Dietary myristic, palmitic, and linoleic acids modulate cholesterolemia in gerbils

Previous studies with cebus monkeys and review of published human data indicated that 85% to 90% of the variation in plasma cholesterol (TC) could be explained on the basis of dietary myristic (14:0) and linoleic (18:2) acid intake in the absence of cholesterol, and that 16:0 contributed to cholesterolemia as dietary cholesterol was increased. Because monkeys are a limited resource, a more convenient, sensitive model was sought for investigating these dietary fatty acid and plasma lipid relationships. Accordingly, this report describes the results of multiple regression analysis of the TC response to dietary fatty acids based on 319 young, male Mongolian gerbils fed a total of 59 purified diets supplying about 40% energy as fat from single or blended fat sources. Gerbils (6-16 animals per dietary group) were fed purified diets (21 with 0.01 to 0.08% cholesterol and 38 cholesterol-free) for 4-week periods. When cholesterol-free diets were fed, dietary 14:0 and 18:2 together accounted for 89% of the observed variation in TC. Although 14:0 consumption increased TC in a linear manner, the independent ability of 18:2 to lower cholesterol was nonlinear and exhibited a threshold effect at 5-6% dietary energy, above which further lowering of TC was less remarkable. In gerbils consuming cholesterol-supplemented diets, 87% of the observed variation in TC could be accounted for by a regression equation that included 14:0, palmitic acid (16:0), and a log function of 18:2 plus dietary cholesterol itself. These results demonstrate the applicability of gerbils for such studies and confirm previous observations in monkeys and humans that dietary 14:0 and 18:2 are the main fatty acids modulating plasma cholesterol under normocholesterolemic circumstances (i.e., when consuming low-cholesterol diets and lipoprotein metabolism is normal) whereas 16:0 also appears modestly hypercholesterolemic when LDL receptors are compromised (i.e., when dietary cholesterol or certain metabolic factors have encumbered lipoprotein metabolism).

Suppressors of nmtl-181, a conditional lethal allele of the Saccharomyces cerevisiae myristoyl-CoA:protein N-myristoyltransferase gene, reveal proteins involved in regulating protein N-myristoylation

Several essential Saccharomyces cerevisiae proteins require myristate to be covalently bound to their amino-terminal glycine for biological activity. Protein N-myristoylation is catalyzed by myristoyl-CoA:protein N-myristoyl-transferase, Nmt1p. nmt1-181 encodes a mutant enzyme with a Gly451-->Asp substitution. nmt181p has a reduced affinity for myristoyl-CoA and produces global defects in protein N-myristoylation at > or = 30 degrees C. nmt1-181 results in growth arrest at various stages of the cell cycle within 1 hr after cells are shifted to > or = 30 degrees C and lethality within 8 hr. The growth-arrest phenotype and loss of viability do not require components of the mating pathway and are associated with lysis sensitivity that may be related to undermyristoylation of two protein phosphatases, Ppz1p and Ppz2p. Growth can be rescued at 30 degrees C by adding myristate or sorbitol to the medium or by removing inosine. Cells can be rescued at 37 degrees C by overexpressing nmt1-181p or Nmt1p or by adding myristate to the medium. Selection of high-copy suppressors of the myristate auxotrophy and lethality observed at 37 degrees C yielded only NMT1, whereas six unlinked suppressors of the myristoylation defect (SMD1-6) were obtained when the screen was conducted at 30 degrees C. The protein products of three SMD loci were identified: (i) cdc39-delta 1.7p, which transactivates NMT1; (ii) Fas1p, the beta subunit of the fatty acid synthetase complex, activates FAS2's promoter and increases myristoylation of Gpa1p; and (iii) Pho5p, the major secreted acid phosphatase produced by this yeast. PHO5 is normally induced when yeast are grown in phosphate-depleted medium. Removal of inorganic phosphate from the medium also rescues nmt1-181 cells at 30 degrees C. PHO5's mechanism of suppression of nmt1-181 appears to involve, at least in part, activation of FAS2 transcription and a resulting effect on FAS1 expression. There is an inverse relationship between cellular N-myristoyltransferase and secreted acid phosphatase activities. These observations provide a potential mechanism for coupling phosphate metabolism with the regulation of myristoyl-CoA synthesis and protein N-myristoylation.

Endothelial nitric oxide synthase membrane targeting. Evidence against involvement of a specific myristate receptor

The endothelial isoform of nitric oxide synthase (ec-NOS) is targeted to the particulate subcellular fraction by means of N-terminal myristoylation. However, the association of ecNOS with the particulate subcellular fraction appears to be dynamically regulated, in that agonist treatment of endothelial cells induces translocation of the enzyme from membrane to cytosol (Michel, T., Li, G., and Busconi, L. (1993) Proc. Natl. Acad. Sci. U.S.A. 90, 6252-6255). cDNA encoding wild-type and myristoylation-deficient mutant (myr-) ecNOS was transcribed and translated in vitro, and we found that the recombinant wild-type but not the myr- mutant protein undergoes myristoylation and is able to associate with biological membranes prepared from diverse cell sources. Treatment of these cell membranes with heat or with trypsin did not affect their ability subsequently to serve as acceptor membranes for the wild-type recombinant enzyme. The wild-type ecNOS, but not the myr- mutant, is able to form stable associations with phospholipid liposomes. We also explored the possibility that a polybasic domain within the ecNOS protein might serve as a secondary structural determinant for ecNOS membrane association and constructed truncation mutants that flank a polybasic domain present in the ecNOS. These truncation mutants, transcribed and translated in vitro or transfected into COS-7 cells, undergo myristoylation and are able to associate with biological membranes in a fashion indistinguishable from the wild type ecNOS. Taken together, these results indicate that ecNOS binding to biological membranes is dependent upon interactions of the N-terminal myristoyl moiety of ecNOS with lipid components of the membrane, and this association does not require a specific membrane protein functioning as a myristate receptor nor the presence of a polybasic domain within the ecNOS.

Conditional lethal expression of the vaccinia virus L1R myristylated protein reveals a role in virion assembly

Within vaccinia virus-infected cells, the product of the L1R open reading frame is covalently modified by myristic acid at the penultimate NH2-terminal glycine residue. Previously we have shown that while the L1R protein is a constituent of both intracellular mature virus particles and extracellular enveloped virions which are released from the infected cell, it is associated exclusively with the primary membranes surrounding the virion core. Given this rather specific localization, it was of interest to study the potential role of this essential gene in virus replication and morphogenesis. To this end, we have constructed a recombinant vaccinia virus in which expression of the L1R gene can be transcriptionally repressed. Without the inducer isopropylthiogalactopyranoside (IPTG), synthesis of the L1R protein was blocked, resulting in a total inhibition of plaque formation. Velocity sedimentation of viral particles labeled in the presence of [3H]thymidine, grown in the absence of IPTG, revealed a substantial reduction in viral DNA incorporation into virions. Likewise, proteolysis of the major core proteins p4a, p4b, and p25K, believed to occur during the final stages of virion maturation, was severely impaired. In the absence of L1R expression, only immature virions could be detected by electron microscopy. Transient expression of a plasmid containing the full-length L1R gene driven by its own promoter was able to complement and rescue the defective phenotype. However, a plasmid bearing a mutation in the myristyl acceptor glycine residue was unable to biologically rescue the recombinant, and the protein was not detected in purified virions.trans complementation using a truncated, myristylated form of the L1R protein partially rescued the defective mutant. Collectively, these data suggest that myristic acid mediates essential interactions of the L1R protein with viral membranes and/or other virion components that lead to the productive assembly, maturation, and release of particles.

Temperature-sensitive lesions in the capsid proteins of the rotavirus mutants tsF and tsG that affect virion assembly

The SA11 rotavirus mutants tsF and tsG contain temperature sensitive (ts) lesions in the capsid proteins VP2 and VP6, respectively, that interfere with their ability to assemble. To understand the nature of their lesions, full-length cDNAs of tsF gene 2 and tsG gene 6 were prepared from viral mRNA by reverse transcription and polymerase chain reaction. Comparative sequence analysis indicated that the ts phenotype of tsF VP2 is due to an Ala-->Asp substitution at position 387. The mutation falls outside of those regions of VP2 previously suggested to be of functional significance and therefore points to a previously unidentified site in VP2 that is important for the assembly of viral cores. Comparative sequence analysis showed that tsG VP6 contains two mutant amino acids, i.e., Thr-10 and His-13, and therefore one or both of these mutations are responsible for the ts phenotype of the mutant VP6. In the case of other group A and group C VP6 sequences, these residues are Ser and Asp, respectively. Characterization of tsG-infected cells by indirect immunofluorescence staining showed that while viroplasmic inclusions are formed at the nonpermissive temperature, the mutant VP6 accumulates in these structures only at the permissive temperature. While influencing intracellular accumulation, the Thr-10-->Ser and His-13-->Asp mutations in tsG VP6 are probably not directly involved in the interaction of VP6 with VP2, as VP6 deletion mutants lacking residues 10 and 13 retain the ability to bind VP2 in vitro. Analysis of VP6 failed to confirm previous reports that the protein was myristylated and thus excludes the possibility that this cotranslational modification is temperature-dependent for tsG VP6. Together, these data suggest that the amino terminus of VP6 plays an essential role in virus assembly in vivo, perhaps by being necessary for the movement of the protein to viroplasmic inclusions, the site of core and single-shelled particle formation.

The Fps/Fes protein-tyrosine kinase promotes angiogenesis in transgenic mice

The fps/fes proto-oncogene encodes a cytoplasmic protein-tyrosine kinase known to be highly expressed in hematopoietic cells. To investigate fps/fes biological function, an activating mutation was introduced into the human fps/fes gene which directs amino-terminal myristylation of the Fps/Fes protein. This mutant, myristylated protein induced transformation of Rat-2 fibroblasts. The mutant fps/fes allele was incorporated into the mouse germ line and was found to be appropriately expressed in transgenic mice, in a tissue-specific pattern indistinguishable from that of the endogenous mouse gene. These mice displayed widespread hypervascularity, progressing to multifocal hemangiomas. High levels of both the transgenic human and endogenous murine fps/fes transcripts were detected in vascular tumors by using RNase protection, and fps/fes transcripts were localized to endothelial cells of both the vascular tumors and normal blood vessels by in situ RNA hybridization. Primary human umbilical vein endothelial cultures were also shown to express fps/fes transcripts and the Fps/Fes tyrosine kinase. These results indicate that fps/fes expression is intrinsic to cells of the vascular endothelial lineage and suggest a direct role of the Fps/Fes protein-tyrosine kinase in the regulation of angiogenesis.

SV40 large T antigen functions at two distinct steps in virion assembly

The SV40 large T antigen mutant 5002 has two amino acid substitutions (L19-F; P28-S) and is defective for productive viral infection as demonstrated by its small plaques that arise very late and by a 100-fold reduced yield of infectious progeny. 5002 replicates viral DNA at the same time postinfection as wild-type SV40, and the production of progeny DNA molecules is only marginally reduced. Furthermore, the viral capsid proteins accumulate to near normal levels following infection with 5002. In this manuscript we report evidence that 5002 infection is blocked at a specific stage of viral assembly. The SV40 viral assembly pathway involves conversion of 75S chromatin complexes to 240S virions. Unlike mutants within the T antigen host range (HR) domain, that are also defective for viral assembly and accumulate 75S particles (Spence and Pipas, 1994), 5002 particles are blocked as 150S previrions containing viral DNA and capsid proteins. We have previously shown that 5002 and HR mutants cooperate to produce viable progeny in trans complementation tests. Thus, by two criteria, SV40 large T antigen encodes two distinct activities that function at different steps in virion assembly.

Efficient particle formation can occur if the matrix domain of human immunodeficiency virus type 1 Gag is substituted by a myristylation signal

Lentiviruses, such as human immunodeficiency virus type 1 (HIV-1), assemble at and bud through the cytoplasmic membrane. Both the matrix (MA) domain of Gag and its amino-terminal myristylation have been implicated in these processes. We have created HIV-1 proviruses lacking the entire matrix domain of gag which either lack or contain an amino-terminal myristate addition sequence at the beginning of the capsid domain. Myristate- and matrix-deficient [myr(-)MA(-)] viruses produced after transient transfection are still able to assemble into particles, although the majority do not form at the plasma membrane or bud efficiently. Myristylation of the amino terminus of the truncated Gag precursor permits a much more efficient release of the mutant virions. While myr(-)MA(-) particles were inefficient in proteolytic processing of the Gag precursor, myristylation enabled efficient proteolysis of the mutant Gag. All matrix-deficient viruses are noninfectious. Particles produced by matrix-deficient mutants contain low levels of glycoproteins, indicating the importance of matrix in either incorporation or stable retention of Env. Since matrix-deficient viruses contain a normal complement of viral genomic RNA, a role for MA in genomic incorporation can be excluded. Contrary to previous reports, the HIV-1 genome does not require sequences between the 5' splice donor site and the gag start codon for efficient packaging.

Flavin supported fatty acid oxidation by the heme domain of Bacillus megaterium cytochrome P450BM-3

Cytochrome P450BM-3 is a fatty acid hydroxylase that consists of a heme domain covalently attached to a diflavin (FMN+FAD) cytochrome P450 reductase domain. The heme and flavin domains can be separately expressed and purified from E. coli recombinant expression systems. Normally P450s require a protein redox partner as a source of electrons. We now have found that the P450BM-3 heme domain can be reduced by NADPH+FMN and that reduced FMN can support the P450 catalyzed hydroxylation of a fatty acid substrate, myristic acid. HPLC profiles show that the "artificial" FMN supported hydroxylation gives the same products as does holo-P450BM-3.

Membrane targeting of the nucleotide exchange factor Sos is sufficient for activating the Ras signaling pathway

Activation of growth factor receptors results in tyrosine autophosphorylation and recruitment of SH2 domain-containing effectors, including Grb2. Grb2 recruitment mediates activation of the Ras nucleotide exchanger Sos by an unknown mechanism. To examine the role of membrane recruitment, we prepared Sos derivatives containing either myristoylation or farnesylation signals. This resulted in plasma membrane targeting of Sos and stimulation of the Ras signaling pathway, including ERK and AP-1 activities leading to oncogenic transformation. Sos derivatives with nonfunctional myristoylation or farnesylation sequences were inactive. Farnesylation of Sos also activated Ras signaling in yeast. In both mammalian cells and yeast, membrane-targeted Sos derivatives lacking the C-terminal region were considerably more active. Therefore, targeting of Sos to the plasma membrane in the vicinity of Ras appears to be the primary mechanism leading to activation of the Ras pathway. A secondary mechanism could involve relief of the inhibitory effect of the Sos C-terminal region.

The palmitoylation status of the G-protein G(o)1 alpha regulates its activity of interaction with the plasma membrane

Plasmids containing cDNAs encoding either the wild-type guanine-nucleotide-binding protein G(o)1 alpha or the palmitoylation-negative cysteine-3-to-serine (C3S) mutant of G(o)1 alpha were transfected into Rat 1 cells, and clones stably expressing immunoreactivity corresponding to these polypeptides were isolated. Clones C5B (expressing wild-type G(o)1 alpha) and D3 (expressing the mutant form) were selected for detailed study. Immunoprecipitation of whole cell lysates of each clone labelled with either [3H]palmitate or [3H]myristate demonstrated incorporation of [3H]myristate into both wild-type and the C3S mutant of G(o)1 alpha, but that incorporation of hydroxylamine-sensitive [3H]palmitate was restricted to the wild type. When membrane and cytoplasmic fractions were prepared from cells of either the C5B or D3 clones, although immunodetection of wild-type G(o)1 alpha was observed only in the membrane fraction, the C3S mutant was present in both membrane and cytoplasmic fractions. Furthermore, a significant proportion of the C3S G(o)1 alpha immunoreactivity was also detected in the cytoplasmic fraction if immunoprecipitation of recently synthesized G(o)1 alpha was performed from fractions derived from cells pulse-labelled with [35S]Trans label. Pretreatment of cells of both clones C5B and D3 with pertussis toxin led to complete ADP-ribosylation of the cellular population of G(o)1 alpha in both cell types, irrespective of whether the polypeptide was subsequently found in the membrane or cytoplasmic fraction following cellular disruption. By contrast, separation of membrane and cytoplasmic fractions before pertussis-toxin-catalysed [32P]ADP-ribosylation allowed modification only of the membrane-associated G(o)1 alpha (whether wild-type or the C3S mutant). This labelling was decreased substantially by incubation of the membranes with guanosine 5'-[beta gamma-imido]triphosphate. No cytoplasmic G-protein beta subunit was detected immunologically, and the non-membrane-associated C3S G(o)1 alpha from D3 cells migrated as an apparently monomeric 40 kDa protein on a Superose 12 gel-filtration column. Membrane-associated wild-type and C3S G(o)1 alpha appeared to interact with guanine nucleotides with similar affinity, as no alteration in the dose-response curves for guanine-nucleotide-induced maintenance of a stable 37 kDa tryptic fragment was noted for the two forms of G(o)1 alpha. Chemical depalmitoylation of membranes of clone C5B with neutral 1 M hydroxylamine caused a release of some 25-30% of each of G(o)1 alpha, Gi2 alpha and Gq alpha/G11 alpha from the membranes. Equivalent treatment of D3 cells caused an equivalent release of Gi2 alpha and Gq alpha/G11 alpha, but was unable to cause any appreciable release of the CS3 form of G(o)1 alpha, which was membrane-bound.

Phosphatidylcholine-specific phospholipase D activity is elevated in v-Fps-transformed cells

Activating the protein-tyrosine kinase of v-Fps results in a rapid increase in diglyceride (DG) in rat fibroblasts. The v-Fps-induced increases in DG were detected only when phospholipids were prelabeled with [3H]-myristate, which is incorporated primarily into phosphatidylcholine (PC). Inhibition of phosphatidic acid (PA) phosphatase (PAP), which converts PA to DG, blocked v-Fps-induced DG production. PA is a primary metabolite of type D phospholipases (PLD). Consistent with these observations, PLD activity was activated in response to the kinase activity of v-Fps. The increased PLD activity was detected only when the cells were prelabeled with the PC-specific [3H]-myristate. These data support the hypothesis that v-Fps-induced DG is derived from PC via the PLD/PAP pathway.

Interaction of Helicobacter pylori and its fatty acids with parietal cells and gastric H+/K(+)-ATPase

Helicobacter pylori and the fatty acids produced by this organism were compared for their acid inhibitory activity in isolated parietal cells and their interaction with gastric H+/K(+)-ATPase. H pylori (intact organisms, sonicates, methanolic extracts, and extracts from culture medium) and the fatty acids cis 9,10-methyleneoctadecanoic acid and tetradecanoic acid inhibited at fairly high concentrations histamine- and dibutyryl cyclic adenosine monophosphate stimulated acid production in isolated parietal cells, dissipated (with a slow onset) the H+/K(+)-ATPase created H+ gradient in gastric membrane vesicles, and inhibited H+/K(+)-ATPase activity in a concentration dependent manner. The inhibitory potency of H pylori and the fatty acids in relation to H+/K(+)-ATPase depended on the amount of membrane protein. Bovine serum albumin prevented enzyme inhibition and proton dissipation from gastric vesicles. The data indicate that H pylori establishes its antisecretory action in parietal cells by blocking H+/K(+)-ATPase activity and also by a detergent action at the apical parietal cell membrane. The fatty acids cis 9,10-methyleneoctadecanoic acid and tetradecanoic acid are probably the acid inhibitory factors secreted by H pylori.

Myristoylation of protein at a distinct position allows its phosphorylation by protein kinase C

A hydrophilic enzyme, lysozyme, was myristoylated in vitro by the N-hydroxysuccinimide ester of myristic acid and three monomyristoylated lysozymes modified at a distinct position (at Lys-13, Lys-33, Lys-97) were isolated by two-step column chromatography. The relationship between membrane binding and phosphorylation by protein kinase C of these monomyristoylated lysozymes were examined using phospholipid vesicles. These three lysozymes bound to phospholipid vesicles to the same extent, whereas the binding of nonmyristoylated native lysozyme was negligible. When native and three monomyristoylated lysozymes were reacted with protein kinase C in a phosphatidylserine (PS)-containing vesicle system, phosphorylation was observed with the myristoylated lysozymes, whereas that of native lysozyme was negligible. However, a remarkable (more than sixfold) difference in the extent of phosphorylation by protein kinase C was observed among three monomyristoylated lysozymes with a different myristoylated position. These results suggest that the membrane binding of substrate protein is not sufficient for the phosphorylation by protein kinase C and the topology of the substrate protein on the membrane play a crucial role in the recognition of substrate protein by protein kinase C. These results further indicate that protein myristoylation can modulate the topology of the membrane-bound protein.

Protein glycosylation and myristylation in Chlorella virus PBCV-1 and its antigenic variants

Chlorella virus PBCV-1 particles contain three glycoproteins, the major capsid protein Vp54 and two minor proteins Vp280 and Vp260. The major capsid protein is myristylated as well as glycosylated. Both modifications are in the carboxyl-terminal portion of the protein. A gene which is modified in a PBCV-1 antiserum-resistant mutant was cloned and sequenced. This gene has an open reading frame of 3099 bases and encodes one of the two large virion glycoproteins (Vp260). Vp260 contains 13 tandem repeats of 61 to 65 amino acids. The mutation deletes the equivalent of four of the amino acid repeat sequences and duplicates one of these sequences.

Agents that increase phosphatidic acid inhibit the LH-induced testosterone production

The results of the present study point to phosphatidic acid (PtdOH) as a possible intracellular messenger, which might be involved in local modulation of testicular testosterone production in vivo. Propranolol (27-266 microM) induced an increased level of [3H]PtdOH in isolated rat Leydig cells, prelabeled with [3H]myristate, and at the same time a strong dose-dependent inhibition of the acute testosterone production stimulated by luteinizing hormone (LH). The inhibition was not bypassed by the addition of dibutyryl-cAMP but was overcome, when 22(R)-hydroxycholesterol was added as a direct substrate for cytochrome P-450 side chain cleavage enzyme. Thus, the inhibition appears to be exerted at a point distal to cAMP-generation but before the first enzyme in the testosterone synthetic pathway. Treatment with other agents (4 beta-phorbol 12-myristate 13-acetate (PMA), A23187, and sphingosine) giving rise to increases in the PtdOH-level resulted in the inhibition of the LH-induced testosterone formation as well, thus indicating a connection between the two effects. Furthermore, we were able to demonstrate a highly significant correlation between the PtdOH-increase and the inhibition of the LH-stimulated testosterone production. This may suggest a causal relationship between these two parameters.

Inhibition of rhodopsin phosphorylation by non-myristoylated recombinant recoverin

Bovine recoverin regulates rhodopsin phosphorylation and controls photoreceptor light sensitivity in a Ca(2+)-dependent manner. Recoverin is post-translationally modified with lipids (myristic acid or related lipids) at its N-terminus. Since with this lipid modification (N-myristoylation), recoverin associates with rod outer segment membranes in a Ca(2+)-dependent manner, N-myristoylation has been suggested to be important for the function of this protein. To study the role of this modification, we obtained recombinant non-myristoylated recoverin in E. coli and studied its functional properties. Here, we report that recombinant non-myristoylated recoverin inhibits rhodopsin phosphorylation at Ca2+ concentrations of 30 nM-10 microM in a similar way as native N-myristoylated recoverin does. Thus, our result showed that N-myristoylation is not essential for the Ca(2+)-dependent inhibition of rhodopsin phosphorylation by recoverin.

Cellular distribution of HIV type 1 Nef protein: identification of domains in Nef required for association with membrane and detergent-insoluble cellular matrix

Cellular distribution of HIV-1 Nef protein was studied by expressing the protein in mammalian cells. Cell extracts were fractionated by low- and high-speed centrifugation and by nonionic detergents. Two Nef-related proteins were expressed in COS cells, Nef-27kD and Nef-25kD. Nef-27kD, an N-myristoylated form of Nef, was found in the cytosol and in association with a particulate fraction of the cytoplasm. Treatment of the particulate cytoplasmic fraction with nonionic detergents, using three different protocols designed to isolate the cytoskeleton matrix, indicated that part of Nef was sensitive and part was resistant to detergent solubilization. These two cellular fractions represent membrane- and cytoskeleton-associated Nef. Nef-25kD, initiated from an in-frame AUG codon, was not modified with myristic acid at the amino terminus. Consequently, this protein was present in a soluble form in the cytosol. Furthermore, a mutant of Nef-27kD, in which the myristoylation signal is deleted, appears as a cytoplasmic soluble protein. To determine domains in Nef that are responsible for its subcellular distribution, successive internal deletions of 14-20 amino acids were introduced at the N-terminal portion of the protein. Five mutants were evaluated with respect to their cellular localization. One mutant (pSVLA-5), from which amino acids 73-88 were deleted, did not copurify with the detergent-insoluble fraction. The protein was, however, present in the particulate cytoplasmic fraction, presumably in association with membranes. Taken together, these results suggest that N-myristoylation of Nef affects its association with both membranes and cytoskeleton.(ABSTRACT TRUNCATED AT 250 WORDS)

The CAAX peptidomimetic compound B581 specifically blocks farnesylated, but not geranylgeranylated or myristylated, oncogenic ras signaling and transformation

Recently developed CAAX peptidomimetic compounds have been shown to be potent and specific inhibitors of farnesyl protein transferase activity and to block the growth of Ras-transformed cells. However, whether this growth inhibitory action is specifically a consequence of blocking oncogenic Ras signaling has not been determined. To address this question, we have utilized mutants of the normally farnesylated oncogenic Ras protein (Ras-F) that are modified by alternative lipids, a geranylgeranyl isoprenoid (Ras-GG) or the fatty acid myristate (Myr-Ras), to determine the specificity of the CAAX peptidomimetic compound, B581. Like Ras-F, both Ras-GG and Myr-Ras are membrane-associated and transforming. Unexpectedly, NIH 3T3 cells transformed by each of the three Ras mutants underwent morphological alteration to a less transformed, but not normal, morphology. However, B581 inhibited the ability of only Ras-F-transformed cells, but not Ras-GG- or Myr-Ras- (or Raf-) transformed cells, to grow in soft agar. Furthermore, although all three lipid-modified versions of Ras stimulated mitogen-activated protein kinase activation, and both Jun and Elk-1 transcriptional activity, B581 inhibited only farnesylated Ras activation of these three downstream components of Ras signaling. Therefore, B581 prevents the growth of Ras-transformed cells by specifically antagonizing Ras-mediated signaling.

Insulin-stimulated hydrolysis of phosphatidylcholine by phospholipase C and phospholipase D in cultured rat hepatocytes

We have investigated the mechanism of action by which insulin increases phosphatidate (PA) and diacylglycerol (DAG) levels in cultured rat hepatocytes. Insulin initially stimulated phosphatidylcholine-dependent phospholipase D (PC-PLD) with a significant increase in both PA and intracellular as well as extracellular choline. The involvement of phospholipase D was confirmed by the formation of PC-derived phosphatidylethanol in the presence of ethanol. The DAG increase appeared to be biphasic. Only the early phase of DAG production was inhibited by propranolol, an inhibitor of the phosphatidate phosphatase (PAP) responsible for the conversion of PA into DAG, suggesting that initially the DAG increase is due to the PLD-PAP pathway. The delayed DAG increase was in parallel with increased intracellular and extracellular phosphocholine and probably derived directly from PC-PLC activity. Experiments performed in the presence of 1 microM phorbol 12-myristate 13-acetate (PMA) indicated that protein kinase C (PKC) mediated the insulin effect on PC-PLC, but not on PC-PLD. These findings were confirmed using the PKC inhibitors calphostin, H7 and staurosporine. The dual activation of these phospholipases with a biphasic elevation of DAG levels and activation of specific PKC isoenzymes could be necessary to elicit both early and delayed effects of insulin.

Increased phospholipase D activity in multidrug resistant breast cancer cells

An adriamycin-resistant MCF-7 cell line (300-fold increased resistance vs. wild type MCF-7) showed an increased phorbol ester-stimulated phospholipase D activity of approximately 4-6 times over that found in wild type cells. Phospholipase D activity was assessed by monitoring the mass of phorbol ester-induced phosphatidylethanol. The cellular phospholipase D activity was time- and phorbol ester-concentration-dependent and was obvious whether measured as an increase in the mass of PEt or in the production of 3H-labeled phosphatidylethanol in cells prelabeled with [3H]myristic acid. Phorbol ester also stimulated increases in the production of the mass of cellular diacylglycerol and phosphatidic acid in the adriamycin-resistant cells vs. the wild-type cells. Tests with a series of drug-resistant MCF-7 cell lines revealed a positive correlation between increased drug resistance and phorbol ester-stimulated phospholipase D activity.

A glycoprotein binding retinoids and fatty acids is present in Drosophila

In the search for a possible Drosophila melanogaster homolog of interphotoreceptor retinoid-binding protein (IRBP), a approximately 140-kDa retinoid- and fatty acid-binding glycoprotein found in vertebrates, the 110,000 g supernatant fraction prepared from homogenates of fly heads was analyzed for the presence of proteins capable of binding radiolabeled retinol and palmitic acid. A soluble protein, which binds concanavalin A and has a retention time on size-exclusion high-performance liquid chromatography identical to that of purified bovine IRBP, was identified as binding both ligands. As assessed by fluorescence titration, the protein fraction obtained by concanavalin A-Sepharose affinity chromatography and size-exclusion chromatography of fly head supernatant had apparent dissociation constants of 2.9 x 10(-7) +/- 0.6 M for all-trans retinol, with the number (n) of independent ligand binding sites per protein molecule = 2, and 3.5 x 10(-7) +/- 0.1 M for 16-[9-anthroyloxy] palmitic acid with n = 7. High-performance liquid chromatography of hexane extracts of this protein fraction resolved several peaks with polarity and relative retention times similar, but not identical to all-trans retinol and retinal and their 9-, 11-, and 13-cis isomers. Gas chromatography/mass spectrometry analysis of fatty acid methyl esters prepared following lipid extraction of the protein identified lauric, myristic, palmitic, palmitoleic, and oleic acids as being covalently bound. Laurate, myristate, palmitate, and stearate were noncovalently bound. The apparent molecular mass of the Drosophila protein as assessed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and silver staining of the retinoid- and fatty acid-binding peak obtained by hydrophobic interaction chromatography of the size-exclusion fraction was approximately 70 kDa.

Membrane interaction of small N-myristoylated peptides: implications for membrane anchoring and protein-protein association

The effect of the covalent attachment of a myristolyl moiety to the N-terminal glycine residue in proteins, N-myristoylation, on lipid-protein interactions was investigated in a model system using magnetic resonance spectroscopic methods. Two peptides with sequences conserved among known N-myristoylated proteins were chosen for this study. Using two-dimensional nuclear magnetic resonance techniques, it was shown that N-myristolylation results in an aggregation of both peptides in solution, although they lack well defined folded conformations in solution either when chemically N-myristolyated or when nonacylated. The interaction of the acylated peptides with lipid bilayers was investigated using spin label electron spin resonance and 2H NMR techniques. The results show that when bound to membranes, the covalently linked myristoyl chain of one of the peptides is directly inserted into or anchored to the lipid bilayer. The binding of the other peptide with membranes is effected by interactions between amino acid residues and the phospholipid headgroups. In this case, the covalently linked myristoyl moiety is most likely not in direct contact with the acyl chains of the host lipid bilayer. Rather, the N-myristoyl chains stabilize the peptide aggregate by forming a hydrophobic core. Measurements of peptide binding to membranes showed that N-myristoylation affects both the lipid:peptide stoichiometry at saturation and the equilibrium binding constant, in a manner that is consistent with the structural information obtained by magnetic resonance methods.

N-terminally myristoylated Ras proteins require palmitoylation or a polybasic domain for plasma membrane localization

Plasma membrane targeting of Ras requires CAAX motif modifications together with a second signal from an adjacent polybasic domain or nearby cysteine palmitoylation sites. N-terminal myristoylation is known to restore membrane binding to H-ras C186S (C-186 is changed to S), a mutant protein in which all CAAX processing is abolished. We show here that myristoylated H-ras C186S is a substrate for palmitoyltransferase, despite the absence of C-terminal farnesylation, and that palmitoylation is absolutely required for plasma membrane targeting of myristoylated H-ras. Similarly, the polybasic domain is required for specific plasma membrane targeting of myristoylated K-ras. In contrast, the combination of myristoylation plus farnesylation results in the mislocalization of Ras to numerous intracellular membranes. Ras that is only myristoylated does not bind with a high affinity to any membrane. The specific targeting of Ras to the plasma membrane is therefore critically dependent on signals that are contained in the hypervariable domain but can be supported by N-terminal myristoylation or C-terminal prenylation. Interestingly, oncogenic Ras G12V that is localized correctly to the plasma membrane leads to mitogen-activated protein kinase activation irrespective of the combination of targeting signals used for localization, whereas Ras G12V that is mislocalized to the cytosol or to other membranes activates mitogen-activated protein kinase only if the Ras protein is farnesylated.

Posttranslational acylation of the transferrin receptor in LSTRA cells with myristate, palmitate and stearate: evidence for distinct acyltransferases

When incubated with [3H]myristate or [3H]palmitate, LSTRA cells, a murine T cell line, incorporated radiolabel into a protein of 95 kDa as analyzed by SDS-polyacrylamide gel electrophoresis. This dually acylated protein was identified as the transferrin receptor by immunoprecipitation with a monoclonal anti-transferrin receptor antibody. Acylation of the transferrin receptor was posttranslational and occurred via ester or thioester linkages. Analysis of radiolabeled transferrin receptor protein from [3H]myristate-labeled cells by acid hydrolysis followed by thin layer chromatography revealed the exclusive presence of [3H]myristate. Labeled transferrin receptor protein from [3H]palmitate-labeled cells contained predominantly [3H]stearate and smaller amounts of [3H]palmitate. This is in contrast to the protein-tyrosine kinase p56lck, which in [3H]palmitate-treated LSTRA cells, incorporated primarily [3H]palmitate. An analog of myristic acid, 5-nonanyloxyfuran-2-carboxylic acid, inhibited the incorporation of [3H]myristate, but not [3H]palmitate or [3H]stearate into transferrin receptor protein, suggesting that these acylation events are distinct. These studies indicate that the murine transferrin receptor is acylated posttranslationally with myristate, palmitate and stearate and suggest that more than one acyltransferase activity is responsible for its acylation.