THEJOURNAL OF BIOLOGICALCHEMISTRY Vol. 264, No. 28, Issue of October 5, pp. 1637&16382,1989 Q 1989 by The American Society for Biochemistryand Molecular Biology, Inc. Printed in U.S. A. , a Novel ras-related Family of Proteins ThatAre Substrates*

(Received for publication, March 7, 1989)

John DidsburyS, Richard F. Weber, Gary M. BokochQT, Tony Evans, and Ralph Snyderman From the Department of Pharmacological Sciences, Genentech, Incorporated, South SanFrancisco, California 94080 and the $Department of Immunology, Research Institute of ScrippsClinic, LA Jolla, California 92037

A new family of ras-related proteins,designated rac (for review see Ref. 1). The a-subunits (Mr 39,000-52,000) (cas-related C3 botulinum toxin substrate) has been bind and hydrolyze GTP, andconsensus sequences are impli- identified. racl and rac2 cDNA clones were isolated cated in these activities (1). The second class are the “low from a differentiated HL-60 library and encode pro- molecular weight GTP-binding proteins” with M, values of teins that are 92%homologous and share58% and 26- 20,000-30,000 of which ras is the archetype (for review see 30% amino acid homology with human rhos and ras, Ref. 2). In addition to H- (3), N- (4), and K-ras(5), this group respectively. Nucleotide sequenceanalysis predicts includes yeast YPTl (6), RHO (7), SEC4 (€9, simian rul (9), both racl and rac2 proteins to contain192 amino acids with molecular masses of 21,450 and 21,429daltons, aplysia and human rhos (10, II), murine yptl (12), four rat respectively. racl and rac2 possess four of the five rub homologues (13), three related bovine smgs (14), three conserved functional domains in ras associated with human rap homologues (15, 16), and bovine ARF (17). Each binding and hydrolysis of guanine nucleotides. They of these low molecular weight GTP-binding proteinspossesses also contain the COOH-terminal consensus sequence consensus sequences for GTP binding and hydrolysis which Cys-X-X-X-COOH which localizes ras tothe inner are homologous to the Ga proteins. To date, definitive func- plasma membrane and the residuesGly” and Alas8, at tions have not been ascribed to any of the ras or ras-related which sites mutations elicit transforming potential to proteins although YPTl and SEC4 have been implicated as ras. The rac transcripts, particularly rac2, display regulators of the secretory pathway of yeast Saccharomyces relative myeloid tissueselectivity. Both racl tran- cereuisiue (18,19). Guanine nucleotides stimulate secretion in scripts (2.4 and 1.1 kilobases (kb)) increase whenHL- mammalian cells independent of calcium increases and phos- 60 cells differentiate to neutrophil-like morphology. pholipase C activation (20,21), and low molecular weight In contrast, differentiationof U937 cells tomonocyte- GTP-binding proteins may be involved in this process since like morphology causes no changein the2.4-kb mRNA microinjected ras protein induces degranulation in mast cells and a decreasein the1.1-kb mRNA species. rac2 (22). Interestingly, low molecular weight GTP-binding pro- mRNA (1.45 kb) increases 7-9-fold and %fold upon teins copurify with the formylpeptide chemoattractant recep- differentiation of HL-60 and U937cells, respectively. tor from differentiated HL-60 cells.’ This receptor stimulates Neither mRNAs are present ina Jurkat cell line, rac T exocytosis as well as chemotaxis of leukocytes (23). and unlikeracl, rac2mRNA is absent in human brain and liver tissue. Transfection experiments permitted ADP-ribosylation of cellular proteins by a number of bac- the demonstration that racl and rac2 are substrates terial toxins (i.e. cholera, pertussis, pseudomonas exotoxin A, for ADP-ribosylation by the C3 component of botuli- and diphtheria) is the primary mechanism for their toxicity num toxin.The datasuggest that racs are plasma mem- (for review see Ref. 24). An NAD-dependent ADP-ribosyl- brane-associated GTP-binding proteins which could transferase is associated with active components of botulinum regulate secretory processes, particularly in myeloid toxins. The botulinum neurotoxins block release of neuro- cells. transmitters at presynaptic nerve terminals by an unknown mechanism (25) and have also been reported to inhibit exo- cytosis in bovine adrenal cells (26). Botulinum toxins ADP- ribosylate 21-26-kDa eukaryotic proteins (27-29). The only Two classes of GTP-binding proteins have been identified substrate for botulinum toxin-inducedADP-ribosylation thus in mammalian tissues. G proteinsare heterotrimers consisting far identified has been the ras-related rho protein which has of a, p, and y subunits. These proteins transduce receptor- been found predominantly in neural tissues (27, 30, 31). mediated signals across the plasma membrane to specific In thispaper we identify a novel class of human rm-related effector systems including adenylate cyclase, cyclic GMP proteins which are substrates for ADP-ribosylation by botu- phosphodiesterase, phospholipases, and certain ion channels linum toxin C3 ADP-ribosyltransferase. Two cDNAs have * This work was supported, in part, by Grant 5R01-DE-03738-16 been isolated encoding proteins 92% homologous to each from the National Institute of Dental Research (to R. S.) and Na- other. The proteins, designated rucl and ruc2, are 58% ho- tional Institute of Health Grant GM39434 (to G. M. B.). The costs mologous with rho proteins and 26-30% homologous with ras of publication of this article were defrayed in partby the payment of proteins. Northern analysis of rucl and ruc2 mRNAs show page charges. This article must therefore be hereby marked “aduer- them to differ from one another in their tissue distribution tisement” in accordance with 18 U.S.C. Section 1734 solely to indicate and their amounts upon cellular differentiation of myeloid this fact. The nucleotide sequencefs) reported in thispaper has been submitted precursor cells to either neutrophil or monocyte-like mor- to the GenBankTM/EMBLData Bank with accession number(s) phology. 505038. 4 To whom correspondence should be addressed. Polakis, P., Evans, T., and Snyderman, R. (1989) Biochem. Bio- 7 Established Investigator of the American Heart Association. phys. Res. Commun. 161, 276-283. 16378 racs, Novel Botulinum ToxinSubstrates 16379

A =3 washing in 0.1 X SSC, 0.1% SDS at 42 “C. Transfections-cDNA inserts for racl and rac2 were cloned in the sense and antisense directions in the eukaryotic expression vector pRK5 under control of a cytomegalovirus enhancer/promoter (39). Transient expression in COS 7 cells was carried out using the DEAE- p16 dextran procedure (40) with DEAE-dextran at 400 Kg/ml and chlo- roquine at 100 KM during the 3-h incubation with DNA. Forty to forty-eighth post-transfection cells were washedwith phosphate- buffered phosphate and removed from the plates with 5 mM EDTA

B 35 in phosphate-buffered phosphate for 5 min at 37 ”C. Cells from each 60-mm plate were resuspended in 100 p1 of 10 mM Tris-HC1, pH 7.6, 1 mM EDTA, and 100 KM phenylmethylsulfonyl fluoride and sub- jected to two cycles of freeze thawing in dry-ice/EtOH and stored at -80 “C. ‘IO ADP-ribosylation by Botulinum Toxin-ADP-ribosylation was per- formed by the procedure described in (41) in the presence of 2 FM [32P]NAD (5000 cpm/pmol) and 0.5 Kg/ml of the purified C3 ADP- ribosyltransferase of botulinum toxin, obtained as in(28) and kindly H300 bp provided by Dr.Klaus Aktories (UniversitatGiessen). G22K was purified as described in (41) and labeled in the presence of 2.5 ~1 of FIG. 1. Schematic representation of rac cDNA clones. A, neutrophil cytosol. racl clones 3 and 16; B, rac2 clones 5 and 10. All cDNAs are EcoRI restriction fragments and aligned 5’ to 3’. The coding regions are RESULTS AND DISCUSSION indicated by shaded boxes. The arrows under the maps indicate the direction and region of DNA sequencing. Isolation of rac cDNA Clones-The amino acid sequence FDTAGQEDYD obtained from purified placental G25K (32) is a highly conserved region of all ras-related proteins thatis EXPERIMENTALPROCEDURES proposed to contribute to the formation of the guanine nu- Cloning and Sequencing of DNA-An oligonucleotide, 5”TT cleotide-binding site (42-45). An oligonucleotide based on this (T/C)GA(T/C)AC AGCTGGCCA(A/G)GA(A/G)GA(T/C)TA(T/C)sequence (see “Experimental Procedures”)was used to screen GA(T/C) based on thesequence of G25K (32) Phe-Asp-Thr-Ala-Gly- Gln-Glu-Asp-Tyr-Asp was synthesized. The oligonucleotide was end 600,000 recombinant phageplaques from a differentiated labeled with T4 polynucleotide kinase and [Y-~’P]~ATP andused to HL-60 cDNA library under low stringency conditions. Sev- screen a XgtlO phage library constructed from poly(A)+ RNA of HL- enteen positive clones were isolated. cDNA inserts from each 60 cells differentiated with dibutyryl cyclic AMP (33). The hybridi- clone were subcloned into M13mp19 andtheir nucleotide zation was carried out under low stringency in 20% formamide, 5 X sequences determined. The aminoacid sequence deducedfrom SSC (20 X SSC is 3 M NaC1, 0.3 M sodium citrate), 0.1%sodium one of the clones containing a 0.8-kb insert revealed astriking pyrophosphate, 50 mM sodium phosphate, pH 6.5, 5 X Denhardt’s solution (34), 0.1% SDS’ and 50 wg/ml denatured herring spermDNA homology with rho proteins (10, 11).The open reading frame at 37 “C with washing at 47 “C in 0.5 X SSC, 0.1% SDS. cDNA was in this clone lackedan initiator methionine andwas presumed labeled by random priming andused for high stringency screening in not to be full length. To isolate full length clones the library the above buffer containing 50% formamide at 42 “Cwith washing at was rescreened with the 0.8-kb insert cDNA, labeled with32P, 42 “C in 0.1 X SSC, 0.1% SDS. cDNA inserts were subcloned into under high stringency conditions. Twenty positive cloneswere M13mp19 (Fig. 1) and sequenced by the method of Sanger et al. (35). Sequence homology alignments were performedaccording tothe isolated. Restriction mapping andnucleotide sequence deter- method of Dayhoff et al. (36). mination of eight clones revealed afull length clone identical RNA Analysis-Human brain, liver, placenta, and HeLacell RNAs to the original cDNA as well as a second highly homologous passaged one timeover oligo(dT)-cellulose were obtained from Clon- full length clone. Fig. 2 shows the nucleotide and deduced tech Laborat,ories Inc. (Palo Alto, CA). RNAs from U937 cells, U937 amino acidsequences of these twoclones which we have cells differentiated with dibutyrylcyclic AMP (37), human peripheral designated racl and rac2 for ~s-related-C3botulinum toxin blood mononuclear cells, and SK7 (T cell hybridoma) cells were also purified by one passage over oligo(dT)-cellulose. RNAs from Jurkat substrate. The two cDNAs contain open reading frames en- (T cell), HL-60, and HL-60cells differentiated with dibutyryl cyclic coding proteins with a length of 192 amino acids with calcu- AMP (33) were purified by two passages over oligo(dT)-cellulose. All lated M, values of 21,450 and 21,429, respectively. The nu- RNAs were isolated by the method of Chirgwin et al. (38). A 45-mer cleotide sequence homology in the coding region of racl and antisense oligonucleotide (5”ATAGAATGTGAGTCTGAACTCT- rac2 is 77%. The differences in nucleotide sequence are - TACAT TTAGAACAAACAAAACCT-3’) complementary to 3’-non- domly distributed suggestingeach cDNA is derived from coding sequenceof racl cDNA was end labeled withT4 polynucleotide kinase and [Y-~’P]~ATP(34) and used as a racl-specific probe. Two separate (Fig. 2). 45-mer oligonucleotides complementary to 3’-noncoding sequence Amino Acid Sequence Homology-The amino acid sequence of rac2 cDNA with a 15-basepair overlap (5“TGTGAGAGTC (Fig. 3) of racl and rac2 are most homologous (58%) with TTCAAACTTTTAAACCTTGCCAGTCAGGACTTTTG-3’ and 5’- human rhos 6, 9, and 12 (10, 11) and with human ras (26- GGGTTGAGT TGGGTTTTCTATTTGCAATAGCAAAAGTCCT- 30%, (2)).Although the rac clones were selected with a probe GACTG-3’) were annealed and labeled by filling in withKlenow enzyme in the presence of [32P]dATP and [32P]dCTP andused as a based on the sequence of G25K, neither rac protein is G25K rac2 specific probe. One pg of each RNA and10 pg of total neutrophil based on the lack of homology with other known regions of RNA was electrophoresed on formaldehyde gels (34) and transferred amino acid sequence for G25K (32). Five regions of the ras to nitrocellulose. Hybridization with the racl oligonucleotide probe molecule are postulated tobe involved in guanine nucleotide was carried out in 50% formamide, 1 X Denhardt’s solution, 5 X SSC, binding and hydrolysis; residues 10-16, 57-62, 116-120, 145- 1%SDS, and 200 pg/ml of denatured herring sperm DNA at 37 “C 147, and 152-165 (2, 45). The sequences of rad and rac2, as and washed in 2 X SSC, 0.1% SDS at 42 “C. Hybridization with the rac2 synthetic DNA was carried out in the same buffer at 42 “C with well as the three humanrhos are highly conserved in four of these regions with considerable divergence in the fifthregion (ras residues 152-165). There are also minor differences in The abbreviationsused are: SDS,sodium dodecyl sulfate; GTP+, guanosine 5’-O-(thio)triphosphate;G25K, a low molecular weight the NH2-terminalregion putatively involved in GTP hydrol- GTP-bindingprotein initiallypurified from humanplacenta; kb, ysis (ras residues 10-16 (45)). The functionaleffects of such kilobases. divergence upon binding andhydrolysis remain tobe assessed. 16380 racs, Novel Botulinum Toxin Substrates

10 UOAIKCVVVGDGAVGKT 123 4 5 6 7 8 9101112 A'IGUGGCU~GX;n;rr~GA;CGGAGCn;TAGGTMMC ...... &""'l)ll- "- - 2.4b rac2 A'IGUGGCCA~Gn;n;nn~AGA~CCG~~GAC MOAIKCVVVGDGAVGKT racl 2n xn CL-~ISYTTNAFP-~EYI I -1.lkb roc1 ~CCTA~TCAGT~~U~CK~~rATATCC...... rac2 CXiCC-~GCTACACCAC~ffiCCGGIGAC CLLISYTTNAFPGEYI

rac1 rac2 -1.45kb rac2 CUCffin;-CMCTA~GC~TG~n~~~GCCA PTVFDNYSANVMVDSXP FIG.4. Northern analysis of racl and rac2RNAs. The fol- "" VNLGLWDTAGOEDYDRL lowing human RNAs are in each lane; I, HeLa cell; 2, liver; 3, brain; roc1 G""WXATACUXlGGA~lTA~CAGA~...... 4, placenta; 5, U937 cell; 6, differentiated U937 cell; 7, SK-7 cell; 8, rac2 Gl'GMCCXGGGCKTGGGACACKClGGGCUXAGGA~ACGACCGTCT Jurkat cell; 9, HL-60 cell; 10, differentiated HL-60 cell; 11, peripheral VHLGLWDTAGOEDYDRL blood mononuclear cell; 12, peripheral blood neutrophil. RNAs were 7n nn RP'LSYPOTDVFLICFS" purified by either one (lanes 1-7, and 11) or two (8-10) passages rac1 ACGCCCCCTATCCTA~CCCAUC4GA~TGlTClTMTlXCnTXCC through oligo(dT)-cellulose, and 1 pg of each was electrophoresed...... (10 roc2 CCGGCCGCTCTCCTATCC4WACGGACGTCTXCP3lC?VXTCTCCC Total RNA pg) from neutrophils was used (lane 12). Autoradi- RPLSYPOTDVFLICFS ography was for 4 days except rac2 lanes 11 and 12 where exposure 90 100 was for 1 day. LVSPASFENVRAKWYPE rac1 Tn;'IGAGTCCTGUTTCCClWX&AGTGGTATCClGAG...... roc2 mTCAGCCUGCCTC?TA~~ffiTCCGrrCGCCCCAIL;n;Gm Two lines of evidence strongly suggest that racl andrac2 are LVSPASYENVRAKWFPE GTP-binding proteins. First,a protein purified from platelets 110 VRHHCPNTPIILVGTKL on thebasis of it's ability to bind GTPwas subjectedto amino rac1 Gn;n;GCACCACn;TCCC~CUTCA~~~CT~CT...... acid sequence analysis, and thesequence obtained from three rac2 G'IG;CGGUCCACKCCCUGUCACCCA~TC~~UC~GCT peptides matched exactly the amino acid sequence encoded VRHHCPSTPIILVGTKL by the racl cDNA (57). Second, racl and rac2 contain all the 120 130 DLRDDKDTIEKLKEKK consensus amino acid sequences that have been identified in n;ATC~ACGGAn;AT~CACGA~A~~~~G~ rac1 ...... ras as GTP-binding domains(2). rac2 GGACCn;CGGGACGAUAGGACACCATCCAGUACXAAGGAGMGMGC DLRDDKDTIEXLXEKK Two amino acids in ras, Gly" and Ala5', which when mu- 140 150 tated causeras to be oncogenic, are also present in rad,rac2, LTPITYPQGLA~AKEIG and rho. This raises the possibility that mutations in these rac1 ...... rac2 ?CGCTCCUTCACCTACCffiCAGGGCC?GGCACKGCCMGGXATKAC sites in rac could alter the regulationof the functionof these LAPITYPOGLALAKEID proteins. 160 ras proteins havebeen localized to the innersurface of the AVKYLECSALTORGLKT rac1 GCXTMMTACC-U-C...... plasma membrane with attachmentrequiring acylation of the rac2 TCU;IC~TACCTU;AGmeCC~CCCA~AGG~C COOH-terminal cysteine by palmitic acid (46). The COOH- SVKYLECSALTORGLKT terminal sequenceCys-X-X-X-COOH found in ras is also 170 180 VFDEAIRAVLCPPPVK found in racl and rac2 and may indicate that these proteins rac1 AG'IG""AffiMCF~ffFCAFF~~~CCCGCCTCCffi~MGA...... can also associate with the plasma membrane. rac2 CG'IGTTCGACGAGGCCATCCGGGCCGTGC'IGTGCCCTCAGCCCAffiCGGC VFDEAIRAVLCPQPTR The areas of greatest amino acid sequence divergence be- 190 tween racl and rac2 and the other low molecular weight G KRKRKCLLL- rac1 AGAGGMGAGPAMTGCC'IGC'IGTIGTM proteins occur at residues 22-27, 49-54, 121-153, and 170- ...... rac2 AGCAGMGCGCGCC'IGCAGCCTCCTCTAG 188 implicating theseregions in effector and/or detectorfunc- OOKRACSLL- tions. The greatest divergence between racl and rac2 occurs FIG.2. Nucleotide and deduced amino acid sequencesof rac within one of the above regions at residues 180-190 which is cDNAs. Identical nucleotides are indicated by a star. Amino acids also the hypervariable region of ras proteins. The variability are numbered. inthis region may give differentfunctional properties to members of the ras family of proteins (2) and potentially to rad versus rac2. Deletion of the hypervariable region in ras does not alter theknown biochemical properties of the mole- rac1 - cule or its transforming ability(47), although biological assays rac2 - rhol2- other than transformation may be needed to fully determine rho9 - rho6 - the role of this region in the functionof ras. N-ras- H-raS- Tissue Distribution of rac RNAs-We examined a number 70 80 90 100 110 120 130 of human tissues andcell lines for expression of rac mRNAs roc1 - LRPLSYPOTDVnfCFSLVSPASFENYRAKWYPFVR-HHCPNTPIILVGT~L~DK-~IE~EK .'...... '.....'....'.~.y.~.....~....-....~.~...... ,.~...... rac2 - .'...... '.....'....'.~.y.~.....~....-....~.~...... ,.~...... by Northernblot analysis. Tworacl mRNA species were rhol2- rho9 - detected with lengthsof 2.4 and 1.1 kb (Fig. 4). Although racl rho6 - N-ras- mRNA was not present ina Jurkat T cell line, asmall amount H-ras- was detected in SK-7 cells, a T cell hybridoma. This may racl - indicate that a racl negative regulatory element is normally rac2 - rho12- present in lymphocytes but is lost uponhybridoma formation. rho9 - rho6 - The ratioof racl 2.4- and 1.1-kb species is roughly equivalent N-ras- H-raS- in U937, HL-60, differentiated HL-60, and human neutrophils (lanes 5, 9, 10, and 12) all of which are myeloid in origin. In FIG.3. Alignment of racl and rac2 protein sequences with human rhos and human N-ras and H-ras. Amino acids are the other tissueshowever, the ratio isdifferent, with the 2.4- designated by single-letter notation. Residues identical to racl amino kb mRNA being enriched over the 1.1-kb mRNA, which is acids are indicated with a star. Reference numbering is that of racl. particularly strikingin placental tissue(lane 4).Furthermore, racs, Novel Botulinum Toxin Substrates 16381 the increased amount of racl 1.1-kb mRNA relative to the 2.4-kb species appears to exhibit neutrophilicselectivity. In- -97kD deed, differentiation of U937 and HL-60 cells to monocyte- -66kD uersus neutrophil-like morphology results in a decrease in the -43kD 1.1-kb mRNA (lanes 5 uersus 6)for differentiated U937 cells and an increase (lanes 9 uersus 10) in differentiated HL-60 cells. -31kD rac2 is encoded by a single 1.45-kb mRNA. (Fig. 4). Com- pared with racl, rac2 mRNA is less ubiquitously distributed -21kD with little or no RNA detected in brain,liver, or Jurkat cells (lanes 2,3, and 8). In contrast toracl mRNA, differentiation -14kD of U937 cells toward monocytescaused a 2-3-fold increase in rac2 mRNA (lanes 5 uersus 6).Differentiation of HL-60 cells 12345678 also resulted in an increase in rac2 mRNA (7-9-fold (lanes 9 uersus 10)).rac2 mRNA is abundant in all cells of myeloid FIG. 5. ADP-ribosylation of rac by botulinum C3 ADP-ri- origin (Fig. 4, lanes 5, and 9-12), in particular, mature bosyltransferase. rac cDNAs, antisense (-) and sense (+), were 6, transfected into COS cells and whole cell lysates were ADP-ribosy- neutrophils (lane 12).Note that therac2 hybridization signal lated as described under “Experimental Procedures,” then run on a for neutrophils (lane 12) was generated from less poly(A)+ 13% polyacrylamide gel and autoradiography performed for 16 h at RNA with an exposure time one-fourth of that for RNAs in -70 “C with an intensifying screen. C3 ADP-ribosyltransferase is lanes 1-10. The only nonmyeloid cell that appears to havea included in each sample except as indicated. Lane I, 100 ng of G22K relatively large amount of rac2 mRNA is HeLa (lune 1 ), the without C3 added; lane 2, 100 ng of G22K; lane 3, 7.5 pl of racl (-) reason for which is unknown. lysate; lane 4,7.5 p1 of racl (+) lysate; lane 5,7.5 pl of rac2 (-) lysate; lane 6, 7.5 pi of rac2 (+) lysate; lone 7, 20 pg of neutrophil plasma While racl and rac2 proteinsare 92% identicalto one membrane without C3 added; lane 8, 20 pg of neutrophil plasma another, suggesting similar functions, there are distinct dif- membrane. G22K isthe endogenous neutrophil botulinum toxin ferencesin the tissue distribution of their RNAs possibly substrate previously described by Bokoch et al. (41). indicating functionsthat are notidentical. Both racl and rac2 could be important in differentiation of myeloid precursor rapl protein (41). rapl protein was cloned from a differen- cells with rac proteins potentiallyplaying a role in neutrophil tiated HL-60 library and expressed but did not serve as a function. The absenceof any detectablerac RNAin the Jurkat botulinum toxin substrate.3 It is thus likely that thepurified T cell line would indicate that racs are not housekeeping G22K preparation (41) contains both rapl and rac but that proteins but required by specialized cells, in particular cells rac represents the botulinum toxin substrate activity in puri- of myeloid origin. It is interesting to note thattissue-specific fied G22K. distribution of another low molecular weight GTP-binding Botulinum toxins have been shown to inhibit neurotrans- protein, smg p25A hasalso been demonstratedin neural mitter release and exocytosis (26, 54). A direct role for botu- tissues (48). linum toxin ADP-ribosylating activityas mediating inhibition rac Proteins as Substrates for Botulinum Toxin-Botulinum of exocytosis has not yet been definitively demonstrated. In toxins C1 and D contain an ADP-ribosyltransferase activity fact, some evidence suggests otherwise. In electropermeabil- that is able to ADP-ribosylate21-26-kDa eukaryotic proteins. ized chromaffin cells, ADP-ribosylating activityof botulinum Ribosylation isenhanced by addition of GDP,GTP, or toxins did not correlate with inhibitionof secretion. Antibody GTPyS (29, 49, 50), suggesting that these toxin substrates to C3 toxin blocked ADP-ribosylation by C3 and D toxins are GTP-binding proteins. It is not yet clear whether the without blocking the inhibitoryeffect of D toxin on secretion ADP-ribosyltransferaseactivities of toxinsC1 and D are (55). In addition, ribosylation of a 22-kDa substrate by botu- integral components of the toxins or due to a distinct third linum toxin D in rabbit neutrophils did not alter the cells ADP-ribosyltransferaseactivity, called C3, isolated from ability to secrete the azurophilic enzyme 8-glucosaminidase these toxin preparations (51,52). (56). However, the authors were able to ADP-ribosylateonly To date, the onlyidentified substrate for ADP-ribosylation 70% of the endogenous 22-kDa botulinum substrate and no by botulinum toxins was rho (27, 30, 31). Based on the high specific granulemarkers were examined. Nonetheless, by degree of homology between racs and rho proteins (Fig. 3), we analogy with yeast YPTl and SEC4 as regulators of secretion sought to determine whetherracs are substrates for botulinum(18, 19) and evidence that microinjected ras induces degran- toxin. racl and rac2 were transiently expressed in COS cells, ulation (22) it islikely that a number of low molecular weight and whole cell lysates of transfected cells weretested for their GTP-binding proteins are involved in regulating secretory ability to be ADP-ribosylated by botulinum C3 ADP-ribosyl- processes. The presence of large amounts of rac RNA in transferase. The results presented in Fig. 5 show significant secretory cells (i.e. neutrophils and monocytes) and the fact increases in ADP-ribosylation of an -22-kDa protein in cells that they are botulinum toxin substratesleads one tohypoth- esize that these proteins are candidatesfor such a regulatory transfected with racl and rac2 sense DNAs (lanes 4 and 6) function. over cells transfected with antisense DNAs (lanes 3 and 5). racl and rac2 comigrate with theendogenous neutrophil mem- REFERENCES brane botulinum toxin substrate (lane 8) as well as with the 1. Gilman, A. G. (1987) Annu. Reo. Biochern. 56,615-649 purified substrate from neutrophils,designated G22K,de- 2. Barbacid, M. (1987) Annu. Reu. Biochern. 56, 779-827 scribed previously (41). racl, rac2, G25K, and the protein(s) 3. Shimizu, K., Goldfarb, M., Perrucho, M., and Wigler, M. (1983) designated G22K migrate with an apparent molecular mass Proc. Natl. Acad. Sci. U. S.A. 80, 383-387 of 25 kDa in an alternativegel system.3 ADP-ribosylation of 4. DeFeo, D., Gonda, M., Young, H., Chang, E., Lowry, D., Scolnick, E.,and Ellis, R. (1981) Proc. Natl. Acnd. Sci.U. S. A. 78,3328- G22K by botulinum toxin C3 component exhibits therequire- 3332 ment for a cytosolic protein for optimal labeling, unlike rho (41, 53), and can also be differentiated from rho by immuno- L. A. Quilliam, C. J. Der, and G. M. Bokoch, unpublished obser- logical criteria (53). The purified G22K was found to contain vations. 16382 rues, Novel Botulinum Toxin Substrates 5. Ellis, R., DeFeo, D., Shih, T., Gonda, M., Young, H., Tsuchida, 33. Didsbury, J. R., and Snyderman, R. (1987) FEBS Lett. 2 19,259- N., Lowry, D., and Scolnick, E. (1981) Nature 292, 506-511 263 6. Gallwitz, D., Donath, C., and Sander,C. (1983) Nature 306,704- 34. Maniatis, T., Fritsch, E. F., and Sambrook, J. (1982) Molecular 707 Cloning: A Laboratory Manual, Cold Spring HarborLaboratory, 7. Madaule, P., Axel, R., and Myers, A. (1987) Proc. Natl. Acad. Sci. Cold Spring Harbor, NY U. S. A. 84, 779-783 35. Sanger, F., Nicklen, S., and Coulson, A. R. (1977) Proc. Natl. 8. Salminen, A., and Novick, P. (1987) Cell 49,527-538 Acad. Sci. U. S. A. 74,5463-5467 9. Chardin, P., and Tavitian,A. (1986) EMBO J. 5, 2203-2208 36. Dayhoff, M. O., Barker, W. C., and Hunt, L. T. (1983) Methods 10. Madaule, P., and Axel, R. (1985) Cell 41,31-40 Enzymol. 91, 524-545 11. Chardin, P., Madaule, P., and Tavitian, A. (1988) Nucleic Acids 37. Didsbury, J. R., Ho, Y.-S., and Snyderman, R. (1987) FEBS Lett. Res. 16, 2717 211,160-164 12. Haubruck, H., Disela, C., Wagner, P., and Gallwitz, D. (1987) 38. Chirgwin, J. M., Przybyla, A. E., MacDonald, R. J., and Rutter, EMBO J. 6,4049-4053 W. J. (1979) Biochemistry 18, 5294-5299 13. Touchot, N., Chardin, P., and Tavitian, A. (1987) Proc. Natl. 39. Eaton, D. L., Wood, W. I., Eaton, D., Hass, P. E., Hollingshead, Acad. Sci. U. S. A. 84,8210-8214 P., Wion, K., Mather, J., Lawn, R. M., Vehar, G. A., and 14. Matsui, Y., Klkuchi, A., Kondo, J., Hishida, T., Teranishi, Y., Gorman, C. (1986) Biochemistry 25,8343-8347 and Takai, Y. (1988) J. Biol. Chem. 263, 11071-11074 40. Gorman, C. (1985) in DNA Cloning ZZ (Glover, D. M., ed.) pp. 15. Pizon, V., Chardin, P., Lerosey, I., Olofsson, B., and Tavitian, A. 143-190, IRL Press, Washington, D. C. (1988) Oncogene 3, 201-204 41. Bokoch, G. M.,Parkos, C. A., and Mumby, S. M. (1988) J. Biol. 16. Pizin, V., Lerosey, I., Chardin, P., and Tavitian,A. (1988) Nucleic Chem. 263,16744-16749 Acids Res. 16, 7719 42. La Cour, T. F. M., Nyborg, J., Thirup, S., and Clark, B. F. C. 17. Sewell, J. L., and Kahn, R. A. (1988) Proc. Natl. Acad. Sci. U. S. (1985) EMBO J. 4, 2385-2388 A. 85,4620-4624 43. Halliday, K. R. (1984) J. Cyclic Nucleotide Protein Phosphoryla- 18. Segev, N., Mulholland, J., and Botstein, D. (1988) Cell 52, 915- tion Res. 9,435-448 44. 924 McCormick, F., Clark, B. F. C., LaCour, T. F. M., Kjeldgaard, M., Norskov-Lauritsen, L., and Nyborg, J. (1985) Science 19. Goud, B., Salminen, A., Walworth, N. C., and Novick, P. J. (1988) 230, 78-82 Cell 53,753-768 45. DeVos, A. M., Tong, L., Milburn, M. V., Matias, P. M., Jancarik, 20. Cockroft, S., Howell, T. W., and Gomperts, B. D. (1987) J. Cell. J., Noguchi, S., Nishimura, S., Miura, K., Ohtsuka, E., and Biol. 105, 2745-2750 Kim, S.-H. (1988)Science238,888-893 21. Barrowman, M. M., Cockroft, S., and Gomperts, B. D. (1986) 46. Fujiyama, A,, and Tamanoi, F. (1986) Proc. Natl. Acad. Sci. U. Nature 319, 504-507 S. A. 83,1266-1270 22. Bar-Sagi, D., and Feramisco, J. R. (1986) Science 233, 1061- 47. Lacal. J. C.. Anderson. P. S., and Aaronson, S. A. (1986) EMBO 1066 J. 5,6791687 23. Snyderman, R., Smith, C. D., and Verghese, M. (1986) J. Leu- 48. Sano. K.. Kikuchi. A.. Matsui. Y.. Teranishi. Y., and Takai, Y. kocyte Biol. 40, 785-800 (1989) 'Biochem.'Biophys. Res. Commun. 168, 377-385 24. Eidels, L., Proia, R. L., and Hart, D.A. (1983) Microbiol. Rev. 49. Ohashi, Y., and Narumiya, S. (1987) J. Biol. Chem. 262, 1430- 47,596-620 1433 25. Simpson, L. L. (1981) Pharmacol. Rev. 33,158-188 50. Aktories, K., Weller, U., and Chhatwal, G. S. (1987) FEBS Lett. 26. Knight, D. E., Tonge, D. A., and Baker, P. F. (1985) Nature 317, 212,109-113 719-721 51. Aktories, K., and Frevert, J. (1987) Biochem. J. 247, 363-368 27. Kikuchi, A,, Yamamoto, K., Fujita, T., and Takai, Y. (1988) J. 52. Rosener, S., Chhatwal, G. S., and Aktories, K. (1987) FEBS Lett. Biol. Chem. 263, 16303-16308 224,38-42 28. Aktories, K., Rosener, S., Blaschke, U., and Chhatwal, G. S. 53. Quilliam, L. A,, Lacal, J.-C., and Bokoch, G. M. (1989) FEBS (1988) Eur. J. Biochem. 172, 445-450 Lett. 247,221-226 29. Matsuoka, I., Syuto, B., Kurihara, K., and Kubo, S. (1987) FEBS 54. Simpson, L. L. (1986) Annu. Rev. Pharmacol. Toricol. 26, 427- Lett. 216,295-299 453 30. Narumiya, S., Sekine, A., and Fujiwara, M. (1988) J. Biol. Chem. 55. Adam-Vizi, V., Rosener, S., Aktories, K., andKnight, D. E. 263,17255-17257 (1988) FEBS Lett. 238,277-280 31. Braun, U.,Habermann, B., Just, I., Aktories, K., and Vandek- 56. Mege, J. L.,Volpi, M., Becker, E. L., and Shaafi, R. I. (1988) erckhove, J. (1989) FEBS Lett. 243,70-76 Biochem. Biophys. Res. Commun. 152,926-932 32. Polakis, P., Snyderman, R., and Evans, T. (1989) Biochem. Bio- 57. Polakis, P. G., Weber, R. F., Nevins, B., Didsbury, J. R., Evans, phys. Res. Commun. 160, 25-32 T., andSnyderman, R. (1989) J. Biol. Chem. 264,16383-16389