Kinases in Neutrophils Activation of the P21-Activated Protein Products

Products of Sphingolipid Catabolism Block Activation of the p21-Activated Protein Kinases in Neutrophils

This information is current as Jian P. Lian, RiYun Huang, Dwight Robinson and John A. of September 23, 2021. Badwey J Immunol 1998; 161:4375-4381; ; http://www.jimmunol.org/content/161/8/4375 Downloaded from

References This article cites 57 articles, 35 of which you can access for free at: http://www.jimmunol.org/content/161/8/4375.full#ref-list-1

Why The JI? Submit online. http://www.jimmunol.org/

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on September 23, 2021

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 1998 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Products of Sphingolipid Catabolism Block Activation of the p21-Activated Protein Kinases in Neutrophils1

Jian P. Lian,* RiYun Huang,† Dwight Robinson,† and John A. Badwey2*‡

Neutrophils stimulated with the chemoatttractant FMLP are known to exhibit a rapid and transient activation of two p21- activated protein kinases (Paks) with molecular masses of approximately 63 and 69 kDa. Paks can be detected by their ability to undergo renaturation and catalyze the phosphorylation of a peptide substrate that corresponds to amino acid residues 297 to 331 of the 47-kDa subunit of the nicotinamide-adenine dinucleotide phosphate-oxidase complex (p47-phox) ﬁxed within a gel. In this study, we demonstrate that N-acetylsphingosine (C2-ceramide) and a variety of sphingoid bases (e.g., D-erythrosphingosine) block activation of the 63- and 69-kDa Paks in neutrophils. The concentrations of these lipids that were effective in blocking Pak activation were similar to those that inhibit a variety of neutrophil responses. Activation of the 63- and 69-kDa Paks was also markedly reduced in neutrophils treated with sphingomyelinase before stimulation. Moreover, we report that addition of C2- ceramide or D-erythrosphingosine to neutrophils after stimulation with FMLP markedly enhances the rate of Pak inactivation. Downloaded from These effects were not mimicked by arachidonate, which is a potent disorganizing agent of neutrophil membranes. These data support and extend the proposal that sphingoid bases may establish a set point in neutrophils for positive stimuli. The Journal of Immunology, 1998, 161: 4375–4381.

eutrophils stimulated with the chemoattractant FMLP ex- A variety of studies over the past decade strongly suggest that

hibit a rapid and transient activation of four renaturable certain sphingolipids may function as second messengers or mod- http://www.jimmunol.org/ N protein kinases with molecular masses of approximately ulators of cellular responses (e.g., for review, see Refs. 21–23). 69, 63, 49, and 40 kDa (1–3). Activation of all four of these kinases Cell-permeable derivatives of ceramide (C2-ceramide) and sphin- appears to be under the control of a single stimulatory pathway that is goid bases (e.g., D-erythrosphingosine) inhibit phagocytosis, de- Ϫ sensitive to antagonists of heterotrimeric G proteins, type 1 and/or 2A granulation, and O2 production by neutrophils (17, 24–28). Inter- protein phosphatases, tyrosine kinases, and phosphatidylinositol 3-ki- estingly, C2-ceramide also inhibits activation of ERK1 and ERK2 nase (1–6). The 63- and 69-kDa enzymes were subsequently identi- in neutrophils (17) and blocks translocation of certain small GT- fied as p21-activated kinases (Paks)3 (7, 8). Paks undergo autophos- Pases (Cdc42) to the membrane of these cells (29). Sphingoid phorylation/activation upon binding the active (GTP-bound) forms of bases inhibit PKC (25) and phosphatidic acid phosphatase in neu- the small GTPases (p21) Rac or Cdc42 (9). Paks can catalyze the trophils (30). Increased amounts of ceramide and sphingosine have by guest on September 23, 2021 phosphorylation of the 47-kDa subunit of the nicotinamide-adenine been observed in FMLP-stimulated neutrophils and neutrophils dinucleotide phosphate-oxidase complex (p47-phox) (7) and certain undergoing IgG-dependent phagocytosis (17, 28, 31). The cellular heavy and light chains of myosin (e.g., 10, 11). Paks are also involved concentrations of these lipids also undergo marked increases in in the activation/potentiation of several distinct MAP kinase cascades. certain genetic diseases (e.g., Farber’s lipogranulomatosis) and Transfection of constitutively active Pak or overexpression of wild- during various infections (e.g., Refs. 32–34). type Pak in certain cells is sufficient to activate the c-jun N-terminal Paks and Pak-related kinases may be regulated by various lipids kinase/stress-activated protein kinase (JNK/SAPK) and, to a lesser (35, 36). In particular, a recent study has demonstrated that high extent, p38 MAP kinase (e.g., 12–14; for review, see Ref. 15). More- concentrations of sphingoid bases (50–200 ␮M) and certain other over, activated Pak can potentiate the ability of wild-type Raf-1 or lipids can activate Pak 1 in vitro in a manner similar to that ob- growth factors to stimulate ERKs and MAP-kinase kinases in numer- served with Cdc42-GTP (36). COS-7 cells expressing Pak 1 ex- ous cell types (14, 16). Selective antagonists of MAP-kinase kinase hibit enhanced activity of this kinase when treated with agents that and p38 MAP kinase inhibit chemotaxis, phagocytosis, degranulation, elevate the content of sphingosine (36). T cells also exhibit an and superoxide production by neutrophils (17–20). Thus, Paks may enhanced activity of Pak when treated with C2-ceramide through participate in a variety of neutrophil responses. an unknown mechanism (37). In this study, we report that products of sphingolipid catabolism block activation of the 63- and 69-kDa Paks in FMLP-stimulated neutrophils. The effective concentrations †Arthritis Unit, Massachusetts General Hospital, Boston, MA 02114; ‡Department of of these lipids are in the range of those that can be achieved under Biological Chemistry and Molecular Pharmacology, Harvard Medical School, Bos- ton, MA 02115; and *Boston Biomedical Research Institute, Boston, MA 02114 various physiologic/pathologic situations. The significance of Received for publication March 24, 1998. Accepted for publication June 22, 1998. these results to neutrophil function and Pak activation under various circumstances is discussed. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 These studies were supported by National Institutes of Health Grants DK 50015, AI Materials and Methods 23323 (to J.A.B.), and AR 43518 (to D.R.). Materials 2 Address correspondence and reprint requests to Dr. John A. Badwey, Boston Bio- medical Research Institute, 20 Staniford St., Boston, MA 02114. D-Erythro-C2-ceramide (N-acetylsphingosine), D-erythrodihydro-C2-cer- 3 Abbreviations used in this paper: Pak, p21-activated protein kinase; ERK, extracel- amide (N-acetyldihydrosphingosine), D-erythrosphingosine, D-erythrodihy- lular-regulated kinase; MAP, mitogen-activated protein; p47-phox, the 47-kDa pro- drosphingosine, and sphingosine-1-phosphate were purchased from Cal- tein component of the phagocyte oxidase; PKC, protein kinase C. biochem (La Jolla, CA). D,L-threodihydrosphingosine was obtained from

FIGURE 1. Effects of C2-ceramides on the activation of the 63- and 69-kDa Paks in neutrophils. The renaturable protein kinases in neutrophils FIGURE 2. Effects of various sphingoid bases on the activation of the 63- and 69-kDa Paks in neutrophils. Autoradiographs demonstrate the ef-

were assayed directly in gels containing the fixed peptide substrate, as Downloaded from referenced under Materials and Methods. A, Cells were incubated in the fects of D-erythrosphingosine (A), D,L-threodihydrosphingosine (B), and ␮ D-erythrodihydrosphingosine (C) on the activation of the 63- and 69-kDa absence (lanes a and b) and presence of 40 MC2-ceramide (lanes c and ␮ Paks in neutrophils. Intact cells were treated with: lane a, 0.25% (v/v) d)or40 MC2-dihydroceramide (lanes e and f) for 5 min prior to treatment for 15 s with 0.25% (v/v) DMSO (lanes a, c, and e; unstimulated DMSO for 5 min, followed by an additional 0.25% (v/v) DMSO for 15 s cells) or 1 ␮M FMLP (lanes b, d, and f; stimulated cells). B, Shows the (unstimulated cells); lane b, 0.25% (v/v) DMSO for 5 min, followed by 1 complete profile for the renaturable kinases from the same experiment, ␮M FMLP for 15 s; and lane c,10␮M sphingoid base for 5 min, followed ␮ except the time of development of the autoradiograph was increased from by 1 M FMLP for 15 s. The concentrations of sphingoid bases utilized in http://www.jimmunol.org/ these studies did not affect cell viability, as measured by the exclusion of 18 to 72 h. The concentrations of C2-ceramide and C2-dihydroceramide utilized in these studies did not affect cell viability, as measured by the trypan blue. Paks were monitored after renaturation by their ability to cat- exclusion of trypan blue. The positions of the 63- and 40-kDa kinases are alyze the phosphorylation of the p47-phox peptide fixed within a gel, as designated by unbroken arrows, whereas the 69- and 49-kDa kinases are described in Materials and Methods. The positions of the 69- and 63-kDa .Denotes a 38/39-kDa kinase that was frequently Paks are designated by an arrowhead and arrow, respectively ,ء .marked by arrowheads observed to undergo activation in cells stimulated with FMLP. Molecular mass markers are provided on the left. Results Effects of various sphingolipids on the activation of the

renaturable protein kinases in neutrophils by guest on September 23, 2021 Neutrophils stimulated with FMLP exhibit rapid activation of four Biomol Research Laboratories (Plymouth Meeting, PA). An affinity-puri- fied, rabbit polyclonal Ab raised against a peptide corresponding to resi- renaturable protein kinases with molecular masses of approxi- dues 525–544 of rat Pak 1 (Pak(C-19) Ab) was purchased from Santa Cruz mately 69, 63, 49, and 40 kDa (2, 3) (Fig. 1). The 63- and 69-kDa Biotechnology (Santa Cruz, CA). Anhydrous sodium sulfate, o-phthalal- enzymes have been identified as Paks (8). Recent studies indicate dehyde, arachidonate, and sphingomyelinase (520 U/mg) from Bacillus that these kinases may be forms of Pak 1 that differ in their content cereus were obtained from Sigma (St. Louis, MO). Sources of all other materials are described elsewhere (1, 2, 8). of phosphate (39). Paks can be detected directly in gels by their ability to undergo renaturation and catalyze the phosphorylation of Methods a peptide substrate uniformly fixed within a gel. The positions of these kinases are visualized by autoradiography after exposure of Preparation of neutrophils. Guinea pig peritoneal neutrophils were pre- ␥ 32 pared as described previously (38). These preparations contained Ͼ90% the gel to [ - P]ATP (2, 3). The peptide substrate utilized corre- neutrophils with viabilities always Ͼ90%. sponds to amino acid residues 297–331 of p47-phox, which con- Detection of renaturable protein kinases in polyacrylamide gels. Paks tains several of the phosphorylation sites of this protein (41). and certain other protein kinases were detected directly in gels by their ␮ Treatment of neutrophils with 40 MC2-ceramide for 5 min ability to undergo renaturation and catalyze the phosphorylation of a pep- before stimulation with 1 ␮M FMLP for 15 s blocked activation of tide substrate fixed within a gel that corresponds to residues 297–331 of p47-phox. This technique was performed as described previously (1, 2). the 63- and 69-kDa Paks along with the 49- and 40-kDa kinases

Stock solutions of FMLP (4 mM), PMA (2 mg/ml), C2-ceramide (16 (Fig. 1). C2-ceramide also reduced the basal activities of these mM), D-erythrosphingosine (16 mM), and D,L-threodihydrosphingosine (8 kinases in unstimulated neutrophils (Fig. 1C). The decreases in mM) were prepared in DMSO. D-erythrodihydroceramide (12 mM) and Pak activity were estimated by comparing the heights of the bands D-erythrodihydrosphingosine (16 mM) were prepared in ethanol. Sphin- gosine-1-phosphate (1.25 mM) was prepared in methanol. All stock solu- in lane b with those in lane d by densitometry. Treatment of neu- ␮ tions were stored at Ϫ20°C and diluted with DMSO or ethanol so that the trophils with 40 MC2-ceramide for 5 min before stimulation ﬁnal amount of solvent in each assay did not exceed 0.50% (v/v) (this with 1 ␮M FMLP for 15 s reduced the content of 32P in the 63- and includes the 0.25% added with the stimulus). These amounts of solvent did 69-kDa bands by 84 Ϯ 15% and 83 Ϯ 14% (SD, n ϭ 6), respec- not cause any of the effects noted. ␮ tively. In contrast, similar treatment of the cells with 40 MC2- Miscellaneous procedures. Procedures for immunoprecipitating Paks 32 from neutrophil lysates with the Pak(C-19) Ab and methods for detecting dihydroceramide reduced the content of P in these bands only these kinases in immune complexes by autophosphorylation are described 15 Ϯ 11% and 23 Ϯ 18% (SD, n ϭ 4), respectively (Fig. 1).

elsewhere (7, 39). Sphingosine was measured by the procedure of Merrill Neither increasing the incubation time of the cells with C2-dihy- and coworkers (31, 40). droceramide to 15 min, nor increasing the concentration of C2- Analysis of data. Unless otherwise noted, all of the autoradiographic ob- ␮ servations were conﬁrmed in at least three separate experiments performed dihydroceramide to 80 M altered these results (data not shown). on different preparations of cells. The number of observations (n) is based Effects of the various sphingoid bases on the activation of the on different cell preparations. 63- and 69-kDa Paks are summarized in Figure 2. Treatment of The Journal of Immunology 4377 Downloaded from

FIGURE 4. Effects of different concentrations of C2-ceramide and D- erythrosphingosine on the activation of the renaturable protein kinases in neutrophils. Neutrophils were treated with various amounts of C -ceramide FIGURE 3. Effects of various sphingolipids on the time course for ac- 2 tivation of the 63- and 69-kDa Paks in stimulated neutrophils. Neutrophils (A)orD-erythrosphingosine (B) for 5 min at 37°C before stimulation with ␮ 1 ␮M FMLP for 15 s. The renaturable kinases were monitored by their were incubated with 0.25% (v/v) DMSO (I; control cells), 40 MC2- ␮ ␮ ability to catalyze the phosphorylation of the p47-phox peptide ﬁxed within ceramide (II), 40 MC-dihydroceramide (III), 10 M D-erythrosphin- http://www.jimmunol.org/ 2 the gel. A, The concentrations of C -ceramide utilized were: lane b,0␮M; gosine (IV), and 10 ␮M D,L-threodihydrosphingosine (V) for 5 min prior to 2 ␮ ␮ ␮ ␮ stimulation with 1 ␮M FMLP. The cells were treated with: lane a, 0.25% lane c,50 M; lane d,40 M; lane e,20 M; and lane f,10 M. B, The ␮ (v/v) DMSO for 15 s (unstimulated cells); lane b, FMLP for 15 s; lane c, concentrations of D-erythrosphingosine utilized were: lane b,0 M; lane c, ␮ ␮ ␮ ␮ FMLP for 30 s; lane d, FMLP for 1 min; lane e, FMLP for 3 min; and lane 15 M; lane d,10 M; lane e,5 M; and lane f, 2.5 M. Lane a in A and B is for unstimulated cells not treated with C -ceramide or D-erythrosphin- f, 0.25% (v/v) DMSO for 3 min. Paks were monitored by their ability to 2 undergo renaturation and catalyze the phosphorylation of the p47-phox gosine. The positions of the 63- and 40-kDa kinases are designated by peptide ﬁxed within a gel. The positions of the 69- and 63-kDa Paks are unbroken arrows, whereas the 69- and 49-kDa kinases are marked by designated by an arrowhead and arrow, respectively. arrowheads. by guest on September 23, 2021

neutrophils with D-erythrosphingosine, D,L-threodihydrosphingosine, or D-erythrodihydrosphingosine, each at 10 ␮M, for 5 min concentrations being 20 to 40 ␮Mand5to10␮M, respectively before stimulation with 1 ␮M FMLP for 15 s reduced the content (Fig. 4). of 32P in the 63- and 69-kDa bands by 82 Ϯ 14% and 77 Ϯ 18% Sphingosine-1-phosphate can function as a second messenger (SD, n ϭ 6), 85 Ϯ 5% and 76 Ϯ 8% (SD, n ϭ 3), and 85 Ϯ 2% that modulates certain cellular responses (e.g., 22). Exposure of and 85 Ϯ 4% (SD, n ϭ 3), respectively. The complete gel profiles neutrophils to this lipid (10 ␮M) for 1 to 10 min did not trigger showing the effects of D-erythrosphingosine on the renaturable activation of the 63- and 69-kDa Paks, nor did it block activation protein kinases in unstimulated and stimulated neutrophils are pre- of these kinases upon subsequent stimulation of the cells with sented in Figure 4. It is noteworthy that the effects of D-erythro- FMLP (n ϭ 2) (data not shown). sphingosine and D,L-threodihydrosphingosine were very similar in Activation of the 63- and 69-kDa Paks can also be monitored by these studies (Fig. 2), even though these compounds differ in their the ability of these kinases to undergo autophosphorylation after stereochemistry at carbon atoms 2 and 3 and in the presence of a immunoprecipitation from lysates of stimulated neutrophils (7, 39) ␮ trans-double bond between carbon atoms 4 and 5. The significance (Fig. 5). Treatment of neutrophils with 40 MC2-ceramide or 10 of these data to the specificity of the relevant target(s) of the sphin- ␮M D-erythrosphingosine for 5 min before stimulation with 1 ␮M goid bases is dealt with in Discussion. The concentrations of sphin- FMLP for 15 s reduced this autophosphorylation reaction (Fig. 5, golipids utilized in these studies did not affect cell viability, as lanes c and e). The amounts of 32P in the 63- and 69-kDa bands ␮ ␮ measured by the exclusion of trypan blue or the release of lactate from stimulated cells treated with 40 MC2-ceramide or 10 M 2ϩ dehydrogenase. Previous studies monitoring Ca fluxes in neu- D-erythrosphingosine were reduced by 78 Ϯ 8% and 84 Ϯ 6%

trophils have established that C2-ceramide and sphingosine do not (SD, n ϭ 3) and 73 Ϯ 3% and 78 Ϯ 4% (range, n ϭ 2), respec- block the binding of FMLP to its receptor (24, 27). tively. Western blotting was performed to establish that equal amounts of Pak were immunoprecipitated in these experiments Kinetics of the inhibition of Pak activation by sphingolipids (data not shown). In contrast, addition of these lipids to the im- The 63- and 69-kDa Paks exhibited maximal activation within 15 s munoprecipitated kinases during the autophosphorylation assay of cell stimulation, followed by signiﬁcant inactivation at 3 min (2) (Fig. 5, lanes d and f) or to the phosphorylation step of the “in

(Fig. 3). Optimal amounts of C2-ceramide and the sphingoid bases gel” renaturation assay with the p47-phox peptide substrate did were effective at blocking activation of the 63- and 69-kDa Paks at not affect these reactions (n ϭ 2; data not shown). These data

all time points examined (Fig. 3, II, IV, and V), whereas C2-dihy- strongly indicate that sphingolipids at these concentrations do droceramide was ineffective at all periods tested (Fig. 3III). C2- not interact with the 63- and 69-kDa Paks themselves, but on ceramide and D-erythrosphingosine blocked activation of the 63- (an) upstream component(s) involved in the activation of these and 69-kDa Paks in a dose-dependent manner, with the effective kinases (see Discussion). 4378 SPHINGOLIPIDS AND p21-ACTIVATED PROTEIN KINASES

FIGURE 5. Effects of C2-ceramide and D-erythrosphingosine on the activation of the 63- and 69-kDa Paks in neutrophils monitored by autophosphorylation. Paks were immunoprecipitated from neutrophil lysates with the Pak(C-19) Ab and monitored by their ability to undergo autophosphorylation, as referenced in Materials and Methods. The autoradiograms shown were derived from Paks immunoprecipitated from lysates of: lane a, unstimulated cells; lane b, stimulated cells; lane c, cells treated with 40 ␮M

C2-ceramide for 5 min prior to stimulation; lane d, stimulated cells with 40 ␮ MC2-ceramide added to the immune complex during the autophosphorylation reaction; lane e, cells treated with 10 ␮M D-erythrosphingosine for 5 min prior to stimulation; and lane f, stimulated cells with 10 ␮M D- erythrosphingosine added to the immune complex during the autophosphorylation reaction. Cells were stimulated with 1 ␮M FMLP for 15 s. The

69- and 63-kDa Paks are designated by the arrowhead and solid arrow, Downloaded from respectively.

Effects of sphingomyelinase on the activation of the 63- and 69- FIGURE 7. Alterations in the activities of the 63- and 69-kDa Paks when D-erythrosphingosine or C -ceramide was added to neutrophils kDa Paks 2 shortly after stimulation with FMLP. Autoradiographs demonstrate the ef-

Treatment of leukocytes and other cell types with sphingomyeli- fects of D-erythrosphingosine (A) and C2-ceramide (B) on the activities of http://www.jimmunol.org/ nase increases the cellular content of natural, long chain ceramides the 63- and 69-kDa Paks when these lipids were added to neutrophils after and their breakdown products (e.g., sphingosine) (37, 42). For ex- stimulation with 1 ␮M FMLP. Paks were monitored by their ability to ample, incubation of HL-60 cells with sphingomyelinase (0.10– undergo renaturation and catalyze the phosphorylation of the p47-phox 0.38 U/ml) for 15 min increases the content of ceramide from 89 peptide ﬁxed within a gel, as described in Materials and Methods. A, Neu- trophils were treated with: lane a, 0.25% (v/v) DMS0 for 15 s (unstimu- to 445 pmol/106 cells (42). Neutrophils were treated with sphin- lated cells); lane b, FMLP for 15 s; lane c, FMLP for 1 min; lane d, FMLP gomyelinase to evaluate possible effects of long chain ceramides for 2 min; lane e, FMLP for 1 min with 10 ␮M D-erythrosphingosine added and/or endogenously generated sphingosine on the activation of 15 s after FMLP; and lane f, FMLP for 2 min with 10 ␮M D-erythrosphin- the 63- and 69-kDa Paks (Fig. 6). Treatment of neutrophils with gosine added 15 s after FMLP. B, Neutrophils were treated with: lane a, sphingomyelinase (2 U/ml) at 37°C for 20 min reduced the amount 0.25% (v/v) DMS0 for 15 s; lane b, FMLP for 15 s; lane c, FMLP for 2 by guest on September 23, 2021 32 ␮ of P in the 63- and 69-kDa bands from neutrophils stimulated min; and lane d, FMLP for 2 min with 40 MC2-ceramide added 15 s after with FMLP for 1 min by 66 Ϯ 11% and 74 Ϯ 10% (SD, n ϭ 3), FMLP. The positions of the 69- and 63-kDa Paks are designed by the respectively. These results appeared to reﬂect the catalytic activity arrowhead and arrow, respectively. The broken arrow indicates the position of sphingomyelinase because the inhibitory effects observed with of a 96-kDa kinase that undergoes activation at time points Ն1 min. this enzyme were both time and dose dependent. For example, blockade of Pak activation was markedly reduced if the incubation time with sphingomyelinase was reduced to 2 to 5 min or if the dose was lowered to 0.03 U/ml. Furthermore, heat inactivation of sphingomyelinase abolished its inhibitory effects on the activation of the 63- and 69-kDa Paks (n ϭ 2; data not shown). Interestingly, treatment of neutrophils with sphingomyelinase alone resulted in the activation of a 45-kDa kinase (Fig. 6, lane d, asterisk). While the identity of this kinase is not known, it did not react with the Pak(C-19) Ab during Western blotting.

Effects of certain sphingolipids on the rate of inactivation of Paks in stimulated neutrophils

FIGURE 6. Effects of sphingomyelinase on the activation of the 63- and Figure 7 summarizes the effects of adding D-erythrosphingosine or 6 69-kDa Paks in neutrophils. Neutrophils (3 ϫ 10 /ml) were incubated for C2-ceramide to neutrophils shortly after stimulation with FMLP. 20 min at 37°C in the standard assay medium in the absence (lanes a, b, Neutrophils stimulated with FMLP for 1 or 2 min with D-erythro- and c) or presence (lanes d, e, and f) of sphingomyelinase (2 U/ml). The sphingosine (10 ␮M) added 15 s after FMLP exhibited signiﬁ- cells were subsequently treated with: lanes a and d, 0.25% (v/v) DMSO for cantly less activity for the 63- and 69-kDa Paks than cells stimu- 15 s (unstimulated cells); lanes b and e,1␮M FMLP for 15 s; and lanes lated with FMLP alone for 1 or 2 min (Fig. 7A). Since these ␮ c and f,1 M FMLP for 1 min. Paks were monitored by their ability to kinases were at maximal activity at the time D-erythrosphingosine undergo renaturation and catalyze the phosphorylation of the p47-phox was added (i.e., lane b), this lipid must therefore enhance the rate peptide ﬁxed with a gel. The 69- and 63-kDa Paks are designated by an of Pak inactivation under these circumstances. Similar results were arrowhead and solid arrow, respectively. The broken arrow indicates the position of a 96-kDa kinase that undergoes activation at time periods Ն1 observed for the 49- and 40-kDa kinases. The enhanced inactiva- min, and the asterisk (*) indicates a 45-kDa kinase that undergoes activation of the 63- and 69-kDa Paks with D-erythrosphingosine was tion in sphingomyelinase-treated cells. Treatment of neutrophils with estimated by comparing the heights of the peaks in lanes c and d sphingomyelinase did not affect cell viability, as measured by the exclusion with those in lanes e and f by densitometry. Adding D-erythro- of trypan blue. sphingosine to stimulated cells reduced the amounts of 32Pinthe The Journal of Immunology 4379 Downloaded from

FIGURE 8. Structures of C2-ceramide, C2-dihydroceramide, sphingosine, and dihydrosphingosine.

63- and 69-kDa bands at 1 and 2 min by 83 Ϯ 7% and 85 Ϯ 10% undergo autophosphorylation/activation upon interacting with the http://www.jimmunol.org/ (range, n ϭ 2) and 97 Ϯ 2% and 93 Ϯ 7% (SD, n ϭ 3), respec- activated forms of Cdc42 or Rac (9). Pak can also undergo a Rac/ ␮ tively. Similarly, addition of C2-ceramide (40 M) to FMLP-stim- Cdc42-independent activation upon association with membrane or ulated neutrophils also resulted in a diminution of these activities certain lipids (36, 44). All of these reactions in vitro require several

(Fig. 7B). When C2-ceramide was added to neutrophils 15 s after minutes to 1 h for maximal activation of Pak to occur (36, 45, 46). FMLP, the amounts of 32P in the 63- and 69-kDa bands at 1 and In contrast, optimal activity of the 63- and 69-kDa Paks is ob- 2 min were reduced by 18 Ϯ 6% and 0 Ϯ 0% (SD, n ϭ 3) and served within 15 s in stimulated neutrophils (2) (Fig. 3). The small 86 Ϯ 9% and 78 Ϯ 9% (SD, n ϭ 6), respectively. The effects with adaptor protein Nck and the ␤-subunit of a heterotrimeric G pro-

D-erythrosphingosine and C2-ceramide were highly speciﬁc for the tein bind speciﬁcally to Pak and may mediate the translocation of

69-, 63-, 49-, and 40-kDa kinases, as none of the other renaturable this kinase to the membrane (44, 47–49). Pak can also form a tight by guest on September 23, 2021 enzymes in unstimulated or stimulated neutrophils exhibited a sim- complex with PIX, a guanine nucleotide exchange factor for Rac ilar loss in activity. Importantly, addition of arachidonate (10 ␮M) (50). Pak, Cdc42, and Nck undergo a very rapid (Յ15 s) translo- to neutrophils 15 s after FMLP reduced the content of 32Pinthe cation to the ruffling membranes of stimulated neutrophils (51). 63- and 69-kDa Paks at 2 min by only 18 Ϯ 4% and 19 Ϯ 8% (SD, The exact events that trigger the rapid activation of the Paks in n ϭ 3), respectively. Arachidonate is known to disorganize/desta- neutrophils remain unknown. bilize the membranes of neutrophils at this concentration and pro- The structures of C2-ceramide, C2-dihydroceramide, sphin- duces striking alterations in the morphology of these cells (43) (see gosine, and dihydrosphingosine are provided in Figure 8. It is dif- Discussion). Arachidonate alone did not trigger activation of the ficult to envision a single target that is sensitive to all four of these 63- and 69-kDa Paks (data not shown). Thus, the effects of sphin- compounds, since the inhibitory effects of C2-ceramide are depen- golipids on Pak cannot be mimicked by an agent that disorganizes dent upon the presence of the double bond between carbon atoms the membranes of these cells. 4 and 5 (Fig. 1), whereas the effects with sphingosine are not (Fig. It is possible that the effects of C -ceramide on neutrophils were 2 2). Thus, it may be reasonable to assume that C2-ceramide and the due to its degradation to D-erythrosphingosine. This possibility sphingoid bases have separate targets in neutrophils. Interestingly, was investigated and eliminated by measuring the increase in D- C -ceramide, but not dihydroceramide or sphingosine, inhibits ϫ 7 2 erythrosphingosine in neutrophils (3 10 /ml) incubated with 40 translocation of Cdc42 to the membrane of FMLP-stimulated neu- ␮ ␮ MC2-ceramide for 5 min, followed by stimulation with 1 M trophils by 82% (29). Whether C2-ceramide blocks GTP/GDP ex- FMLP for 15 s (see Materials and Methods). In three experiments, change on Cdc42 or inhibits the interaction of this small GTPase the increases in D-erythrosphingosine were 2.5, 0, and 0%. An ϩ with the membrane after activation is not known. The concentra- earlier study measuring Ca2 fluxes also concluded that significant ␮ tion of C2-ceramide that was effective against Cdc42 (50 M) (29) conversion of C2-ceramide to D-erythrosphingosine did not occur was similar to that which inhibits Pak activation (20–40 ␮M) (Fig. in human neutrophils over the duration of the experiments (27). ␮ 4). Neutrophils treated with 50 MC2-ceramide for 1 h (versus the Whether C2-ceramide was metabolized to other biologically active 5-min period utilized in these studies) contain 32 Ϯ 6 pmol of this lipids (e.g., short chain sphingomyelin, glycosphingolipids) during lipid/nmol of phospholipid (29). The concentration of long chain these experiments is not known. ceramides in unstimulated neutrophils is approximately 2.5 pmol/ nmol phospholipid (17, 29), and this value can increase 10-fold Discussion during cell stimulation due to the action of a neutral sphingomy-

In this work, we report that products of sphingolipid catabolism elinase (26, 28). Thus, the concentrations of C2-ceramide that are block activation of the 63- and 69-kDa Paks in neutrophils. More- effective against Pak are reasonable physiologically.

over, we demonstrate that these compounds can also increase the C2-ceramide, but not C2-dihydroceramide, can destabilize mem- rate of Pak inactivation when added to cells after stimulation. Paks branes at ceramide:lipid ratios of 0.2 to 0.3 (52). Data presented 4380 SPHINGOLIPIDS AND p21-ACTIVATED PROTEIN KINASES

above indicate that C2-ceramide blocks neutrophil responses at phosphorylation in stimulated neutrophils, with the phosphoryla- ratios of Յ0.033 (Fig. 4 and Ref. 29). Even if 100% of the 40 ␮M tion reaction predominating immediately after cell stimulation. In-

C2-ceramide partitioned into the membranes of guinea pig neutro- terruption of the phosphorylation reaction by sphingolipids would phils, the ceramide:lipid ratio would be only approximately 0.10. allow the dephosphorylation reaction to predominate. Second, Moreover, arachidonate markedly increases the fraction of the sphingolipids may disrupt complexes or associations that are nec- membrane in neutrophils which is liquid-cyrstalline (destabilized) essary for Pak activation (cf 50, 54) or that shield Paks from phos- as measured in ﬂuorescence polarization studies (43), but did not phatases. Third, sphingolipids may activate protein phosphatases mimick the effects of sphingolipids on Pak (see Results). These that utilize Pak as a substrate (cf 56). In all three of these situa- data indicate that the effects of C2-ceramide on the 63- and 69-kDa tions, the same targets for sphingolipids could both block activa- Paks are not the result of destabilizing the membranes of these tion of the 63- and 69-kDa Paks (e.g., Figs. 1–5) and mediate the cells. inactivation of these kinases (Fig. 7), depending upon whether

In contrast to the C2-ceramides, the ability of sphingoid bases to these lipids were added to the cells before or after stimulation. block Pak activation was not dependent upon the trans-double The levels of D-erythrosphingosine and ceramide are increased bond between carbon atoms 4 and 5, or a particular stereochem- in neutrophils during a variety of physiologic situations (26, 28, istry for the hydroxyl groups at carbon atoms 2 and 3. A similar 31). In particular, neutrophils exhibit significant increases in both situation has been reported earlier for the inhibition of PKC and sphingosine and ceramide at 15 min to 2 h after stimulation with Ϫ O2 release from neutrophils (25). One explanation for this effect is FMLP (26, 31). Since activation of Paks in FMLP-stimulated neu- that positively charged sphingoid bases may partition into regions trophils is over by 5 min (Fig. 3), it is unlikely that sphingolipids of the plasmalemma that contain acidic phospholipids, neutralize are involved in terminating the activation of these kinases, but Downloaded from these molecules, and disrupt complexes required for cell stimula- rather are active in establishing a set point for a subsequent stim- tion (e.g., 25). We have reported previously that 3-phosphorylated ulus. A similar situation has been described earlier for PKC (57). inositides are involved in the activation of the 63- and 69-kDa Paks This set point is likely to function to prevent accidental stimulation (6). D-3-phosphoinositides can activate exchange factors for Rac of neutrophils. Finally, high concentrations of ceramide and sphin- (e.g., 53). Similarly, Rac forms a stable signaling complex with goid bases are present in patients with various sphingolipid storage both a type I phosphatidylinositol-4-phosphate 5-kinase and a di- diseases (e.g., Farber’s lipogranulomatosis, type C Niemann-Pick http://www.jimmunol.org/ acylglycerol kinase (54). The acidic lipids that are products of disease) (32, 33) or cells infected with certain mycotoxins (fumo- these lipid kinases may also function in the activation of Rac (54). nisins) (e.g., 34). The possibility thus exists that inhibition of Pak It should also be noted that all four stereoisomers of sphingosine activation by these lipids may contribute to at least some of the provide the same headgroup conformation and vary only in the pathologic events associated with these disorders. position of the alkyl chain (25). A specific target for sphingoid bases could therefore exist that recognizes only the headgroup of Acknowledgments these molecules and not the hydrophobic region. Thus, few con- We are grateful to Ms. Paula Geary for her help with these experiments, clusions can be drawn from the data presented herein as to whether and Ms. Angela DiPerri for typing this paper. sphingoid bases affect a specific target protein that functions in the by guest on September 23, 2021 activation of Pak. Importantly, the amounts of endogenous sphin- References gosine in neutrophils are within an order of magnitude of the ex- ␮ Ϫ 1. Ding, J., and J. A. Badwey. 1993. Neutrophils stimulated with a chemotactic ogenous concentrations (5–10 M) that inhibit O2 release (24, 25) peptide or a phorbol ester exhibit different alterations in the activities of a battery and block activation of Pak (Fig. 4) (55). of protein kinases. J. Biol. Chem. 268:5234. 2. Ding, J., and J. A. Badwey, 1993. Stimulation of neutrophils with a chemoat- As noted above, Pak 1 can be directly activated in vitro by a tractant activates several novel protein kinases that can catalyze the phosphory- variety of sphingoid bases at concentrations of 50 to 400 ␮M (36). lation of peptides derived from p47-phox and MARKS. J. Biol. Chem. 268: In contrast, neutrophils treated with various sphingoid bases at 10 17326. ␮ 3. Grinstein, S., W. Furuya, J. R. Butler, and J. Tseng. 1993. Receptor mediated MorC2-ceramide for 5 min exhibited a diminution in the basal activation of multiple serine/threonine kinases in human neutrophils. J. Biol. activities of the 63- and 69-kDa Paks (Figs. 1 and 3, compare lane Chem. 268:20233. a in I, IV, and V) and a marked reduction in the activation of these 4. Huang, C.-K., G. F. Laramee, M. Yamazaki, and R. I. Sha’afi. 1990. Stimulation of a histone H4 protein kinase in Triton X-100 lysates of rabbit peritoneal neu- kinases after stimulation with FMLP (Figs. 2, 3, and 5). Sphingoid trophils treated with chemotactic factors: lack of requirements of calcium mobi- ␮ ␮ lization and protein kinase C activation. J. Cell. Biochem. 44:221. bases at 10 M and C2-ceramide (10–50 M) are known to block 5. Brumell, J. H., and S. Grinstein. 1994. Serine/threonine kinase activation in hu- a variety of functional responses in neutrophils (24–29). These man neutrophils: relationship to tyrosine phosphorylation. Am. J. Physiol. 267: data indicate that the effects of sphingolipids on Paks are likely to C1574. be dependent upon a number of factors that include the concen- 6. Ding, J., C. J. Vlahos, R. Liu, R. F. Brown, and J. A. Badwey. 1995. Antagonists of phosphatidylinositol 3-kinase block activation of several novel protein kinases tration of lipid, cell type employed, and/or the stimulatory pathway in neutrophils. J. Biol. Chem. 270:11684. utilized to activate these kinases. 7. Knaus, U. G., S. Morris, H.-J. Dong, J. Chernoff, and G. M. Bokoch. 1995. Regardless of the exact nature of the upstream events that trig- Regulation of human leukocyte p21-activated kinases through G-protein coupled receptors. Science 269:221. ger activation of the 63- and 69-kDa Paks, these kinases must 8. Ding, J., U. G. Knaus, J. P. Lian, G. M. Bokoch, and J. A. Badwey, 1996. The clearly undergo covalent modification during activation (i.e., phos- renaturable 69 and 63 kDa protein kinases that undergo rapid activation in chemoattractant-stimulated guinea pig neutrophils are p21-activated kinases (Paks). phorylation) (3) since the enhanced activity persists even after J. Biol. Chem. 271:24869. SDS-PAGE and the denaturation/renaturation steps (cf 1–3). It is 9. Manser, E., T. Leung, H. Salihuddin, Z.-S. Zhao, and L. Lim. 1994. A brain therefore particularly interesting that addition of C -ceramide or serine/threonine protein kinase activated by Cdc42 and Rac 1. Nature 367:40. 2 10. Wu, C., S.-F. Lee, E. Furmaniak-Kazmierczak, G. P. Cote, D. Y. Thomas, and D-erythrosphingosine to neutrophils 15 s after stimulation with E. Leberer. 1996. Activation of myosin-1 by members of the Ste20p protein FMLP increases the rate of Pak inactivation (Fig. 7). Since these kinase family. J. Biol. Chem. 271:31787. kinases were at maximal activation when the sphingolipid was 11. Brzeska, H., U. G. Knaus, Z.-Y. Wang, G. M. Bokoch, and E. D. Korn. 1997. p21-activated kinase has substrate specificity similar to Acanthamoeba myosin-1 added, the simplest explanation for this effect is that these com- heavy chain kinase and activates Acanthamoeba myosin-1. Proc. Natl. Acad. Sci. pounds enhance the dephosphorylation of Pak under these circum- USA 94:1092. 12. Zhang, S., J. Han, M. A. Sells, J. Chernoff, U. G. Knaus, R. J. Ulevitch, and stances. There are several possible mechanisms that can explain G. M. Bokoch. 1995. Rho family GTPases regulate p38 MAP kinase through the this effect. First, Paks may undergo both phosphorylation and de- downstream mediator Pak 1. J. Biol. Chem. 270:23934. The Journal of Immunology 4381

13. Bagrodia, S., B. Derijard, R. J. Davis, and R. A. Cerione. 1995. Cdc42 and 35. Brzeska, H., J. Szczepanowska, J. Hoey, and E. D. Korn. 1996. The catalytic Pak-mediated signaling leads to JUN kinase and p38 MAP kinase activation. domain of Acanthamoeba myosin I heavy chain kinase. II. Expression of active J. Biol. Chem. 270:22731. catalytic domain and sequence homology to p21-activated kinase (PAK). J. Biol. 14. Frost, J. A., S. C. Xu, M. R. Hutchison, S. Marcus, and M. H. Cobb. 1996. Chem. 271:27056. Actions of Rho family small G-proteins and p21-activated protein kinases on 36. Bokoch, G. M., A. M. Reilly, R. H. Daniels, C. C. King, A. Olivera, S. Spiegel, mitogen activated protein kinase family members. Mol. Cell. Biol. 16:3707. and U. G. Knaus. 1998. A GTPase-independent mechanism of Pak activation: 15. Manser, E., H.-Y. Huang, T.-H. Loo, X.-Q. Chen, J.-M. Dong, T. Leung, and regulation by sphingosine and other biologically active lipids. J. Biol. Chem. L. Lim. 1997. Regulation of phosphorylation pathways by p21- GTPases: the p21 273:8137. Ras-related Rho subfamily and its role in phosphorylation signalling pathways. 37. Kaga, S., S. Ragg, K. A. Rogers, and A. Ochi. 1998. Activation of p21-Cdc42/ Eur. J. Biochem. 242:171. Rac-activated kinases by CD28 signalling: p21-activated kinase (PAK) and MEK 16. Frost, J. A., H. Steen, P. Shapiro, T. Lewis, N. Ahn, P. E. Shaw, and M. H. Cobb. kinase 1 (MEKK1) may mediate the interplay between CD3 and CD28 signals. 1997. Cross-cascade activation of ERKs and ternary complex factors by Rho J. Immunol. 160:4182. family proteins. EMBO J. 16:6426. 38. Badwey, J. A., and M. L. Karnovsky. 1986. NADH-oxidase and aldehyde oxidase 17. Suchard, S. J., P. J. Mansﬁeld, L. A. Boxer, and J. A. Shayman. 1997. Mitogen- from polymorphonuclear leukocytes. Methods Enzymol. 132:365. activated protein kinase activation during IgG-dependent phagocytosis in human 39. Lian, J. P., and J. A. Badwey. 1997. Activation of the p21-activated protein neutrophils. J. Immunol. 158:4961. kinases from neutrophils with an antibody that reacts with the N-terminal region 18. Kuroki, M., and J. T. O’Flaherty. 1997. Differential effects of a mitogen-activated of Pak 1. FEBS Lett. 404:211. protein kinase inhibitor on human neutrophil responses to chemotactic factors. 40. Merrill, A. H., Jr., E. Wang, R. E. Mullins, W. C. Jamison, S. Nimkar, and Biochem. Biophys. Res. Commun. 232:474. D. C. Liotta. 1988. Quantitation of free sphingosine in liver by high performance 19. Downey, G. P., J. R. Butler, H. Tapper, L. Fialkow, A. R. Saltiel, B. R. Rubin, liquid chromatography. Anal. Biochem. 171:373. and S. Grinstein. 1997. Importance of MEK in neutrophil microbicidal respon- 41. El Benna, J., L. P. Faust, and B. M. Babior. 1994. The phosphorylation of the siveness. J. Immunol. 160:434. respiratory burst oxidase component p47-phox during neutrophil activation. 20. Zu, Y.-L., J. Qi, A. Gilchrist, G. A. Fernandez, D. Vazquez-Abad, D. L. Kreutzer, J. Biol. Chem. 269:23431. C.-K. Huang, and R. I. Sha’aﬁ. 1998. p38-mitogen activated protein kinase ac- 42. Yang, Z., M. Costanzo, D. W. Golde, and R. N. Kolesnick. 1993. Tumor necrosis tivation is required for human neutrophil function triggered by TNF-␣ or fMLP

factor activation of the sphingomyelin pathway signals nuclear translocation of Downloaded from stimulation. J. Immunol. 160:1982. ␬B translocation in intact HL-60 cells. J. Biol. Chem. 268:20520. 21. Hannun, Y. A. 1996. Functions of ceramide in coordinating cellular responses to 43. Badwey, J. A., J. T. Curnutte, J. M. Robinson, C. B. Berde, M. J. Karnovsky, and stress. Science 274:1855. M. L. Karnovsky. 1984. Effects of free fatty acids on release of superoxide and 22. Spiegel, S., and A. H. Merrill, Jr. 1996. Sphingolipid metabolism and cell growth on change of shape by human neutrophils: reversibility by albumin. J. Biol. regulation. FASEB J. 10:1388. Chem. 259:7870. 23. Kolesnick, R., and D. W. Golde. 1994. The sphingomyelin pathway in tumor 44. Lu, W., S. Katz, R. Gupta, and B. J. Mayer. 1997. Activation of Pak by membrane necrosis factor and interleukin-1 signalling. Cell 77:325. localization mediated by Nck. Curr. Biol. 7:85. 24. Lambeth, J. D., D. N. Burnham, and S. R. Tyagi. 1988. Sphingonine effects on 45. Martin, G. A., G. Bollag, F. McCormick, and A. Abo. 1995. A novel serine chemoattractant-induced diacylglycerol generation, calcium fluxes, superoxide http://www.jimmunol.org/ kinase activated by Rac1/Cdc42 Hs-dependent autophosphorylation is related to production, and on cell viability in the human neutrophil. J. Biol. Chem. 263: Pak 65 and STE 20. EMBO J. 14:1970. 3818. 46. Manser, E., C. Chong, Z.-S. Zhao, T. Leung, G. Michael, C. Hall, and L. Lim. 25. Merrill, A. H., Jr., S. Nimkar, D. Menaldino, Y. A. Hannun, C. Loomis, 1995. Molecular cloning of a new member of the p21-Cdc42/Rac activated kinase R. M. Bell, S. R. Tyagi, J. D. Lambeth, V. L. Stevens, R. Hunter, and D. C. Liotta. (PAK) family. J. Biol. Chem. 270:25070. 1989. Structural requirements for long-chain (sphingoid) base inhibition of pro- 47. Bokoch, G. M., Y. Wang, B. P. Bohl, M. A. Sells, L. A. Quilliam, and tein kinase C in vitro and for the cellular effects of these compounds. Biochem- U. G. Knaus. 1996. Interaction of the Nck adapter protein with p21-activated istry 28:3138. kinase (Pak 1). J. Biol. Chem. 271:25746. 26. Nakamura, T., A. Abe, K. J. Balazovich, D. Wu, S. J. Suchard, L. A. Boxer, and J. A. Shayman. 1994. Ceramide regulates oxidant release in adherent human 48. Galisteo, M. L., J. Chernoff, Y.-C. Su, E. Skolnick, and J. Schlessinger, 1996. The neutrophils. J. Biol. Chem. 269:18384. adaptor protein Nck links receptor tyrosine kinases with the serine-threonine kinase Pak 1. J. Biol. Chem. 271:20997. 27. Wong, K., X.-B. Li, and N. Hunchuk. 1995. N-acetylsphinosine (C2-ceramide) inhibited neutrophil superoxide formation and calcium influx. J. Biol. Chem. 49. Leeuw, T., C. Wu, J. D. Shrag, M. Whiteway, D. Y. Thomas, and E. Leberer. by guest on September 23, 2021 ␤ 270:3056. 1998. Interaction of a G-protein -subunit with a conserved sequence in Ste20/ 28. Suchard, S. J., V. Hinkovska-Galcheva, P. J. Mansfield, L. A. Boxer, and PAK family protein kinases. Nature 391:191. J. A. Shayman. 1997. Ceramide inhibits IgG-dependent phagocytosis in human 50. Manser, E., T.-H. Loo, C.-G. Koh, Z.-S. Zhao, Z.-Q. Chen, L. Tan, I. Tan, polymorphonuclear leukocytes. Blood 89:2139. T. Leung, and L. Lim. 1998. Pak kinases are directly coupled to the PIX family 29. Abousalham, A., C. Liossis, L. O’Brien, and D. N. Brindley. 1997. Cell-perme- of nucleotide exchange factors. Mol. Cell 1:183. able ceramides prevent the activation of phospholipase D by ADP-ribosylation 51. Robinson, J. M., J. A. Badwey, and P. G. Heyworth. 1997. Rapid accumulation factor and RhoA. J. Biol. Chem. 272:1069. of key signal transduction molecules to leukocyte phagosomes. Mol. Cell. Biol. 30. Mullmann, T. J., M. I. Siegel, R. W. Egan, and M. M. Billah. 1991. Sphingosine 8:258a (Abstr.). inhibits phosphatidate phosphohydrolase in human neutrophils by a protein ki- 52. Simon, C. G., Jr., and A. R. L. Gear. 1998. Membrane-destabilizing properties of nase C-independent mechanism. J. Biol. Chem. 266:2013. C2-ceramide may be responsible for its ability to inhibit platelet aggregation. 31. Wilson, E., E. Wang, R. E. Mullins, D. J. Uhlinger, D. C. Liotta, J. D. Lambeth, Biochemistry 37:2059. and A. H. Merrill, Jr. 1988. Modulation of the free sphingosine levels in human 53. Nimnual, A. S., B. A. Yatsula, and D. Bar-Sagi. 1998. Coupling of Ras and Rac neutrophils by phorbol esters and other factors. J. Biol. Chem. 263:9304. guanosine triphosphatases through the Ras exchanger Sos. Science 279:560. 32. Moser, H. W., A. L. Prensky, H. J. Wolfe, N. P. Rosman, S. Carr, and G. Ferreira. 54. Tolias, K. F., A. D. Couvillon, L. C. Cantley, and C. L. Carpenter. 1998. Char- 1969. Farber’s lipogranulomatosis: report of a case and demonstration of an ex- acterization of a Rac1- and RhoGDI-associated lipid kinase signaling complex. cess of free ceramide and ganglioside. Am. J. Med. 47:869. Mol. Cell. Biol. 18:762. 33. Goldin, E., C. F. Roff, S. P. F. Miller, C. Rodriguez-Lafrasse, M. T. Vanier, 55. Merrill, A. H., Jr. 1991. Cell regulation by sphingosine and more complex sphin- R. O. Brady, and P. G. Pentchev. 1992. Type C Niemann-Pick disease: a murine golipids. J. Bioenerg. Biomembr. 23:83. model of the lysosomal cholesterol lipidosis accumulates sphingosine and sphin- 56. Dobrowsky, R. T., and Y. A. Hannun. 1992. Ceramide stimulates a cytosolic ganine in liver. Biochim. Biophys. Acta 1127:303. protein phosphatase. J. Biol. Chem. 267:5048. 34. Wang, E., W. P. Norred, C. W. Bacon, R. T. Riley, and A. H. Merrill, Jr. 1991. 57. Merrill, A. H., Jr., and V. L. Stevens. 1989. Modulation of protein kinase C and Inhibition of sphingolipid biosynthesis by fumonisms: implication for diseases diverse cell functions by shingosine: a pharmacologically interesting compound associated with Fusarium moniliforme. J. Biol. Chem. 266:14486. linking sphingolipids and signal transduction. Biochim. Biophys. Acta 1010:131.