Characterization, expression and chromosomal localization of ... - Nature

2 downloads 0 Views 584KB Size Report
gene homologous to the mouse Lsc oncogene, with strongest expression in hematopoetic tissues. Hans-Christian Aasheim1, Florence Pedeutour2 and Erlend B ...
Oncogene (1997) 14, 1747 ± 1752  1997 Stockton Press All rights reserved 0950 ± 9232/97 $12.00

SHORT REPORT

Characterization, expression and chromosomal localization of a human gene homologous to the mouse Lsc oncogene, with strongest expression in hematopoetic tissues Hans-Christian Aasheim1, Florence Pedeutour2 and Erlend B Smeland1 1 2

Department of Immunology, Institute for Cancer Research, The Norwegian Radium Hospital, Montebello, N-0310, Oslo, Norway; URA CNRS 1462, Faculty de Medecine, Avenue de Valombrose, Nice, France

A human cDNA clone, denoted sub1.5, was isolated from cDNA library generated from human T cells. The sub1.5 cDNA sequence was novel and was not identical to any known cDNA sequences in the GenBank. Recently, however, a mouse cDNA (Lsc) with high homology to sub1.5 was identi®ed, indicating that the sub1.5 sequence may represent the human homologue of the mouse Lsc gene. The sub1.5 cDNA includes an open reading frame of 875 amino acids, predicting a protein with molecular weight of 97 kDa. Like Lsc, sub1.5 shows homology to the previous described oncogene Lbc, in particular to two functional domains in the Lbc protein; the Dbl protooncogene homology domain and the pleckstrin homology domain. Lsc is proposed to be an oncogene and is a member of a growing family of proteins that may function as activators of the Rho family GTPases. Members of the Rho family regulates the polymerization of actin to produce stress ®bers. Activation of Rho GTPases by sub1.5 is also indicated by our studies, as stress ®ber formation is observed in serum-starved stable NIH3T3 sub1.5 transfectants. Sub1.5 cDNA hybridizes to two major transcripts of 3.5 and 5 kb size and the strongest expression is seen in hematopoietic tissues like thymus, lymph nodes, peripheral blood leukocytes and spleen. We also show that both puri®ed B and T cells express sub1.5. In addition, our data indicate that sub1.5 mRNA is abundantly expressed in CD34+ human progenitor cells. Fluorescent in situ hybridisation, using sub1.5 cDNA as a probe on human metaphases, shows that the sub1.5 gene is localized to chromosome 19q13.13. Keywords: Lsc homologue; Rho GEF; Dbl-family; hematopoetic expression Members of the Ras superfamily of proteins function as molecular switches in a diversity of cellular signalling pathways, in¯uencing processes such as cytoskeletal organization, development, vesicular transport, cell polarity and cell motility (Adams et al., 1990; Boguski and McCormic, 1993; Ridley et al., 1992; Ridley and Hall, 1992). The Ras superfamily consists of more than 50 members that share the ability to bind and hydrolyze GTP. The superfamily can be divided into several subfamilies, designated Ras, Rho/Rac, Rab, Ran, and Arf/Sar (For review see Boguski and Correspondence: H-C Aasheim Received 8 November 1996; revised 17 December 1996; accepted 17 December 1996

McCormick, 1993; Quilliam et al., 1995; Denhardt, 1996). The Ras family itself is involved in triggering cell proliferation in response to mitogens and growth factors via the Raf/MEK/mitogen-activated protein kinase signalling pathway leading to a phosphorylation cascade which activates transcription factors to induce gene expression (Denhardt, 1996). The Rho/Rac family is involved in cytoskeletal organisation and focal contacs (Ridley and Hall, 1992; Ridley et al., 1992) and has been suggested to participate in growth factor signalling (Nobes and Hall, 1995; Coso et al., 1995; Minden et al., 1995; Hill et al., 1995) and to be involved in regulation of receptor mediated endocytosis through coated pits (Lamaze et al., 1996). The Ras superfamily of GTP-binding proteins alternate between alternate active GTP-bound and inactive GDP-bound states (for review see Quilliam et al., 1995). The GTP-binding/GTPase cycles are tightly controlled, with guanidine nucleotide exchange factors (GEFs) catalyzing their conversion to the GTP-bound active state and GTPase-activating proteins (GAPs) ensuring their return to an inactive, basal state through the stimulation of GTP hydrolysis (Boguski and McCormick, 1993). GEFs have been described for virtually all the Ras GTPase families (Boguski and McCormick, 1993). Recently, the product of the Dbloncogene was shown to serve as a GEF for speci®c members of the Rho branch of the Ras superfamily (RhoA, Rac1, CDC42HS; Hart et al., 1994). It was shown that both the GEF activity and the transforming activity of the Dbl protein was localized to a 240 amino acid motif between residues 498 and 738 in the protein (Hart et al., 1994). An emerging group of proteins conferring homology to this 240 amino acid motif in the Dbl oncogene has been identi®ed including Vav (Katzav et al., 1989), Ect-2 (Miki et al., 1993), Lbc (Toksoz and Williams, 1994), Lfc (Whitehead et al., 1995), Tim (Chan et al., 1994), Tiam (Habets et al., 1994), Cdc24 (from Saccharomysis cerevisia; Drubin, 1991), the breakpoint cluster region protein (Bcr; Hariharan and Adams, 1987), and the mammalian ras guanidine releasing factor (Shou et al., 1992). All Dbl family members possess a second shared domain, designated the plekcstrin homology (PH) domain. The PH domain was initially identi®ed as a region of sequence homology, of approximately 120 amino acids, which is duplicated in pleckstrin (reviewed in Lemmon et al., 1996; Gibson et al., 1994). Pleckstrin is the major substrate of protein kinase C in platelets. The PH domain is found in a large variety of proteins involved in cellular signalling and cytoskeletal func-

Identification of the homologue to the Lsc oncogene H-C Aasheim et al

1748

tions. It is believed that the function of the PH domain is to target the host protein to the cell membrane, by binding to lipids or proteins (Lemmon et al., 1996), to facilitate and regulate enzymatic activities. cDNA sequence and predicted protein structure of the sub1.5 cDNA sequence We have identi®ed and sequenced a novel cDNA encoding a putative protein with homology to two recently described members of the Dbl-family of oncogenes, Lbc and Lfc (Toksoz and Williams, 1994; Whitehead et al., 1995). A novel 3.4 kb cDNA was isolated from a human TPA-stimulated T cell cDNA library using a subtractive strategy recently described (Aasheim et al., 1994, 1996). The sequence revealed an open reading frame (ORF) of 875 amino acids, starting with an ATG codon at nucleotide 436 and terminating with a stop codon at nucleotide 3044 (data not shown, accession number Y09160). The ®rst ATG codon is in moderately good context for translation initiation with the sequence, CACCTCATGG as compared with the Kozak consensus sequence (XCXGCCATGG;Kozak, 1994). Recently, a mouse cDNA (Lsc, Whitehead et al., 1996) was published with 87% overall homology at the amino acid level to the sub1.5 sequence. The overall amino acid homology, the similar expression pattern and the transcript sizes strongly suggest that sub1.5 is the human homolog of the corresponding mouse gene. The mouse protein is denoted Lsc (Lbc's second cousin) and is suggested to be an oncogene (Whitehead et al., 1996). The Lsc cDNA was isolated based on the ability to induce strong oncogenic transformation when expressed in NIH3T3 ®broblast cells. The Lsc sequence is proposed to start at an ATG 53 amino acids upstream of the start of the sub1.5 sequence (Figure 1). Sub1.5 and Lsc diverge from nucleotide number 334 in the sub1.5 sequence and 5' upstream. Here, the two sequences do not show any homology to each other. The sub1.5 sequence does not harbour an ATG for translation start of the corresponding site where Lsc is proposed to start (data not shown). The sub1.5 cDNA encodes a protein with a predicted molecular weight of 97 kDa with several interesting features. Thus, based on homology to previous known genes, the predicted sub1.5 protein and Lsc show homology to the Dbl family of GEF's, and most closely related to the Dbl family member Lbc, an exchange factor with speci®city for Rho family GTPases (Toksoz and Williams, 1994; Zheng et al., 1995). Like all Dbl family members, sub1.5 contains a PH domain in tandem with the DH-domain. Both Sub1.5 and Lsc are distinguished from Lbc by an extended N-terminus (more than 400 amino acids) that does not appear to be necessary for, or inhibitory to cellular transformation (Whitehead et al., 1996). This area does not align signi®cantly to any other proteins when performing GenBank searches. The DH-domain is found in a number of proteins suspected to be involved in cell growth regulation. These include Cdc24 (Zheng et al., 1994), Tiam-1 (Habets et al., 1994), Vav (Katzav et al., 1989), Ect2 (Miki et al., 1993), Ost (Horii et al., 1994), Lbc (Zheng et al., 1995) and Lfc (Whitehead et al., 1995). This domain has been shown to be both essential for the

Figure 1 Comparison of the amino acid sequence of sub1.5 and Lsc. The sequences were optimally aligned on the basis of identical residues (vertical lines). The DH-domain and the PHdomain is localized from amino acid 359 to 710 in the sub1.5 sequence

transforming activity and the GEF activity of oncogenic Dbl. So far GEF activity has been shown for Dbl (Hart et al., 1994), Ost (Horii et al., 1994), Cdc24 (Zheng et al., 1994), Tiam-1 (Van Leeuwen et al., 1995) and Lbc (Zheng et al., 1995), all exhibiting exchange activity in vitro on Rho family GTPases. All DH family members, including sub1.5 and Lsc, also possess a PH domain. PH domains are thought to anchore proteins to cell membranes, either through interaction with lipids or with proteins (Gibson et al., 1994; Lemmon et al., 1996). Both sub1.5 and Lsc possess a PH domain. The transforming activity of both the Lfc and Lbc oncogene has been shown to be dependent of the PH domain (Whitehead et al., 1995, 1996) indicating that a PH-dependent recruitment of Dbl family members to the plasma membrane may be a necessary step for transformation in NIH3T3 cells. NIH3T3 stable transfectants and stress ®ber formation Whitehead et al. (1996) have shown that the transforming activity of the Lsc protein is dependent of exchange

Identification of the homologue to the Lsc oncogene H-C Aasheim et al

activity for Rho family GTPases. The experiments were performed by cotransfection of Lsc with p190 GAP, which has GAP activity for several Rho family GTPases (RhoA, RhoB, Rac1, Rac2, and Cdc42) in in vitro assays (Settleman et al., 1992), and by use of dominant-negative mutants of Rac1, RhoA and Cdc42. The transforming activity of Lsc, as observed by focus formation of NIH3T3 cells, was blocked by all the above mentioned constructs. These data provide support for Lsc-triggered activation of Rho

Fluorescence

signalling pathways and are consistent with the Rho family exchange activity that has been attributed to Dbl family GEF's. In mammalian cells, RhoA regulates actin stress ®ber formation and focal adhesion assembly following growth factor stimulation (Ridley and Hall, 1992). The expression of the Lbc oncogene was found to be associated with stress ®ber formation in serum starved NIH3T3 cells (Zheng et al., 1995), a typical response of Rho stimulation. Rho activation by the sub1.5 gene product is indicated

Phase contrast

Figure 2 E€ects of sub1.5 expression on actin stress ®bers formation. NIH3T3 cells were cotransfected with the sub1.5 cDNA, in the plasmid pCDM8, and the plasmid pPUR (Clontech, CA) conferring resistance to puromycin, using the lipophilic agent DOTAP (Boehringer Mannheim, Germany) essentially as described by the manufacturer (5 mg sub1.5 plasmid and 0.5 mg of pPUR plasmid per 10 cm plate). Control cells were transfected with the pPUR plasmid alone. The cells were grown in selection medium containing 4 mg/ml puromycin. Sub1.5 expression in stable transfectants was con®rmed by Northern analysis (data not shown). Actin ®laments were detected by staining with rhodamine-labelled phalloidin, an actin binding protein as previously described (Ridley et al., 1992). Brie¯y, cells were seeded in 15 mm wells containing 13 mm circular coverslips. The cells were serum starved in DMEM containing 0.5% serum over night, after which they were ®xed in 4% paraformaldehyde for 15 min, permeabilized with 0.1% Triton X-100, and incubated with rhodamine-labelled phalloidin (Molecular probes), to localize actin. Cells were viewed on a Zeiss microscope and photographed. Phalloidin stained sub1.5 NIH3T3 transfectants (a) and phalloidin stained control transfectants (c). Corresponding phase contrast is shown in (b) (sub1.5 transfectants) and d (control transfectants)

1749

Identification of the homologue to the Lsc oncogene H-C Aasheim et al

1750

by our studies showing formation of stress ®bers ending in focal adhesions in stable sub1.5 NIH3T3 transfectants after over-night serum starvation as compared to control cells (Figure 2). The control cells showed few stress ®bers and a punctuated distribution of ®lamentous actin at the plasma membrane, as previously described for Swiss 3T3 cells (Ridley and Hall, 1992). We also observed a better survival of the sub1.5 transfectants after serum starvation than control transfectants. These results support the data from studies with the mouse Lsc which indicate activation of the Rho signalling pathways (Whitehead et al., 1996). Chromosomal localization of the sub1.5 gene Using the FISH technique and the sub1.5 cDNA as a probe, we demonstrate that the chromosomal localization of sub1.5 was to chromosome 19q13.13 (Figure 3). The chromosomal localization was con®rmed by simultaneous hybridization of sub1.5 with a control probe located on 19q13.1 and using two-color detection. Signals of the two probes were shown to be colocalized on the long arm of chromosome 19, the sub1.5 signal appeared to be slightly distal to a control probe (chromosome 19q13.1 probe, data not shown). Other genes localized to chromosome 19q13.1 are myeloid b (A4) precursor like protein, hepsin, CD22, ATP4, Cytochrome C oxidase subunit VIb and VIIa polypeptide 1, glucose phosphatate isomerase, ryanodine receptor 1 peptidase D and CAAT/Enhancer binding protein a (source LLNL Human Genome Center). Genomic Southern blot data indicate that the two major transcripts are trancribed from the same gene (data not shown).

Expression pattern of the sub1.5 gene Sub1.5 was expressed at relatively high levels in blood and lymphoid tissues (thymus, spleen, tonsils and lymph nodes), and at weaker levels in most other tissues (Figure 4 a ± c). Two main transcripts were detected with the sizes of 3.5 and 5 kb, respectively, and the relative expression levels of these two transcripts varied between tissues. The expression pattern of sub1.5 roughly parallells the expression of Lsc, which also showed a particular abundant expression in murine hematopoetic tissues. In mice, Lsc showed a high expression in total bone marrow, while the Sub1.5 gene was expressed only at moderate levels in total bone marrow in humans. This di€erence may be due to di€erent regulation of the genes between mouse and man, or to di€erences in the isolation procedure of bone-marrow cells. Highly puri®ed B cells and T cells from blood and tonsils expressed sub1.5 at high levels, in agreement with the tissue blot data (Figure 4d). While more mature B lymphoid cells lines (Raji, Daudi and U266) all abundantly expressed sub1.5, the pre-B cell lines Nalm-6 and Reh expressed lower levels of sub1.5 mRNA. A di€erence in the level of sub1.5 expression during B lymphopoesis was further supported by semi-quantitative PCR analysis, demonstrating low levels of sub1.5 transcripts in early B lymphoid cells (pro- and pre-B cells) and higher levels of expression in bone marrow derived B cells expressing surface IgM (Figure 5). From the Northern blot data on thymus, which show strong expression of both transcripts, one would suspect that T cells express the gene from the early di€erentiation stages, indicating that the regulation of sub1.5 expression is di€erently regulated during T cell development versus

Figure 3 Chromosomal localization of the sub1.5 gene. Fluorescence in situ hybridization (FISH) was performed as previously described (Pedetour et al., 1994). Biotinylated sub1.5 probe was hybridized to 20 metaphases from three unrelated healthy donors (two males and one female). Revelation of the signal was done using avidin-¯uorescein. The chromosomes were counterstained with propidium iodide (a and b) and 4,5-diamino-2-phenyleindole (DAPI, for chromosome identi®cation; c and d). Slides were observed using a Zeiss Axiophot microscope and metaphases were analysed using an image processor (Perspective Scienti®c International, Chester, UK). Arrowheads indicate chromosome 19 with a ¯uorescent signal at 19q13.13 band. The idiogram illustrates the chromosomal localization at 19q13.13

Identification of the homologue to the Lsc oncogene H-C Aasheim et al

—4.4 kb

a spleen

—7.5 kb

1751

H2O CD34+,CD38– CD34+,CD38+ CD34+,CD19–

thymus

CD34+,CD19+

prostate

CD19+,IgM– CD19+,IgM+

testis

PBL B cells ovary 26x

small intestine

sub1.5

colon PBL

b heart brain placenta lung liver skeletal muscle kidney pancreas

c fetal liver bone marrow PBL appendix thymus lymph node

—3.5 kb

spleen

d B cells

30x

B cells TPA B cells tonsils CD4 cells CD4 cells TPA CD8 cells CD8 cells TPA Figure 4 Expression of sub1.5 mRNAs in di€erent human tissues. (a, b and c) Represent di€erent commercial tissue blots (Multiple human tissue blot 1 and 2, human immunsystem multiple tissue blot; Clontech, CA). Each lane shows hybridization to 2 mg mRNA of the indicated tissue. (d) Represents a Northern blot of B cells, CD4 cells and CD8 cells isolated from blood or tonsils by positive selection using magnetic Dynabeads coated with anti-CD19, anti-CD4 and anti-CD8, respectively (Funderud et al., 1990, Rasmussen et al., 1992). mRNA was isolated from the cells using Dynabeads oligo-(dT)25 essentially as described (Aasheim et al., 1994). Each lane in d shows hybridization to 1 mg mRNA of the indicated cell type. Molecular weights are indicated in kb on top of a. PBL = peripheral blood leucocytes, TPA = 12-O-tetra decanoylphorbol-

26x actin

Figure 5 Semi-quantitative PCR analysis of levels of sub1.5 in di€erent hematopoetic cell populations. CD34+ cells were positively selected from adult bone marrow aspirates as described (Rusten et al., 1994) and costained with phycoerythrin (PE)conjugated anti-CD34 and ¯uorescein (FITC) conjugated antiCD38, or PE anti-CD34 and FITC anti-CD19. CD19+B cells (depleted for CD34+ cells) were positively selected from adult bone marrow aspirates as described (Funderud et al., 1990; Rasmussen et al., 1992) and costained with PE anti-CD19 and FITC anti-IgM. B cells were also isolated from blood as described (Funderud et al., 1990). Cells were sorted on a FACSortTM+ instrument (Becton Dickinson) after appropriate compensation, as described elsewhere (Rusten et al., 1994). After sorting, the cells were centrifuged and lysed with cold lysis bu€er (phosphate bu€ered saline (PBS), 1% NP-40) at concentrations of 26104 to 16105 cells/mL. mRNA was isolated and cDNA was generated as described (Aasheim et al., 1996). cDNA generated from mRNA isolated from 5 6 103 cells were used as template in the ®rst PCR (32 cycles) using sub1.5 speci®c primers 1 and 2 hybridizing to region 2045 ± 2065 and 2382 ± 2400 respectively. One ml of the ®rst PCR reaction was used as template in the second PCR using a second inner set of sub1.5 speci®c primers (primer 3 and 4) hybrizing to region 2096 ± 2113 and 2178 ± 2199 respectively. Primer 1:5'-AAA TTC TAC ACC ACG TCA ACC, Primer 2: 5'-GGT CAT GGC GGA GGT GAG C, Primer 3: 5'-GGC TCA AGG ACT ATC AGC G, Primer 4: 5'-CAA TTT CTT CTT GGT GAT GTC C. Actin speci®c primers were used for normalization of the amount of cDNA from each cell population (right panel). Here only a ®rst PCR reaction was performed starting with 2 6 103 cells. Sub1.5 speci®c PCR results are shown in left and middle panel, actin speci®c PCR results are shown in right panel. Cycle numbers in the second PCR (sub1.5) and ®rst PCR (actin) are shown below the panels. Five ml aliquots were removed from the PCR reaction mixture at the indicated cycle number, seperated on an 2% agarose gel and processed for hybridization with random labelled [32P]dCTP sub1.5 insert or actin cDNA

B cell development. However, we can not exclude the possibility that the 5 kb transcript is expressed during early B cell di€erentiation, but is not recognized by the primer combinations we have used, due to lack of knowledge about where the sequence di€erence between the two di€erent transcripts is located. Interestingly, we also demonstrated high levels of sub1.5 expression in early hematopoetic progenitor cells (Figure 5). Thus, sub1.5 was abundantly expressed in both CD34+CD387 cells, which are highly enriched in the most primitive hematopoetic progenitor cells, as well as CD34+CD38+ cells, containing primarily early di€erentiation stages of the lymphoid and myeloid lineages. However, it should be noted that these semi-

13-acetate. The Northern blots were hybridized with random labelled [32P]dCTP sub1.5 insert

Identification of the homologue to the Lsc oncogene H-C Aasheim et al

1752

quantitative PCR data only give indications with regard to the expression of the sub1.5 gene and have to be further con®rmed by protein expression studies using antibodies. We are currently in the process of generating such antibodies. In conclusion a human cDNA clone, sub1.5, was isolated from a T cell cDNA library, and showed to be strongest expressed in lymphoid tissues. The predicted protein confers homology to the Dbl-family of GEF's, harbouring both a DH-domain and a PH-domain in addition to an extended N-terminus without known function. Stable NIH3T3 transfectants showed stress ®ber formation upon serum starvation, indicating activation of Rho-like GTPases by sub1.5. Sub1.5 is

highly homologous to the previously published mouse cDNA, Lsc. Based on the similarities in the amino acid sequence, and in the expression pattern of sub1.5 and Lsc, we propose that sub1.5 most probably represents the human homologue of Lsc.

Acknowledgements This work was supported by the Norwegian Cancer Society. We are grateful to Toril Larsen and Ruth Solem for technical assistance, and Raisa Gurvitsj for secretarial help. We also wish to thank Dr Harald Stenmark for valuable discussions and suggestions. Sub1.5 accession number is Y09160.

References Aasheim H-C, Deggerdal A, Smeland EB and Hornes E. (1994). Biotechniques, 16, 716 ± 721. Aasheim H-C, Logtenberg T and Larsen F. (1996). Methods in Molecular Biology ± cDNA Library Protocols. Cowell I and Austin C. (eds.). Humana Press. Vol 69. Adams AEM, Johnson DI, Longnecker RM, Sloat BF and Pringle JR. (1990). J. Cell Biol., 111, 131 ± 142. Boguski MS and McCormick F. (1993). Nature, 366, 643 ± 654. Chan AML, McGovern ES, Catalano G, Fleming TP and Miki T. (1994). Oncogene, 9, 1057 ± 1063. Coso OA, Chiariello M, Yu J-C, Teramoto H, Crespo AÊ , Xu N, Miki N and Gutkind JS. (1995). Cell, 81, 1137 ± 1146. Denhardt DT. (1996). Biochem. J., 318, 729 ± 747. Drubin DG. (1991). Cell, 65, 1093 ± 1096. Funderud S, Eriksten B, AÊsheim H-C, Nustad K, Blomho€ HK, Holte H and Smeland EB. (1990). Eur. J. Immunol., 20, 201 ± 206. Gibson JB, HyvoÈnen M, Musacchio A and Saraste M. (1994). Trends Biochem. Sci., 19, 349 ± 353. Habets GGM, Scholtes EHM, Zuydgeest D, van der Kammen RA, Stam JC, Berns Allmenn and Collard JG. (1994). Cell, 77, 537 ± 549. Hariharan IK and Adams JM. (1987). EMBO J., 6, 115 ± 119. Hart MJ, Eva A, Zangrilli D, Aaronsen SA, Evans T, Cerione RA and Zheng Y. (1994). J. Biol. Chem., 269, 62 ± 65. Hill CS et al. (1995). Cell, 81, 1159 ± 1170. Horii Y, Beeler JF, Sakaguchi K, Tachibana M and Miki T. (1994). EMBO J., 13, 4776 ± 4786. Katzav S, Martin-Zanca D and Barbacid M. (1989). EMBO J., 8, 2283 ± 2290. Kozak M. (1994). J. Biol. Chem., 266, 19867 ± 19873. Lamaze C, Chuang T-H, Terlecky LJ, Bokoch GM and Schmid S. (1996). Nature, 382, 177 ± 179.

Lemmon MA, Ferguson KM and Schlessinger J. (1996). Cell, 85, 621 ± 624. Miki T, Smith CL, Long JE, Eva A and Fleming TP. (1993). Nature, 362, 462 ± 465. Minden A, Lin A, Claret F-X, Abo A and Karin M. (1995). Cell, 81, 1147 ± 1157. Nobes CD and Hall A (1995). Cell, 81, 53 ± 62. Pedetour F, Szpirer C and Nahon JL. (1994). Genomics, 19, 31 ± 33. Rasmussen A-M, Smeland EB, Erikstein B, Caignault L and Funderud S. (1992). J. Imm. Methods., 146, 195 ± 202. Ridley AJ and Hall A. (1992). Cell, 7, 389 ± 399. Ridley AJ, Paterson HF, Johnston CL, Diekman D and Hall A. (1992). Cell, 70, 401 ± 410. Rusten LS, Jacobsen SEW, Kaalhus O, Veiby OP, Funderud SF and Smeland EB. (1994). Blood, 84, 1473 ± 1481. Settleman J, Albright CF, Foster LC and Weinberg RA. (1992). Nature, 359, 153 ± 154. Shou C, Farnsworth CL, Neel BG and Feig LA. (1992). Nature, 358, 351 ± 354. Toksoz D and Williams DA. (1994). Oncogene, 9, 621 ± 628. Quilliam LA, Khosravi-Far R, Hu€ SY and Der CJ. (1995). BioEssays, 17, 395 ± 404. Van Leeuwen FN, van der Kammen RA, Habets GGM and Collard JG. (1995). Oncogene, 11, 2215 ± 2221. Whitehead I, Kirk H, Tognon C, Trigo-Conzales G and Kay R. (1995). J. Biol. Chem., 270, 18388 ± 18395. Whitehead I, Khosravi-Far R, Kirk H, Trigo-Conzales G, Der CJ and Kay R. (1996). J. Biol. Chem., 271, 18643 ± 18650. Zheng Y, Olson MF, Hall Allmenn, Cerione RA and Toksoz D. (1995). J. Biol. Chem., 270, 9031 ± 9034. Zheng Y, Cerione R and Bender A. (1994). J. Biol. Chem., 269, 2369 ± 2372.