Dissociation of the chemotactic and mitogenic activities of platelet- derived growth factor by human neutrophil elastase

Because platelet-derived growth factor (PDGF) may be released at sites where neutrophil proteinases may also be released, we examined the effects of neutrophil elastase and cathepsin G upon the chemotactic and mitogenic activities of PDGF. Elastase abolished the chemotactic activity of PDGF for fibroblasts but had no effect on its chemotactic activity for monocytes, or on its mitogenic activity for 3T3 cells or its capacity to bind to 3T3 cells. Cathepsin G had no effect upon the chemotactic or mitogenic activities of PDGF. In contrast, trypsin eliminated the chemotactic activity of PDGF for monocytes and fibroblasts and the mitogenic activity of PDGF. After reduction and alkylation, PDGF retained full chemotactic activity for fibroblasts and monocytes but exhibited no mitogenic activity and only limited binding to 3T3 cells. These results indicate separate domains on PDGF for fibroblast chemotactic and mitogenic activity and for monocyte and fibroblast chemotactic activity and raise the possibility that the biological activities of PDGF may be modified selectively in vivo. The findings further suggest that the majority of PDGF receptors on fibroblasts mediate mitogenic activity and that only a minority of the PDGF receptors on fibroblasts are responsible for chemotactic activity.     

Differential regulation of responsiveness to fMLP and C5a upon dendritic cell maturation: correlation with receptor expression

The trafficking of immature and mature dendritic cells (DCs) to different anatomical sites in vivo is critical for fulfilling their roles in the induction of Ag-specific immune responses. Although this process is complex and regulated by many mediators, the capacity of DCs to migrate is predominantly dependent on the expression of particular chemotactic receptors on the surface of DCs that enable them to move along chemotactic gradients formed by the corresponding chemokines and/or classical chemoattractants. Here we show that immature DCs (iDCs) respond to both fMLP and C5a as determined by chemotaxis and Ca2+ mobilization, whereas mature DCs (mDCs) respond to C5a, but not fMLP. Additionally, iDCs express the receptors for both fMLP and C5a at mRNA and protein levels. Upon maturation of DCs, fMLP receptor expression is almost completely absent, whereas C5a receptor mRNA and protein expression is maintained. Concomitantly, mDCs migrate chemotactically and mobilize intracellular Ca2+ in response to C5a, but not fMLP. Thus the interaction between C5a and its receptor is likely involved in the regulation of trafficking of both iDCs and mDCs, whereas fMLP mobilizes only iDCs. The differential responsiveness to fMLP and C5a of iDCs and mDCs suggests that they play different roles in the initiation of immune responses.     

Differential migratory response of U-2 OS osteosarcoma cell to the various forms of platelet-derived growth factor.

U-2 OS osteosarcoma cells are mesenchymal-derived transformed cells spontaneously expressing both platelet-derived growth factor (PDGF) A- and B-chain genes, and releasing PDGF AA dimers in culture. Using modified Boyden chemotactic chambers, platelet-purified PDGF was shown to be a chemoattractant for U-2 OS cells. More specifically, U-2 OS cells migrated in the presence of PDGF AB and BB dimers but not in the presence of PDGF AA dimers. This pattern of response was similar to that observed with human fibroblasts and this similarity is consistent with the fact that U-2 OS cells express PDGF receptor alpha- and beta-subunits in a similar fashion to human fibroblasts.     

Identification of neutrophil granule protein cathepsin G as a novel chemotactic agonist for the G protein-coupled formyl peptide receptor

The antimicrobial and proinflammatory neutrophil granule protein cathepsin G (CaG) has been reported as a chemoattractant for human phagocytic leukocytes by using a putative G protein coupled receptor. In an effort to identify potential CaG receptor(s), we found that CaG-induced phagocyte migration was specifically attenuated by the bacterial chemotactic peptide fMLP, suggesting these two chemoattractants might share a receptor. In fact, CaG chemoattracts rat basophilic leukemia cells (RBL cells) expressing the high affinity human fMLP receptor FPR, but not parental RBL cells or cells transfected with other chemoattractant receptors. In addition, a specific FPR Ab and a defined FPR antagonist, cyclosporin H, abolished the chemotactic response of phagocytes and FPR-transfected cells to CaG. Furthermore, CaG down-regulated the cell surface expression of FPR in association with receptor internalization. Unlike fMLP, CaG did not induce potent Ca(2+) flux and was a relatively weaker activator of MAPKs through FPR. Yet CaG activated an atypical protein kinase C isozyme, protein kinase Czeta, which was essential for FPR to mediate the chemotactic activity of CaG. Thus, our studies identify CaG as a novel, host-derived chemotactic agonist for FPR and expand the functional scope of this receptor in inflammatory and immune responses.     

Chemokine production by G protein-coupled receptor activation in a human mast cell line: roles of extracellular signal-regulated kinase and NFAT

Chemoattractants are thought to be the first mediators generated at sites of bacterial infection. We hypothesized that signaling through G protein-coupled chemoattractant receptors may stimulate cytokine production. To test this hypothesis, a human mast cell line (HMC-1) that normally expresses receptors for complement components C3a and C5a at low levels was stably transfected to express physiologic levels of fMLP receptors. We found that fMLP, but not C3a or C5a, induced macrophage inflammatory protein (MIP)-1ss (CCL4) and monocyte chemoattractant protein-1 (CCL2) mRNA and protein. Although fMLP stimulated both sustained Ca(2+) mobilization and phosphorylation of extracellular signal-regulated kinase (ERK), these responses to C3a or C5a were transient. However, transient expression of C3a receptors in HMC-1 cells rendered the cells responsive to C3a for sustained Ca(2+) mobilization and MIP-1ss production. The fMLP-induced chemokine production was blocked by pertussis toxin, PD98059, and cyclosporin A, which respectively inhibit G(i)alpha activation, mitgen-activated protein kinase kinase-mediated ERK phosphorylation, and calcineurin-mediated activation of NFAT. Furthermore, fMLP, but not C5a, stimulated NFAT activation in HMC-1 cells. These data indicate that chemoattractant receptors induce chemokine production in HMC-1 cells with a selectivity that depends on the level of receptor expression, the length of their signaling time, and the synergistic interaction of multiple signaling pathways, including extracellular signal-regulated kinase phosphorylation, sustained Ca(2+) mobilization and NFAT activation.     

Phospholipase A2 activation in human neutrophils. Differential actions of diacylglycerols and alkylacylglycerols in priming cells for stimulation by N-formyl-Met-Leu-Phe

Both 1,2-diacyl- and 1-O-alkyl-2-acylglycerols are formed during stimulation of human neutrophils (PMN), and both can prime respiratory burst responses for stimulation by the chemotactic peptide, N-formyl-Met-Leu-Phe (fMLP); however, mechanisms of priming are unknown. Arachidonic acid (AA) release through phospholipase A2 activation and metabolism by 5-lipoxygenase are important activities of PMN during inflammation and could be involved in the process of primed stimulation. Therefore, we have examined the ability of diacyl- and alkylacylglycerols to act as priming agents for AA release and metabolism in human neutrophils. After prelabeling PMN phospholipids with [3H]AA, priming was tested by incubating human PMN with the diacylglycerol, 1-oleoyl-2-acetylglycerol (OAG), or its alkylacyl analog, 1-O-delta 9-octadecenyl-2-acetylglycerol (EAG) before stimulating with fMLP. fMLP (1 microM), OAG (20 microM), or EAG (20 microM) individually caused little or no release of labeled AA. However, after priming PMN with the same concentrations of either OAG or EAG, stimulation with 1 microM fMLP caused rapid (peak after 1 min) release of 6-8% of [3H]AA from cellular phospholipids; total release was similar with either diglyceride. Priming cells with OAG also enhanced conversion of released AA to leukotriene B4 (LTB4) and 5-hydroxyeicosatetraenoic acid (5-HETE) upon subsequent fMLP stimulation, but AA metabolites were not increased in EAG-primed PMN. If fMLP was replaced with the calcium ionophore A23187 (which directly causes release of AA and production of LTB4 and 5-HETE), priming by both diglycerides again enhanced release of [3H]AA, but only OAG priming increased lipoxygenase activity. Indeed, EAG pretreatment markedly reduced LTB4 and 5-HETE production. Thus, both diglycerides prime release of AA from membrane phospholipids but have opposite actions on the subsequent metabolism of AA.     

Cryptic chemotactic activity of fibronectin for human monocytes resides in the 120-kDa fibroblastic cell-binding fragment

Monocytes and lymphocytes form a second wave of infiltrating blood leukocytes in areas of tissue injury. The mechanisms for monocyte accumulation at these sites are not completely understood. Recently, however, fragments from extracellular matrix proteins including collagen, elastin, and fibronectin have been shown to induce monocyte chemotaxis. In this report we demonstrate that chemotactic activity for human monocytes is expressed when a 120-kDa fragment containing the RGDS cell-binding peptide is released from intact fibronectin or from larger fibronectin fragments. Monocytes, either from mononuclear cell Ficoll-Hypaque preparations (10-20% monocytes, 89-90% lymphocytes) or from elutriation preparations (95% monocytes, 5% lymphocytes), but not lymphocytes, migrated toward 120-kDa fragment preparations (10(-7) M) in blind-end chambers when the cells were separated from the chemoattractant by a 5-micron pore polycarbonate filter either alone or overlying a 0.45-micron pore nitrocellulose filter. Neutrophils migrated toward zymosan-activated serum but not toward 10(-5)-10(-8) M concentrations of the 120-kDa fragment. Intact fibronectin had no chemotactic activity for human monocytes. Fibronectin was isolated from citrated human plasma by sequential gelatin-Sepharose affinity and DEAE ion-exchange chromatography in the presence of buffers containing 1 mM phenylmethylsulfonyl fluoride to prevent fragmentation. Controlled enzymatic digestion with thermolysin cleaved fibronectin into 30 kDa fibrin, 45 kDa collagen, and 150/160-kDa cell and heparin domains. Upon prolonged digestion, purified 150/160-kDa fragments were cleaved into 120-kDa cell and 30/40-kDa heparin-binding fragments. Even though the intact fibronectin molecule, the 150/160-kDa fragments, and the 120-kDa fragment, have cell binding activity for Chinese hamster ovary fibroblasts, only the 120-kDa fragment expressed chemotactic activity for human monocytes. Thus, the 120-kDa fibroblastic cell-binding fragment contains a cryptic site for monocyte chemotaxis which is expressed upon enzymatic cleavage of fibronectin.     

Thioredoxin specifically cross-desensitizes monocytes to MCP-1

Thioredoxin (Trx) is a protein disulfide oxidoreductase which can be secreted and acts as a cytokine. As we recently reported that Trx is chemotactic, we investigated whether it desensitizes monocytes or PMN to other chemokines. Preincubation for 15 min with Trx inhibited the chemotactic response of monocytes to MCP-1, but not to fMLP. This effect was independent of whether Trx was present during the chemotaxis assay or only during the preincubation. Preincubation (5 min) with Trx also inhibited the increase in intracellular Ca(2+) induced by MCP-1 in monocytes, but not that induced by fMLP. Preincubation with Trx did not affect the chemotactic response induced in PMN by IL-8. The inhibition of chemotactic and Ca(2+) responses to MCP-1 in monocytes was not due to a down-regulation of the MCP-1 receptor, as shown by receptor binding studies. The Ca(2+) response to MCP-1 was also inhibited by Trx in a CCR2-transfected cell line. It is suggested that Trx inhibits monocyte responses to chemokines by acting downstream of the chemokine receptors. Since there are high concentrations of circulating Trx in infection and inflammatory diseases, this might act as an inhibitor of monocyte migration in vivo.     

Time dependence of transmembrane potential changes and intracellular calcium flux in stimulated human monocytes

An important characteristic of the functional differentiation of the blood monocyte is the development of its capacity to recognize and respond to stimuli. This ability is mediated to a large extent by specific receptor glycoproteins located on the cell surface. Stimulation of mononuclear phagocytes via these receptors results in a rapid rise in intracellular Ca++ concentration, accompanied or followed by a change in membrane potential, generation of oxidative products, degranulation, and effector functions such as phagocytosis, aggregation, or locomotion. While the development of these characteristics is difficult to characterize in vivo, several investigators have demonstrated in vitro changes in these cells that correlate with the development of effector function. To examine the mechanisms of specific membrane-stimulus interactions of monocytes as they differentiate into macrophage-like cells, we studied the responses of human monocytes and of monocytes incubated in serum-containing medium for up to 96 hr to the chemotactic peptide formyl-methionyl-leucyl-phenylalanine (fMLP). Freshly isolated monocytes exhibited little change in transmembrane potential following stimulation with an optimal concentration of peptide and underwent a significant increase only after 48 hr in culture. While constant resting intracellular Ca++ concentrations were maintained during the culture period, intracellular Ca++ levels following fMLP stimulation increased with with incubation in serum, for up to 96 hr. In contrast, fMLP-induced respiratory burst activity increased from 0 to 24 hr in culture; it remained elevated at 48 hr but declined again by 96 hr. Incubation of the cells for 24 hr increased their random (unstimulated) motility in modified Boyden chambers but did not alter the cells' directed (chemotactic) response to fMLP in comparison to the response of freshly isolated monocytes. Peptide binding to the cells did not increase during the incubation period, indicating that an increase in receptor number or in affinity for fMLP was not responsible for the enhanced responsiveness to fMLP as incubation time increased. These studies indicate that incubation of monocytes in serum-containing medium leads to a complex, altered series of responses to fMLP that correlate with the differentiation of the original monocytes in vitro and may relate to the in vivo differentiation of monocytes to macrophages.     

Alteration of the chemotactic response of NIH/3T3 cells to PDGF by growth factors, transformation, and tumor promoters

The platelet-derived growth factor (PDGF) is a potent chemoattractant for cells that respond to PDGF as a mitogen. The chemotactic response of these cells to PDGF is inversely related to their rate of proliferation, with quiescent cells exhibiting a 25-fold greater chemotactic response than exponentially growing cells. Factors that stimulate the growth of quiescent cells (EGF, FGF, PDGF, and serum) decrease the cells' migratory response to PDGF but not to fibronectin, suggesting that the decreased migration is not due to a general paralysis of cell motility. Transformed lines of NIH/3T3 cells lose their ability to respond to PDGF as a chemoattractant but can still migrate in response to fibronectin. Similarly, after treatment of 3T3 cells with the tumor-promoter phorbol myristate acetate, which induces a transformation-like phenotype, the cells no longer respond to PDGF as a chemoattractant but retain their migratory response to fibronectin. Thus it appears that the growth state of the cells can alter their migratory response to PDGF. These data suggest that growth factors, transformation, and tumor promoters specifically alter the cells' ability to respond to the PDGF-mediated chemotactic signal. It appears that both transformation and tumor promoters accomplish this by altering PDGF-binding to the cell surface.     

A guanine nucleotide regulatory protein controls polyphosphoinositide metabolism, Ca2+ mobilization, and cellular responses to chemoattractants in human monocytes

Previous studies demonstrated that oligopeptide chemoattractant receptors on PMN and macrophages exist in high and low affinity states which are interconvertible by guanosine di- and triphosphates. These observations suggest that guanine nucleotide regulatory (N) proteins play a role in phagocyte activation by chemotactic factors. The data presented here indicate that chemotactic factor receptors on monocytes utilize an N protein to activate phospholipase C and subsequent biologic responses by the cells. This conclusion is based on the findings that inactivation of an N protein of 41,000 m.w. by Bordetella pertussis toxin (PT) treatment abolishes monocyte responsiveness to chemoattractants but not to lectins, PMA, or the Ca2+ ionophore A23187. Treatment with PT inhibited IP3 production, Ca2+ mobilization, and cellular activation as assessed by chemotaxis and changes in forward light scattering in response to the chemoattractants by at least 80%. Therefore, a PT-sensitive N protein plays an important role in the activation of monocytes by chemoattractants.     

Platelet-derived growth factor in chemotactic for fibroblasts

Chemotaxis assays in modified Boyden chambers were used to detect fibroblast chemoattractants in materials released from early-stage inflammatory cells, namely, mast cells, platelets, and neutrophils. Strong attractant activity was found in substances released from platelets. This activity was accounted for mainly by the platelet- derived growth factor (PDGF), which is released from the platelets and which was active as a chemoattractant at 0.5-1.0 mitogenic units/ml. The mitogenic activity of purified PDGF, measured by [3H]thymidine incorporation, occurs at a similar concentration range. By varying the gradient of PDGF, we demonstrated that PDGF stimulates chemotaxis rather than random motility. Preincubation of suspensions of fibroblasts in the presence of PDGF decreased the subsequent migration of cells to a gradient of PDGF as well as to a gradient of fibronectin, which is also in attractant for fibroblasts. The chemotactic response of fibroblasts to PDGF was not inhibited by hydroxyurea or azidocytidine but was inhibited by actinomycin D and cycloheximide, suggesting that synthesis of RNA and proteins but not of DNA is required for the chemotactic response to occur. Fibroblast growth factor, epidermal growth factor, nerve growth factor, and insulin were not chemotactic for human skin fibroblasts, suggesting that the chemoattractant activity of PDGF for fibroblasts is not a general property of growth factors and mitogens. These results suggest that PDGF could have two functions in wound healing: to attract fibroblasts to migrate into the clot and then to induce their proliferation.     

Neutrophils and B cells express XCR1 receptor and chemotactically respond to lymphotactin

The C chemokine lymphotactin (Lptn) has been reported to act specifically on CD4(+) and CD8(+) T lymphocytes and natural killer (NK) cells, but not monocytes. However, the chemotactic effect of Lptn on other types of hematopoietic cells has not been well studied. In this study we investigated (i) the chemotactic influences of Lptn on T and B lymphocytes, neutrophils, monocytes, and dendritic cells, and (ii) the expression of the Lptn receptor (XCR1) by these cells, using RT-PCR. Our data showed that Lptn is chemotactic for B lymphocytes and neutrophils as well as T lymphocytes, but not for monocytes or dendritic cells, and that XCR1 expression is found only in association with T and B lymphocytes and neutrophils, but not monocytes or dendritic cells. Thus, this study is the first demonstration of a chemotactic effect of Lptn on neutrophils and confirms the association of this effect with expression of the XCR1 receptor on these cells. These data suggest that Lptn could potentially be an important protein in the regulation of T and B lymphocytes and neutrophil trafficking, and thereby also their roles in inflammatory and immunological responses.     

Neutrophils Stimulated with a Variety of Chemoattractants Exhibit Rapid Activation of p21-Activated Kinases (Paks): Separate Signals Are Required for Activation and Inactivation of Paks

Activation of the p21-activated protein kinases (Paks) was compared in neutrophils stimulated with a wide variety of agonists that bind to receptors coupled to heterotrimeric G proteins. Neutrophils stimulated with sulfatide, a ligand for the L-selectin receptor, or the chemoattractant fMet-Leu-Phe (fMLP), platelet-activating factor, leukotriene B₄, interleukin-8, or the chemokine RANTES exhibited a rapid and transient activation of the 63- and 69-kDa Paks. These kinases exhibited maximal activation with each of these agonists within 15 s followed by significant inactivation at 3 min. In contrast, neutrophils treated with the chemoattractant and anaphylatoxin C5a exhibited a prolonged activation (>15 min) of these Paks even though the receptor for this ligand may activate the same overall population of complex G proteins as the fMLP receptor. Addition of fMLP to neutrophils already stimulated with C5a resulted in the inactivation of the 63- and 69-kDa Paks. Optimal activation of Paks could be observed at concentrations of these agonists that elicited only shape changes and chemotaxis in neutrophils. While all of the agonists listed above triggered quantitatively similar activation of the 63- and 69-kDa Paks, fMLP was far superior to the other stimuli in triggering activation of the c-Jun N-terminal kinase (JNK) and the p38 mitogen-activated protein kinase (MAPK). These data indicate that separate signals are required for activation and inactivation of Paks and that, in contrast to other cell types, activated Pak does not trigger activation of JNK or p38-MAPK in neutrophils. These results are consistent with the recent hypothesis that G-protein-coupled receptors may initiate signals independent of those transmitted by the α and βγ subunits of complex G proteins.     

rC5a directs the in vitro migration of human memory and naive tonsillar B lymphocytes: implications for B cell trafficking in secondary lymphoid tissues.

Human C5a is a potent chemoattractant for granulocytes, monocytes, and dendritic cells. In mice C5a has been shown to be chemotactic for germinal center (GC) B cells. To date, no information is available on the effects of C5a on human B cell locomotion. Here we demonstrate that rC5a increases polarization and migration of human tonsillar B cells. The locomotory response was due to both chemokinetic and chemotactic activities of rC5a. Moreover, memory and, at a lesser extent, naive B cell fractions from purified tonsillar populations displayed rC5a-enhanced migratory properties, whereas GC cells did not. Flow cytometry revealed C5aR (CD88) on approximately 40% memory and 10% naive cells, respectively, whereas GC cells were negative. Immunohistochemistry showed that a few CD88+ cells were of the B cell lineage and localized in tonsillar subepithelial areas, where the majority of memory B cells settle. Pretreatment of memory B cells with the CD88 mAb abolished their migratory responsiveness to rC5a. Finally, the C5 gene was found to be expressed in naive, GC, and memory B lymphocytes at both the mRNA and the protein level. This study delineates a novel role for C5a as a regulator of the trafficking of human memory and naive B lymphocytes and supports the hypothesis that the B cells themselves may serve as source of C5 in secondary lymphoid tissues.     

Modulation of monocyte chemotactic function in inflammatory lesions. Role of inflammatory mediators

Monocyte recruitment and accumulation in the synovial tissue is pivotal in the evolution of rheumatoid arthritis (RA). In the present study we examined the chemotactic potential of monocytes obtained from synovial fluid (SF) of patients with RA. Functionally, SF monocytes exhibited greatly diminished chemotactic activity to C5a compared with monocytes from the peripheral blood. In contrast, their chemotactic responsiveness to the synthetic peptide, FMLP, was nearly normal. To define a mechanism for this differential chemotactic dysfunction, cell-surface receptors for C5a (C5aR) and FMLP (FMLP-R) were evaluated. Whereas FMLP-R expression was similar on both blood and inflammatory monocytes, C5aR expression was markedly reduced on SF cells. Because decreased C5a binding in certain RA SF samples could not be attributed to free C5a, known or suspected components of inflammatory SF were evaluated for their ability to modulate chemotactic ligand receptors. Bacterial products including LPS and streptococcal cell walls, which are potent monocyte activators, down-regulated C5aR without affecting FMLP-R. Moreover, the cytokines IFN-gamma and granulocyte-macrophage-CSF selectively decreased C5aR in parallel with decreased in vitro chemotactic activity to C5a. Thus, these data indicate that 1) synovial effusions may contain C5a and/or inflammatory mediators that modulate phenotypic and functional changes in monocytes, 2) chemotactic ligand receptors are independently regulated in inflammatory lesions, and 3) decreased C5aR expression and chemotactic potential likely provide a mechanism whereby monocyte-macrophages persist within the inflamed synovium.     

Chemotactic responses of human peripheral blood monocytes to the complement-derived peptides C5a and C5a des Arg

We examined responses of human peripheral blood polymorphonuclear leukocytes (PMN) and monocytes to the highly purified human complement-derived peptides C5a and C5a des Arg. As reported previously, C5a proved to be approximately 10- to 20-fold more potent than C5a des Arg as a chemoattractant for human PMN. C5a also was more potent than C5a des Arg in causing PMN to acquire a polarized morphology. In contrast, we found that human monocytes do not distinguish between C5a and C5a des Arg when these peptides are used as chemoattractants. In two different assay systems, both peptides acted at identical concentrations to stimulate suboptimal and optimal migration of monocytes. Human monocytes also did not distinguish between C5a and C5a des Arg when these peptides were used as inducers of polarization. Studies performed with functionally active, [125I]-labeled C5a and C5a des Arg, however, demonstrated that binding of C5a des Arg to monocytes differed from binding of C5a. Although [125I]-C5a des Arg appeared to bind to the same receptor as [125I]-C5a, binding of labeled C5a des Arg occurred with an affinity that was approximately 100-fold less than that observed with labeled C5a. These results indicate that leukocyte chemotactic and polarization responses to C5a and C5a des Arg vary, depending on the target cell type.     

Enhancement of endogenous production of 12-L-hydroxy-5,8,10,14-eicosatetraenoic acid in aortic smooth muscle cells by platelet-derived growth factor

Platelet-derived growth factor (PDGF) has a chemotactic effect on smooth muscle cells, which is inhibited by lipoxygenase inhibitor caffeic acid. In order to study the role of endogenous lipoxygenase products of arachidonic acid on the chemotactic action of PDGF, effects of PDGF on the lipoxygenase pathway in smooth muscle cells were examined. Lipoxygenase products were analyzed by high-performance liquid chromatography. 15-, 5- and 12-lipoxygenase activities, in order of magnitude, were found in smooth muscle cell homogenate. However, when the lipoxygenase products were analyzed using intact cells prelabelled with [14C]arachidonic acid, only 12-L-hydroxy-5,8,10,14-eicosatetraenoic acid (HETE) was found to be produced endogenously. In addition, 12-HETE was not released into the medium. Treatment of the cells with PDGF increased the endogenous production of 12-HETE. The amounts of intracellular 12-HETE in PDGF-treated cells were 126, 132 and 146% at 1, 3, and 10 hr's after the initiation of PDGF treatment, respectively, when control value at each time point was considered as 100%. Caffeic acid (10(-4) M) completely inhibited the PDGF effect on 12-HETE production. However, PDGF treatment did not significantly alter the 12-lipoxygenase activity. These results suggest that the stimulatory effect of PDGF on 12-HETE production was not mediated by the activation of 12-lipoxygenase activity. Since 12-HETE itself is a potent chemoattractant for smooth muscle cells, the present dat strongly suggest that 12-HETE could be an important intracellular mediator of the chemotactic action of PDGF on aortic smooth muscle cells.