The role of macrophages in the regulation of inflamation and immunity is, in part, due to their secretory repertoire. Among the important mediators released by macrophages are the products of both the cyclooxygenase and lipoxygenase pathways of arachidonic acid (20:4) metabolism. The principal focus of this paper is the mechanism by which bacterial lipopolysaccharides (LPS) regulate 20:4 metabolism in macrophages. LPS has the capacity to prime macrophages for greatly enhanced 20:4 metabolism when the cells are subsequently challenged with a spectrum of triggers. Concomitant with priming, LPS also promotes the covalent attachment of myristic acid to a set of macrophage proteins. The time and concentration dependence of LPS-induced protein myristoylation is consistent with a role for myristoylation in LPS priming of the 20:4 cascade. One of the myristoylated proteins is a 68K (K = 103Mr) protein kinase C substrate which associates with membranes upon myristoylation. LPS-primed macrophages show greatly increased phosphorylation of the 68K protein when the cells are subsequently treated with protein kinase C activating phorbol esters. It is proposed that the myristoylation of the 68K protein promotes its attachment to the membrane where it is more closely associated with activated protein kinase C (PKC), an association which would ensure more efficient catalysis during the mobilization and oxygenation of 20:4.

This paper also examines protein myristoylation during T-cell-mediated activation of macrophages. Immune-activated macrophages have an enhanced capacity to kill several infectious agents by oxidative mechanisms. The lymphokine γ-interferon (IFNγ) rapidly induces the myristoylation of a 48K protein. This 48K protein is also myristoylated in murine macrophages that have been activated in vivo by intraperitoneal injection of Corynebacterium parvum, suggesting that it may be an important intermediate in the activation of macrophages for enhanced microbicidal capacity.

During the course of our investigation into the molecular mechanism by which bacterial lipopolysaccharide (LPS) primes macrophages for enhanced arachidonic acid (20:4) metabolism we discovered that LPS also promotes the covalent attachment of the 14-carbon fatty acid, myristic acid, to a select group of macrophage proteins, including a major, specific substrate of protein kinase C. While the focus of this paper is on stimulus-dependent protein myristoylation in the murine peritoneal macrophage, I have also included information to situate these acylation reactions within a broader context. The chapter therefore begins with a discussion of 20:4 metabolism in macrophages and its modulation by LPS. An account of protein myristoylation in macrophages and its effects on protein kinase C dependent phosphorylation follows. Protein myristoylation during immune activation of macrophages is considered and finally, the function and enzymology of myristoylation in other cells are briefly discussed.

Macrophages are a major source of arachidonic acid (20:4) metabolites, which are important mediators of inflammation (Davies et al. 1984). When macrophage phagocytic receptors interact with particulate ligands such as immune complexes or bacteria, phospholipases are activated and 20:4 is released from membrane phospholipids. The free 20:4 is then oxygenated via the cyclooxygenase pathway to prostaglandin E2 (PGE2) and prostacyclin (PGI2), or along the lipoxygenase pathway to leukotriene C (LTC) and hydroxyeicosatetranoic acids (HETEs) (Scott et al. 1980; Bonney et al. 1978, 1979; Rouzer et al. 1980).

The cyclooxygenase products (prostaglandins, prostacyclin and thromboxane) have been shown to influence smooth muscle tone, platelet aggregation, lymphocyte function, myelopoiesis, vascular permeability as well as pain and fever. The lipoxygenase product leukotriene B (LTB) mediates neutrophil aggregation, secretion and adhesion to vascular endothelium, as well as chemotactic activity for a variety of white blood cells: LTC mediates smooth muscle contraction and increased vascular permeability (reviewed in Davies et al. 1984).

Regulation of the lipoxygenase and cyclooxygenase pathways in macrophages is exerted at a number of levels. Briefly, the amount and types of 20:4 metabolites vary with the stimulus, the activation state of the cell, the tissue of origin and the species. For example:

(1) The ratio of cyclooxygenase to lipoxygenase metabolites secreted by macrophages depends on the agonist. Thus the soluble agonist, phorbol myristate acetate (PMA), stimulates the release of only cyclooxygenase products, while particulate stimuli such as zymosan and immune complexes trigger the release of both cyclooxygenase and lipoxygenase metabolites (Humes et al. 1982; Tripp et al. 1985; Aderem et al. 1986a). The profile of 20 :4 metabolites produced appears to be related to the capacity of the stimulus to trigger Ca2+ transients (Tripp et al. 1985; Aderem & Cohn, 1986, 1988) and will be discussed in detail below.

(2) Murine resident peritoneal macrophages have low microbicidal and tumoricidal activity and secrete large quantities of 20:4 metabolites, while BCG-activated peritoneal macrophages have high microbicidal and tumoricidal capacity and secrete small amounts of 20:4 metabolites (Scott et al. 1982). Resident cells secrete predominantly PGI2, PGE2 and LTC while BCG-activated macrophages only release small quantities of thromboxane B2 (TxB2) and PGE2 (Scott et al. 1982).

(3) The profile of 20:4 metabolites also varies with tissue of origin. While both murine resident peritoneal macrophages and pulmonary tissue macrophages secrete LTC, only peritoneal macrophages secrete significant quantities of PGE2 and PGI2 (Rouzer et al. 1982).

(4) There are differences in 20:4 metabolites secreted by macrophages of different species. For example, in contrast to the mouse, the major products secreted by human monocytes are TxB, LTB, LTC and HETEs white PGE2 is produced in very low concentration and prostacyclin is completely absent (Pawlowski et al. 1983).

(5) Bacterial lipopolysaccharide has the capacity to prime macrophages for enhanced 20:4 metabolism when the cells are subsequently stimulated (Aderem et al. 1986a). While the amount of 20:4 metabolites released to the medium is greatly increased, LPS treatment does not alter the proportion of cyclooxygenase and lipoxygenase products (see below).

LPS induces the secretion of a large number of macrophage products including plasminogen activator, tumour necrosis factor, interleukin-1 and granulocyte macrophage colony stimulating factor (GM-CSF) (for reviews see Morrison & Ulevitch, 1978; Homma et al. 1984). Furthermore, LPS primes macrophages (Pabst & Johnston, 1980) and neutrophils (Guthrie et al. 1984) for enhanced release of reactive oxygen intermediates on subsequent challenge with a variety of triggers.

Little is known about the mechanisms by which LPS stimulates these complex responses in macrophages. Early cellular responses to LPS treatment include inositol lipid turnover (Ogmundsdottir & Weir, 1979), protein phosphorylation (Weiel et al. 1986) and the synthesis of specific proteins (Hamilton et al. 1986). We have recently reported that LPS promotes the myristoylation of three macrophage proteins (Aderem et al. 1986b), including a major protein kinase C substrate. We describe data below that suggest that these novel transacylation reactions might constitute a major transduction pathway for LPS-induced response.

LPS is by itself a poor trigger of the 20:4 cascade (Fig. 1); (Aderem et al. 1986b). However, pre-incubation of cells for 30-60 min with LPS (10 ng ml-1) enhances both the amount and the rate of 20:4 secretion in response to a second stimulus. For example, PMA-induced 20:4 secretion is increased 3-to 15-fold in cells pretreated with LPS (Fig. 1). Similarly, LPS also synergizes with zymosan (yeast cell wall), immune complexes and the Ca2+ ionophore, A23187 (Aderem et al. 1986a). When unprimed cells are stimulated, secretion of 20:4 commences only after a lag phase. This lag phase is significantly shortened in LPS-treated cells (Fig. 1, inset). Treatment with LPS thus speeds and potentiates 20:4 release presumably by establishing one or more preconditions necessary for activation of the phospholipase.

Further support for this hypothesis is provided by the ability of LPS to render macrophages responsive to latex particles. Latex particles are ingested by macrophages, but without the release and metabolism of 20:4. However, when cells are primed with LPS for at least 45 min before ingestion of the latex beads, large amounts of PGI2 and PGE2 are secreted (Fig. 2).

These and other data led us to postulate that activation of 20:4 metabolism in macrophages requires two distinct steps; ? priming step followed by a triggering step (Aderem et al. 1986a). Receptor-mediated particulate stimulation of the 20:4 cascade is capable of both priming and triggering, LPS is capable of priming and latex particles can only generate the triggering step. We suggest below that the second, or triggering step, is an increase in intracellular Ca2+, and that the first, or priming step, is related to stimulus-dependent myristoylation of proteins involved in signal response coupling.

Since the phagocytosis of latex is accompanied by an increase in intracellular calcium, and latex is only capable of generating the triggering signal, we tested the possibility that an increase in intracellular calcium constituted the triggering signal (Aderem & Cohn, 1988). Low concentrations (0·l μM) of the calcium ionophore, A23187, do not trigger the 20: 4 cascade in macrophages. However, when the cells are first primed with LPS (10 ngml-1) for 60min and then challenged with A23187, a rapid release of a large amount of 20:4 metabolites ensues. The order in which stimuli are added is critical. If the cells are first challenged with A23187 and subsequently treated with LPS, no release of 20:4 metabolites is observed.

To be sure that the calcium, and not A23187, was critical for the triggering step, cells were cultured in Ca2+-free medium and exposed sequentially to LPS and A23187. In the absence of external Ca2+ very little 20:4 was secreted into the medium. After the addition of Ca2+ to the external medium a rapid release of 20:4 metabolites occurs, confirming that an increase in intracellular Ca2+ triggers 20:4 metabolite secretion in the LPS-primed cell.

A second role for calcium in the control of macrophage 20:4 metabolism has been found. It is to increase the activity of the lipoxygenase pathway. Low concentrations of LPS (10 ngml-1) or A23187 (0·l μM) do not promote the secretion of any 20:4 metabolites. High concentrations of LPS (1 μgml-1) cause the release of cyclooxygenase products only. However, when macrophages are treated sequentially with LPS and A23187 both cyclooxygenase and lipoxygenase pathways are activated, LTC production being increased 150-fold (Aderem & Cohn, 1988).

A similar shift in profile of 20:4 metabolite secretion has been reported for macrophages treated with PMA and A23187 (Tripp et al. 1985; Aderem & Cohn, 1986). While PMA alone stimulates macrophages to secrete cyclooxygenase products, a combination of PMA and A23187 promotes the release of large amounts of LTC as well. Therefore, increasing the Ca2+ concentration within macrophages with A23187 results in the activation of the lipoxygenase pathway.

These data imply that the cyclooxygenase and lipoxygenase pathways have different Ca2+ requirements. We have shown this to be true by using LPS-primed macrophages that have been treated with A23187 in the absence of external Ca2+ as a model system in which to titrate the Ca2+ concentration dependence of each pathway. Neither cyclooxygenase nor lipoxygenase metabolites are secreted in the absence of Ca2+ in the medium. This is consistent both with our previous finding that 20:4 metabolite secretion in macrophages is completely dependent on extracellular Ca2+ regardless of the nature of the stimulus (Aderem et al. 1986c), and with the requirements of macrophage phospholipases for Ca2+ (Wightman et al. 198la,b). When Ca2+ is titrated into the system, cyclooxygenase products are detected at one order of magnitude lower Ca2+ concentration in the medium than are lipoxygenase metabolites.

The data also suggest that Ca2+ regulates macrophage 20:4 metabolism at two distinct steps. First, Ca2+ is required for the activation of the phospholipases that promote the release of 20:4 from the membrane phospholipid. Second, relatively higher concentrations of Ca2+ are required both to activate the 5-lipoxygenase enzyme and to promote its reversible association with the membrane fraction (Rouzer & Samuelsson, 1987). Thus agonists such as PMA, which mobilize relatively small amounts of intracellular Ca2+ (Di Virgilio et al. 1984), promote the secretion of only cyclooxygenase products, whereas particulate agonists such as immune complexes, which stimulate rather larger increases in intracellular Ca2+ (Young et al. 1984), cause the secretion of both cyclooxygenase and lipoxygenase products.

While investigating the biochemical mechanism of LPS priming of macrophages for enhanced 20:4 metabolism, we discovered that LPS stimulates the selective myristoylation of proteins with molecular mass of 68K (K = 103Mr) and a doublet of 36-42K (Aderem et al. 1986b) (Fig. 3) The myristate is linked to the proteins through amide bonds, since it is not released by treatment with hydroxylamine. The reaction appears to be specific for myristic acid, as palmitate and arachidonate are not incorporated into these proteins.

The temporal response and LPS concentration dependence of the myristoylation reaction is compatible with it having a role in LPS-dependent priming of macrophages for enhanced 20:4 metabolism (Aderem et al. 1986a,b).

We postulated above that agonists, such as zymosan, that promote 20:4 secretion from macrophages must be capable of generating both the priming and triggering signals. It is of interest, therefore, that zymosan also promotes the myristoylation of three proteins with molecular masses identical to those acylated in response to LPS (Aderem et al. 1986b).

We have examined the effect of protein synthesis inhibitors on the myristoylation of macrophage proteins and on 20:4 secretion. A combination of cycloheximide (5 μgml-1) and emetine (5 μgml-1), which inhibits macrophage protein synthesis by 97 % (Aderem et al. 1986c), inhibits protein myristoylation at the earliest time point measurable (<5min). This suggests either that myristoylation is coupled to translation or that the myristoyl transferase is turning over very rapidly. Regardless of which of these two possibilities is correct, it is of interest that 20:4 secretion is also inhibited by protein synthesis inhibitors on the same time scale (Aderem et al. 1986c; Bonney et al. 1980). The exquisite sensitivity of 20:4 metabolism to protein synthesis inhibitors could be explained if protein myristoylation is an intermediate in the activation of 20:4 metabolism.

The 68K protein whose myristoylation is induced by LPS is similar or identical to the 80/87K protein, a major specific substrate for protein kinase C found in brain and fibroblasts (Wu et al. 1982; Albert et al. 1986; Blackshear et al. 1986; Rozengurt et al. 1983; for reviews see Neidel & Blackshear, 1986; Woodgett et al. 1987). The macrophage 68K protein can be immune precipitated with a polyclonal rabbit antiserum raised against the “87 kD” major specific protein kinase C substrate of bovine brain (Aderem et al. 1988a). Further evidence that the 68K macrophage protein is a protein kinase C substrate is provided by the observation that PMA triggers its phosphorylation. The Staphylococcus aureus V8 protease map of the phosphorylated 68K murine protein is identical to that of the bovine 87K phosphoprotein. The two-dimensional phosphopeptide maps following thermolysin digestion of the murine macrophage and bovine brain proteins are also identical. When subjected to isoelectric focusing, the phosphorylated and myristoylated protein from murine macrophages and the phosphorylated proteins from murine, rat and bovine brain were found to be acidic with a pl of approximately 4·5.

The apparent molecular mass of the brain and fibroblast 80/87K protein has been reported to vary with the species and with the analytical gel system used (Albert et al. 1986, 1987; Blackshear et al. 1986; Neidel & Blackshear, 1986; Woodgett et al. 1987). To monitor the electrophoretic behaviour of the PKC substrate protein from a variety of cells and tissues, the mouse macrophage, mouse brain, rat brain and bovine brain proteins were electrophoresed together in two gel systems. On a 6%-12% linear gradient SDS-polyacrylamide gel (Neville’s buffer), the apparent molecular mass values obtained were 68K, 69K, 70K and 76K, respectively. On an 8% SDS-polyacrylamide gel (Laemmli buffer) the molecular mass values were 79K, 80K, 83K and 87K respectively. The variability in apparent molecular mass observed with different gel systems may be explained by the results of hydrodynamic studies of the bovine brain 87K protein which show a calculated molecular mass of 68K, and suggest that the protein is an asymmetric, highly elongated monomer (Albert etal. 1987).

We have compared the distribution, between the membrane and cytosolic fractions, of the myristoylated and phosphorylated 68K PKC substrate. The major portion (>90%) of the myristolated 68K protein was found in the membrane fraction, whereas the major portion (>75 %) of the phosphorylated 68K protein was observed in the cytosolic fraction. As a working hypothesis we propose the following sequence of events to explain the observed distributions of the myristoylated and phosphorylated forms of the 68K protein (Fig. 4). (1) LPS promotes the myristoylation of the 68K protein. (2) The myristoylated protein moves to the membrane where it is more closely associated with PKC, which is known to be active at the membrane (Nishizuka, 1986). (3) PKC phosphorylates the 68K protein. (4) The phosphorylated 68K protein is then de-myristoylated and released to the cytoplasm. This hypothetical sequence of events is supported by several lines of evidence. First, in all systems described thus far myristoylation appears to occur co-translationally or very soon after translation (Wilcox et al. 1987; for a review see Sefton & Buss, 1987). This also appears to be the case in macrophages, since LPS-induced myristoylation of the 68K protein is rapidly blocked by protein synthesis inhibitors as described above. Second, in at least two instances protein-bound myristic acid has been shown to have a role in targeting proteins to the membrane. Myristoylation of p60src and Moloney murine leukemia virus p65gag protein is necessary for plasma membrane association, transforming potential and viral budding (Pellman et al. 1985; Cross et al. 1984; Kamps et al. 1985; Rein et al. 1986). Third, LPS pre-treatment of macrophages results in greatly increased PMA-dependent phosphorylation of the 68K protein and also results in a shift in the PMA dose-response curve such that half-maximal phosphorylation of p68 in LPS-treated cells is obtained with concentrations of PMA 10-fold less than those required in macrophages not treated with LPS (A. Rosen & A. Aderem, unpublished results). Fourth, pulse-chase experiments with [3H]myristic acid, 32Pi and [35S]methionine suggest that PMA induces both the phosphorylation and de-myristoylation of the 68K protein (A. Rosen & A. A. Aderem, preliminary results).

The myristic acid moiety of a protein could facilitate protein association with membranes by simple insertion into the membrane. However, myristic aciddependent targeting of the protein to the membrane appears more selective. Thus myristoylated p60src associates only with the plasma membrane, suggesting that a receptor for myristoylated proteins might exist (Pellman et al. 1985). Protein-protein interaction also appears to be important since myristoylated p60src preferentially associates with a detergent-resistant cellular fraction in fibroblasts (Hamaguchi & Hanafusa, 1987) and the myristoylated 68K protein is associated with a detergentresistant fraction in macrophages (A. Rosen & A. Aderem, unpublished results). Phosphorylation of the myristoylated 68K protein might interrupt its association with the membrane without necessarily promoting its de-myristoylation. Mechanisms by which this might be affected include the introduction of a phosphate group near the site of interaction or a conformational transition of the protein.

Two lines of evidence suggest that LPS is capable of influencing PKC-dependent pathways within macrophages. First, treatment of macrophages with LPS results in increased phosphorylation of the 68K PKC substrate. However, LPS-depen-dent phosphorylation of the 68K protein occurs more slowly than does PMAdependent phosphorylation (half-maximal phosphorylation with LPS in 60min; half maximal phosphorylation with PMA in 7min). Furthermore, the maximal level of phosphorylation obtained with LPS is approximately 50% of that obtained with PMA. Second, the synergy between LPS and A23187 in activating the lipoxygenase pathway can be mimicked by PMA, a known activator of PKC. LPS could either activate PKC directly or potentiate its activity in an alternative way.

It has been reported that LPS does activate partially purified PKC in vitro (Wightman & Raetz, 1984). In this system, it is possible that LPS is substituting for phosphatidyl serine, a necessary cofactor of PKC. The structure of lipid A resembles this phospholipid, and the classic inhibitor of LPS action (Morrison & Ulevitch, 1978), polymyxin B, is known to inhibit PKC activity by neutralizing phosphatidyl serine (Mazzei et al. 1982). However, we have shown that LPS does not activate highly purified PKC (A. Aderem & K. Albert, unpublished observation).

Since LPS does not activate purified PKC directly, we favour the hypothesis that LPS treatment potentiates PKC activity in response to other signals. Stimulation of protein kinase C activity in intact cells causes a shift in the subcellular location of the enzyme from the cytosol to the membrane (Nishizuka, 1986). In macrophages, as mentioned above, LPS-induced myristoylation of the 68K PKC substrate promotes its association with the membrane. LPS might therefore potentiate PKC activity by promoting subcellular association of PKC and its substrates. This hypothesis is supported by the observation that LPS-primed cells show greatly increased phosphorylation of the 68K protein when the cells are subsequently challenged with PMA. The observation that LPS-dependent activation of the PKC pathway is much slower than the PMA response also supports the ‘potentiation’ hypothesis.

Prolonged exposure to PMA results in many cell types becoming partially or completely deficient in specific phorbol ester binding sites, immunoreactive PKC and PKC activity (for a review see Neidel & Blackshear, 1986). This down-regulation has been exploited to assess the role of PKC in a variety of cellular events. We found that macrophages that had been treated with 250 nM-PMA for 24 h showed a 95 % decrease in [3H] phorbol dibutyrate binding, no detectable immunoreactive PKC, and no PMA-dependent phosphorylation of the 68K specific PKC substrate. PMA-dependent activation of 20:4 metabolism was completely inhibited, suggesting that PMA stimulates 20:4 metabolism via a PKC-dependent pathway.

Treatment of macrophages with LPS for 20-40 min results in the cells becoming primed for greatly enhanced 20:4 metabolism when subsequently challenged with a variety of stimuli. Coincident with the establishment of the primed state, LPS also promotes the myristoylation of three proteins. One of these proteins (68K) is a major substrate for PKC. Myristoylation of this protein directs it to the membrane where it is more closely associated with activated PKC (Fig. 4). This multimolecular assembly at the membrane results in greater catalytic efficiency and potentiated 20:4 metabolism.

It is becoming evident that a number of cellular and retroviral proteins have the capacity to become myristoylated. Most of these proteins have a role in cellular regulation, and they include the catalytic subunit of cAMP-dependent protein kinase (Carr et al. 1982), calcineurin B, a component of a calmodulin-binding phosphatase (Aitken et al. 1982), NADH-cytochrome b5 reductase (Ozols et al. 1984), a Mτ 56 000 tyrosine protein kinase expressed in the murine lymphoma cell line LSTRA (Marchildon et al. 1984) and p60src, the tyrosine kinase encoded by the Rous sarcoma virus (Schultz et al. 1985; Buss & Sefton, 1985).

The function of protein myristoylation is only beginning to be understood. Acylation of p60src is essential for association of the protein with the plasma membrane and for expression of this kinase’s transforming potential (Cross et al. 1984; Kamps et al. 1985). Since non-myristoylated p60v-src retains its tyrosine kinase activity (Cross et al. 1984), it is probable that myristoylation functions to target the enzyme to the plasma membrane where it exerts its biological effects. The myristoylated retroviral gag proteins may exploit this cellular targeting mechanism to localize viral budding since the myristoylation of the Moloney murine leukemia virus Pr65gag protein is essential for membrane association and virus particle formation (Rein et al. 1986). However, some myristoylated proteins appear to be soluble (Carr et al. 1982; Olson et al. 1985), suggesting that protein acylation may serve functions other than facilitating protein association with membranes.

We have not yet characterized the myristoyl transferase(s), the enzyme(s) that covalently attach myristic acid to macrophage proteins. However, Towler and colleagues have recently purified an N-myristoyl transferase (NMT) from Saccharomyces cerevisiae, which catalyses the covalent attachment of myristic acid to synthetic peptides whose sequences were derived from those of known N-myristoyl proteins (Towler et al. 1987a,b; Towler & Glaser, 1986). The native NMT is a 55K monomer, which has a strict specificity for myristoyl CoA as the acyl donor. The enzyme activity is present in both crude membrane and soluble fractions prepared from yeast, as well as BC3Hl murine muscle cells (D. A. Towler, personal communication), and has a slightly alkaline pH optimum which suggests that it is not lysosomal. The purified NMT does not exhibit intrinsic methionase activity, suggesting that another enzyme must first remove the initiator methionine from the nascent apo-myristoyl proteins before myristoylation can occur.

NMT exhibits a high degree of selectivity for the sequence of its substrate peptide and a consensus sequence required for effective myristoylation has been elucidated (Towler et al. 1987a). The following rules apply: (1) NMT has an absolute requirement for amino-terminal Gly in its peptide substrates. (2) Neutral amino acids (Asn, Ser, Gin, Ala, Vai) are allowed at position two while charged residues and Pro are not. Peptides with aromatic or bulky hydrophobic residues at position two bind to the enzyme but are poor substrates. (3) A wider spectrum of amino acids can be accommodated at positions three and four in NMT substrates, although peptides with uncharged residues at position three are bound with greater affinity. (4) Position five is critical in NMT-ligand interactions, with Ser5 being a highly favoured residue. A charged residue such as Asp5 renders a peptide totally inactive as a substrate. Peptides with Pro at positions five or six do not function as NMT substrates. (5) Residues beyond the first six amino acids play a role in substrate recognition by NMT but the precise structural requirement in this region of NMT ligands remain to be elucidated.

The consensus data described above are supported by the sequences of known myristoylated proteins. Since the substrate specificity of the NMT is phylogenically conserved, it is possible to identify potential N-myristoyl proteins by examining eukaryotic sequences. For example, it was shown that synthetic peptides derived from the N termini of the HTLV-III gag protein and from the guanine nucleotidebinding proteins (G proteins), Gα,i, could be myristoylated in vitro by NMT. Later it was noted that the gag and Gαi proteins were myristoylated in situ (Buss et al. 1987; Paul et al. 1987). It is also possible that some myristoyl proteins may be acylated at a cryptic site exposed by proteolytic cleavage.

It should be borne in mind that the NMT from yeast was isolated using a peptide substrate derived from the sequence of known myristoylated proteins. The method is therefore likely to select for an enzyme whose specificity approximates the sequence of this peptide and it raises the possibility that other myristoyl transferase enzymes may exist which have different substrate specificities. However, the observation that all the known myristoylated proteins contain the consensus sequence seems to argue against the existence of multiple myristoyl transferases.

The elucidation of a consensus sequence required for myristoylation allows one to predict whether a protein is a candidate for myristoylation or not. Thus, the prediction by Towler et al. that the G proteins and the gag protein of HIV would be myristoylated has been borne out (Buss et al. 1987; Paul et al. 1987). However, certain proteins that contain the myristoylation sequence have been found to be nonmyristoylated. For example, a subset of G proteins, Gα45 and Gα52, contain the N-terminal glycine but are not myristoylated (Buss et al. 1987). This is a very interesting observation since it suggests either that myristoylation of a consensus site does not occur obligatorily or that de-myristoylation can occur.

The palmitic acid attached to the transferrin receptor has been shown to turnover more rapidly than the receptor protein (Omary & Trowbridge, 1981). Whether myristic acid is removed enzymatically from acylated proteins prior to the degradation of the protein has yet to be addressed. In the case of p60src, the half-life of the myristoyl group is the same as that of the polypeptide, suggesting that this protein is not de-acylated (Busset al. 1984). However, we have observed that PMA-dependent phosphorylation of the 68K protein alters its subcellular location from the membrane to the cytosol, and preliminary data suggest that PMA-dependent de-acylation might accompany this change in subcellular distribution (A. Rosen & A. Aderem, unpublished results).

The intracellular events leading to the activation of macrophages by antigen-specific T cells are largely unknown. During cell-mediated immunity, macrophages acquire an increased capacity to secrete reactive oxygen intermediates and to kill microbes and tumour cells. Treatment of macrophages with γ interferon (IFNγ), and with other cytokines like interleukin-4 and GM-CSF, induces properties that are similar to those of macrophages activated in situ (Nathan et al. 1983; Schreiber et al. 1983).

We have investigated the effect of IFNγ on macrophage protein myristoylation (Aderem et al. 1986b). Treatment of murine resident peritoneal macrophages with recombinant IFNγ (10 Uml-1) greatly enhances the myristoylation of a 48K protein. This effect is observed within 1 h and is maximal by 3 h. Neither IFNα nor IFNβ induce the myristoylation of the 48K protein, suggesting that this event is related to the macrophage activating capacity of IFNγ, and not to its antiviral activity. IFNγ-induced myristoylation of the 48K protein appears to be specific for macrophages since it does not occur in human umbilical vein endothelial cells or murine 3T3 fibroblasts.

Consistent with it having a role in macrophage activation, the 48K protein is constitutively myristoylated in macrophages activated in vivo by intraperitoneal injection of heat-killed C. parvum. The identity of the 48K protein is not yet known. A number of workers have described a 47K protein that is phosphorylated in normal neutrophils upon exposure to a variety of activators of the respiratory burst. This 47K protein remains unphosphorylated in neutrophils from patients with autosomal recessive chronic granulamatous disease, which are incapable of generating reactive oxygen intermediates (Heyworth & Segal, 1986). Since activated macrophages have a greatly increased capacity to generate reactive oxygen metabolites, it is important to investigate the possible relationship between the myristoylated 48K macrophage protein and the phosphorylated 47K neutrophil protein.

The LPS-primed macrophage has proved an excellent model system in which to study the convergence of three distinct signal transduction systems, namely those mediated by protein myristoylation, protein phosphorylation and calcium. The biological read-outs, such as 20:4 metabolism and the release of reactive oxygen intermediates, are clear and distinct, and since the cells are terminally differentiated the signals are not obscured by those involved in the regulation of the cell cycle. A number of proteins that are likely to have a role in signal transduction have been identified. They include the 68K PKC substrate and the 48K protein whose myristoylation is induced by IFNγ. Future studies will concentrate on the identity and function of these proteins.

I thank Dr Zanvil A. Cohn for support and guidance, Drs Paul Greengard and Antony Rosen for many stimulating discussions, Doug Marratta and Matthew Keum for excellent technical assistance, and Drs Kathy Barker and Ralph Steinman for critically reading the manuscript. The work on the 68K protein was done in collaboration with K. A. Albert, J. K.-T. Wang and Paul Greengard. This work was supported by National Institutes of Health Grant AI 25032 and AI 07012 and by a grant-in-aid from the Squibb Institute for Medical Research. A. A. Aderem is a Pew Scholar in the Biomedical Sciences.

Aderem
,
A. A.
,
Albert
,
K. A.
,
Keum
,
M. M.
,
Wang
,
J. K. T.
,
Greengard
,
P.
&
Cohn
,
Z. A.
(
1988a
).
Stimulus-dependent myristoylation of major substrate for protein kinase C
.
Nature, Land
.
332
,
362
364
.
Aderem
,
A. A.
,
Cohen
,
D. S.
,
Wright
,
S. D.
&
Cohn
,
Z. A.
(
1986a
).
Bacterial lipopolysaccharides prime macrophages for enhanced release of arachidonic acid metabolites
.
J. exp. Med
.
164
,
165
179
.
Aderem
,
A. A.
,
Keum
,
M. M.
,
Pure
,
E.
&
Cohn
,
Z. A.
(
1986b
).
Bacterial lipopolysaccharides, phorbol myristate acetate, and zymosan induce the myristoylation of specific macrophage proteins
.
Proc. natn. Acad. Sci. U.S.A
.
83
,
5817
5821
.
Aderem
,
A. A.
&
Cohn
,
Z. A.
(
1986
).
Bacterial lipopolysaccharides modify signal transduction in the arachidonic acid cascade in macrophages
.
In Biochemistry of macrophages, Ciba Foundation Symposium 118
(ed.
M.
O’Conner
), pp.
196
210
.
London
:
Pitman
.
Aderem
,
A. A.
&
Cohn
,
Z. A.
(
1988
).
Calcium ionophore synergizes with bacterial lipopolysaccharides in activating macrophage arachidonic acid metabolism
.
J. exp. Med
.
167
,
623
631
.
Aderem
,
A. A.
,
Marratγa
,
D. E.
&
Cohn
,
Z. A.
(
1988b
).
Interferon-gamma induces the myristoylation of a 48K protein in macrophages
.
Proc. natn. Acad. Sci. U.SA. (in press)
.
Aderem
,
A. A.
,
Scott
,
W. A.
&
Cohn
,
Z. A.
(
1986c
).
Evidence for sequential signals in the induction of the archidonic acid cascade in macrophages
.
J. exp. Med
.
163
,
139
154
.
Aitken
,
A.
,
Cohen
,
P.
,
Santikarn
,
S.
,
Williams
,
D. H.
,
Calder
,
A. G.
,
Smith
,
A.
&
Klee
,
C. B.
(
1982
).
Identification of the NH2-terminal blocking group of calcineurin B as myristic acid
.
FEBS. Lett
.
150
,
314
318
.
Albert
,
K. A.
,
Walaas
,
S. L
,
Wang
,
J. K.-T.
&
Greengard
,
P.
(
1986
).
Widespread occurrence of “87kDa”, a major specific substrate for protein kinase C
.
Proc. natn. Acad. Sci. U.SA
.
83
,
2822
2826
.
Albert
,
K. A.
,
Nairn
,
A. C.
&
Greengard
,
P.
(
1987
).
The 87K protein, a major specific substrate for protein kinase C: Purification from bovine brain and characterization
.
Proc. natn. Acad. Sci. U.SA
.
84
,
7046
7050
.
Blackshear
,
P. J.
,
Wen
,
L.
,
Glynn
,
B. P.
&
Witters
,
L. A.
(
1986
).
Protein kinase C-stimulated phosphorylation in vitro of aMr80 000 protein phosphorylated in response to phorbol esters and growth factors in intact fibroblasts. Distinction from protein kinase C and prominence in brain
.
J. BIol. Chem
.
261
,
1459
1469
.
Bonney
,
R. J.
,
Naruns
,
P.
,
Davies
,
P.
&
Humes
,
J. L.
(
1979
).
Antigen-antibody complexes stimulate the synthesis and release of prostaglandins by mouse peritoneal macrophages
.
Prostaglandins
18
,
605
614
.
Bonney
,
R. J.
,
Wightman
,
P. D.
,
Dahlgren
,
M. E.
,
Davies
,
P.
,
Kuehl
,
F. A.
&
Humes
,
J. L.
(
1978
).
Regulation of prostaglandin synthesis and of selective release of lysosomal hydrolases by mouse peritoneal macrophages
.
Biochem. J
.
176
,
433
444
.
Bonney
,
R. J.
,
Wightman
,
P. D.
,
Dahlgren
,
M. E.
,
Davies
,
P.
,
Kuehl
,
F. A.
Jr
&
Humes
,
J. L.
(
1980
).
Effect of RNA and protein synthesis inhibitors on the release of inflammatory mediators by macrophages responding to phorbol myristate acetate
.
Biochim. BIophys. Acta
633
,
410
421
.
Buss
,
J. E.
,
Kamps
,
M. P.
&
Sefton
,
B. M.
(
1984
).
Myristic acid is attached to the transforming protein of Rous sarcoma virus during or immediately after synthesis and is present in both soluble and membrane-bound forms of the protein
.
Molec. cell. Biol
.
4
,
2697
2704
.
Buss
,
J. E.
,
Mumby
,
S. M.
,
Casey
,
P. J.
,
Gilman
,
A. G.
&
Sefton
,
B. M.
(
1987
).
Myristylated alpha subunits of guanine nucleotide-binding regulatory proteins
.
Proc. natn. Acad. Sci. U.SA
.
84
,
7493
7497
.
Buss
,
J. E.
&
Sefton
,
B. M.
(
1985
).
Myristic acid, a rare fatty acid, is the lipid attached to the transforming protein of Rous sarcoma virus and its cellular homolog
.
J. Virol
.
53
,
7
12
.
Carr
,
S. A.
,
Bieman
,
K.
,
Shoji
,
S.
,
Parmelee
,
D. C.
&
Titani
,
K.
(
1982
).
n-Tetradecanoyl is the NH2-terminal blocking group of the catalytic subunit of cyclic AMP-dependent protein kinase from bovine cardiac muscle
.
Proc. natn. Acad. Sci. U.SA
.
79
,
6128
6131
.
Cross
,
F. R.
,
Garber
,
E. A.
,
Pellman
,
D.
&
Hanafusa
,
H.
(
1984
).
A short sequence in the p6Osrc N terminus is required for p6Osrc myristylation and membrane association and for cell transformation
.
Molec. cell. Biol
.
4
,
1834
1842
.
Davies
,
P.
,
Bailey
,
P. J.
,
Goldenberg
,
M. M.
&
Ford-Hutchinson
,
A. W.
(
1984
).
The role of arachidonic acid oxygenation products in pain and inflammation. A
.
Rev. Immun
.
2
,
335
357
.
Di Virgilio
,
F.
,
Lew
,
D. P.
&
Pozzan
,
T.
(
1984
).
Protein kinase C activation of physiological processes in human neutrophils at vanishingly small cytosolic Caz+ levels
.
Nature, Lond
.
310
,
691
693
.
Guthrie
,
L. A.
,
McPhail
,
L. C.
,
Henson
,
P. M.
&
Johnston
,
R. B.
Jr
(
1984
).
Priming of neutrophils for enhanced release of oxygen metabolites by bacterial lipopolysaccharide. Evidence for increased activity of the superoxide-producing enzyme
.
J. exp. Med
.
160
,
1656
1671
.
Hamaguchi
,
M.
&
Hanafusa
,
H.
(
1987
).
Association of p6Osrc with Triton XI00-resistant cellular structure correlates with morphological transformation
.
Proc. natn. Acad. Sci. U.SA
.
84
,
2312
2316
.
Hamilton
,
T. A.
,
Jansen
,
M. M.
,
Somers
,
S. D.
&
Adams
,
D. O.
(
1986
).
Effects of bacterial lipopolysaccharide on protein synthesis in murine peritoneal macrophages: relationship to activation for macrophage tumoricidal function
.
J. cell Physiol
.
128
,
9
17
.
Heyworth
,
P. G.
&
Segal
,
A. W.
(
1986
).
Further evidence for the involvement of a phosphoprotein in the respiratory burst oxidase of human neutrophils
.
Biochem. J
.
239
,
723
731
.
Homma
,
J. Y.
,
Kanegasaki
,
S.
,
Luderitz
,
O.
,
Shiba
,
T.
&
Westphal
,
O.
(
1984
).
Bacterial endotoxin: Chemical, biological and clinical aspects
.
Weinheim
:
Verlag Chemie
.
Humes
,
J. L.
,
Sadowski
,
S.
,
Galavage
,
M.
,
Goldenberg
,
M.
,
subers
,
E.
,
Bonney
,
R. J.
&
Kuehl
,
F. A.
Jr
(
1982
).
Evidence for two sources of arachidonic acid for oxidative metabolism by mouse peritoneal macrophages
.
J. biol. Chem
.
257
,
1591
1594
.
Kamps
,
M. P.
,
Buss
,
J. E.
&
Sefton
,
B. M.
(
1985
).
Mutation of NH2-terminal glycine of p6Osrc prevents both myristoylation and morphological transformation
.
Proc. natn. Acad. Sci. U.SA
.
82
,
4625
4628
.
Marchildon
,
G. A.
,
Casnellie
,
J. E.
,
Walsh
,
K. A.
&
Krebs
,
E. G.
(
1984
).
Covalently bound myristate in a lymphoma tyrosine protein kinase
.
Proc. natn. Acad. Sci. U.SA
.
81
,
7679
7682
.
Mazzei
,
G. J.
,
Katoh
,
N.
&
Kuo
,
J. F.
(
1982
).
Polymyxin B is a more selective inhibitor for phospholipid-sensitive Caz+-dependent protein kinase than for calmodulin-sensitive Caz+-dependent protein kinase
.
Biochem. BIophys. Res. Commun
.
109
,
1129
1133
.
Morrison
,
D. C.
&
Ulevitch
,
R. J.
(
1978
).
The effects of bacterial endotoxins on host mediation systems
.
Am. J. Path
.
93
,
527
617
.
Nathan
,
C. F.
,
Murray
,
H. W.
,
Wiebe
,
M. E.
&
Rubin
,
B. Y.
(
1983
).
Identification of interferon-gamma as the lymphokine that activates human macrophage oxidative metabolism and antimicrobial activity
.
J exp. Med
.
158
,
670
689
.
Neidel
,
J. E.
&
Blackshear
,
P. J.
(
1986
).
Protein kinase C
.
In Phosphoinositides and receptor mechanisms
, vol.
7
(ed.
J.
Putney
), pp.
47
88
.
New York
:
Alan R. Liss
.
Nishizuka
,
Y.
(
1986
).
Studies arid perspectives of protein kinase C
.
Science
233
,
305
312
.
Ogmundsdotπr
,
H. M.
&
Weir
,
D. M.
(
1979
).
Stimulation of phosphatidylinositol turnover in the macrophage plasma membrane: a possible mechanism for signal transmission
.
Immunology
37
,
689
696
.
Olson
,
E. N.
,
Towler
,
D. A.
&
Glaser
,
L.
(
1985
).
Specificity of fatty acid acylation of cellular proteins
.
J. biol. Chem
.
260
,
3784
3790
.
Omary
,
M. B.
&
Trowbridge
,
I. S.
(
1981
).
Biosynthesis of the human transferrin receptor in cultured cells
.
J. biol. Chem
.
256
,
12 888
12 892
.
Ozols
,
J.
,
Carr
,
S. A.
&
Strittmatter
,
P.
(
1984
).
Identification of the NH2-terminal blocking group of NADH-cytochrome b5 reductase as myristic acid and the complete amino acid sequence of the membrane-binding domain
.
J. biol. Chem
.
259
,
13 349
13 354
.
Pabst
,
M. J.
&
Johnston
,
R. B.
(
1980
).
Increased production of superoxide anion by macrophages exposed in vitro to muramyl dipeptide or lipopolysaccharide
.
J. exp. Med
.
151
,
101
114
.
Paul
,
A. V.
,
Schultz
,
A.
,
Pincus
,
S. E.
,
Oroszlan
,
S.
&
Wimmer
,
E.
(
1987
).
Capsid protein VP4 of poliovirus is N-myristoylated
.
Proc. natn. Acad. Sci. U.SA
.
84
,
7827
7831
.
Pawlowski
,
N. A.
,
Kaplan
,
G.
,
Hamill
,
A. L.
,
Cohn
,
Z. A.
&
Scott
,
W. A.
(
1983
).
Arachidonic acid metabolism by human monocytes. Studies with platelet-depleted cultures
.
J. exp. Med
.
158
,
393
412
.
Pellman
,
D.
,
Garber
,
E. A.
,
Cross
,
F. R.
&
Hanafusa
,
H.
(
1985
).
An N-terminal peptide from p6Osrc can direct myristylation and plasma membrane localization when fused to heterologous proteins
.
Nature, Lond
.
314
,
374
377
.
Rein
,
A.
,
McClure
,
M. R.
,
Rice
,
N. R.
,
Luftig
,
R. B.
&
Schultz
,
A. M.
(
1986
).
Myristylation site in Pr65gag is essential for virus particle formation by Moloney murine leukemia virus
.
Proc, natn. Acad. Sci. U.SA
.
83
,
7246
7250
.
Rouzer
,
C. A.
&
Samuelsson
,
B.
(
1987
).
Reversible, calcium-dependent membrane association of human leukocyte 5-lipoxygenase
.
Proc. natn. Acad. Sci. U.SA
.
84
,
7393
7397
.
Rouzer
,
C. A.
,
Scott
,
W. A.
,
Cohn
,
Z. A.
,
Blackburn
,
P.
&
Manning
,
J. M.
(
1980
).
Mouse peritoneal macrophages release leukotriene C in response to a phagocytic stimulus
.
Proc. natn. Acad. Sci. U.SA
.
77
,
4928
4932
.
Rouzer
,
C. A.
,
Scott
,
W. A.
,
Hamill
,
A. L.
&
Cohn
,
Z. A.
(
1982
).
Synthesis of leukotriene C and other arachidonic acid metabolites by mouse pulmonary macrophages
.
J. exp. Med
.
155
,
720
733
.
Rozengurt
,
E.
,
Rodriguez-Pena
,
M.
&
Smith
,
K. A.
(
1983
).
Phorbol esters, phospholipase C, and growth factors rapidly stimulate the phosphorylation of a Mr 80000 protein in intact quiescent 3T3 cells
.
Proc. natn. Acad. Sci. U.SA
.
80
,
7244
7248
.
Schreiber
,
R. D.
,
Pace
,
J. L.
,
Russell
,
S. W.
,
Altman
,
A.
&
Katz
,
D. H.
(
1983
).
Macrophageactivating factor produced by a T cell hybridoma: physiochemical and biosynthetic resemblance to gamma-interferon
.
J. Immun
.
131
,
826
832
.
Schultz
,
A. M.
,
Henderson
,
L. E.
,
Oroszlan
,
S.
,
Garber
,
E. A.
&
Hanafusa
,
H.
(
1985
).
Amino terminal myristylation of the protein kinase p6Osrc, a retroviral transforming protein
.
Science
227
,
427
429
.
Scott
,
W. A.
,
Pawlowski
,
N. A.
,
Murray
,
H. W.
,
Andreach
,
M.
,
Zrike
,
J.
&
Cohn
,
Z. A.
(
1982
).
Regulation of arachidonic acid metabolism by macrophage activation
J. exp. Med
.
155
,
1148
1160
.
Scott
,
W. A.
,
Zrike
,
J. M.
,
Hamill
,
A. L.
,
Kempe
,
J.
&
Cohn
,
Z. A.
(
1980
).
Regulation of arachidonic acid metabolites in macrophages
.
J. exp. Med
.
152
,
324
333
.
Sefton
,
B. M.
&
Buss
,
J. E.
(
1987
).
The covalent modification of eukaryotic proteins with lipid
.
J. Cell Biol
.
104
,
1449
1453
.
Towler
,
D. A.
,
Adams
,
S. P.
,
Eubanks
,
S. R.
,
Towery
,
D. S.
,
Jackson-Machelski
,
E.
,
Glaser
,
L.
&
Gordon
,
J. I.
(
1987b
).
Purification and characterization of yeast myristoyl CoA: protein 7V-myristoyl transferase
.
Proc. natn. Acad. Sci. U.SA
.
84
,
2708
2712
.
Towler
,
D. A.
,
Eubanks
,
S. R.
,
Towery
,
D. S.
,
Adams
,
S. P.
&
Glaser
,
L.
(
1978a
)
Aminoterminal processing of proteins by N-myristoylation: substrate specificity of N-myristoyl transferase
.
J. BIol. Chem
.
262
,
1030
1036
.
Towler
,
D.
&
Glaser
,
L.
(
1986
).
Protein fatty acid acylation: enzymatic synthesis of an N- myristoylglycyl peptide
.
Proc. natn. Acad. Sci. U.SA
.
83
,
2812
2816
.
Tripp
,
C. S.
,
Mahoney
,
M.
&
Needleman
,
P.
(
1985
).
Calcium ionophore enables soluble agonists to stimulate macrophage 5-lipoxygenase
.
J. BIol. Chem
.
260
,
5895
5898
.
Weiel
,
J. E.
,
Hamilton
,
T. A.
&
Adams
,
D. O.
(
1986
).
LPS induces altered phosphate labeling of proteins in murine peritoneal macrophages
.
J. Immun
.
136
,
3012
3018
.
Wightman
,
P. D.
,
Dahlgren
,
M. E.
,
Hall
,
J. C.
,
Davies
,
P.
&
Bonney
,
R. J.
(
1981b
).
Identification and characterization of a phospholipase C activity in resident mouse peritoneal macrophages. Inhibition of the enzyme by phenothiazines
.
Biochem. J
.
197
,
523
526
.
Wightman
,
P. D.
,
Humes
,
J. L.
,
Davies
,
P.
&
Bonney
,
R. J.
(
1981a
).
Identification and characterization of two phospholipase A2 activities in resident mouse peritoneal macrophages
.
Biochem. J
.
195
427
433
.
Wightman
,
P. D.
&
Raetz
,
C. R.
(
1984
).
The activation of protein kinase by biologically active lipid moieties of lipopolysaccharide
.
J. BIol. Chem
.
259
,
10048
10052
.
Wilcox
,
C.
,
Hu
,
J. S.
&
Olson
,
E. N.
(
1987
).
Acylation of proteins with myristic acid occurs cotranslationally
.
Science
238
,
1275
1278
.
Woodgett
,
J. R.
,
Hunter
,
T.
&
Gould
,
K. L.
(
1987
).
In Cell membranes: methods and reviews
, vol.
3
(ed.
E. L.
Elson
,
W. A.
Frazier
&
L.
Glazer
), pp.
215
340
.
New York
:
Plenum
.
Wu
,
W. S.
,
Walaas
,
S. I.
,
Nairn
,
A. C.
&
Greengard
,
P.
(
1982
).
Calcium/phospholipid regulates phosphorylation of a Mr “87k” substrate protein in brain synaptosomes
.
Proc. natn. Acad. Sci. U.SA
.
79
,
5249
5253
.
Young
,
J. D-E.
,
Ko
,
S. S.
&
Cohn
,
Z. A.
(
1984
).
The increase in intracellular free calcium associated with IgG2b/1 Fc receptor-ligand interactions: role in phagocytosis
.
Proc. natn. Acad. Aci. U.SA
.
81
,
5430
5434
.