MOLECULAR AND CELLULAR BIOLOGY, JUlY 1992, p. 3138-3148 0270-7306/92/073138-11$02.00/0

Vol. 12, No. 7

Growth-Regulated Expression of rhoG, a New Member of the ras Homolog Gene Family SYLVIE VINCENT, PHILIPPE JEANTEUR, AND PHILIPPE FORT*

URA CNRS 1191 "Genetique Moleculaire, " Universite Montpellier II Sciences et Techniques du Languedoc, Place E. Bataillon CPIOI, F-34095 Montpellier Cedex 5, France Received 8 October 1991/Accepted 17 April 1992

Cellular transition from the resting state to DNA synthesis involves master switches genes encoding transcriptional factors (e.g.,fos,jun, and egr genes), whose targets remain to be fully characterized. To isolate coding sequences specifically accumulated in late Gl, a differential screening was performed on a cDNA library prepared from hamster lung fibroblasts stimulated for 5 h with serum. One of the positive clones which displayed a sevenfold induction, turned out to code for a protein sharing homology to Ras-like products. Cloning and sequence analysis of the human homolog revealed that this putative new small GTPase, referred to as rhoG, is more closely related to the rac, CDC42, and TC10 members of the rho (ras homolog) gene family and might have diverged very early during evolution. rhoG mRNA accumulates in proportion to the mitogenic strength of various purified growth factors used for the stimulation, as a consequence of transcriptional activation. Gl-specific RNA accumulation is impaired upon addition of antimitogenic cyclic AMP and is enhanced when protein synthesis is inhibited, mainly as a result of RNA stabilization. rhoG mRNA expression is observed in a wide variety of human organs but reaches a particularly high level in lung and placental tissues.

Regulation of cell proliferation is a complex process involving coordinate expression of a limited number of genes. In particular, the trigger from the resting state to proliferation requires a rapid transduction of the mitogenic signal through biochemical events involving cascades of gene activations. Previous works have led to the identification of a set of immediate-early RNA species which rapidly and transiently accumulate upon growth stimulation (2, 26, 37, 38) and of some genes specific to growth-arrested cells, whose RNAs are accumulated during serum starvation and rapidly degraded when proliferation resumes (17, 59). Some encoded products, such as Fos- and Jun-related proteins, turned out to be transcriptional factors whose targets are far from being fully characterized. However, the tumorigenic potential of these switch genes stresses the important function of the target gene products in cell proliferation. Other early-induced genes were shown to code for cytoskeletal and extracellular matrix proteins, transmembrane proteins, and cytokines (reviewed in reference 31). Recently, two mRNAs encoding Ras-related proteins were shown to be mitogen inducible: rhoB, rapidly and transiently induced in growthstimulated rat fibroblasts (33), and rac2, slowly and stably induced in phytohemagglutinin A-stimulated human peripheral blood lymphocytes (56). Proteins that belong to the Ras superfamily share sequence identities with ras proto-oncogene products, especially with GTP-binding and hydrolysis domains, and can be divided into four major groups on the basis of amino acid sequence (10, 11): (i) products of the ras subgroup, that includes ras proto-oncogenes (H-ras, K-ras, and N-ras) (23, 39) and ral and rap genes (16, 52), (ii) products of the rab or YPT subgroup (62), (iii) the rho gene products (41, 53), and (iv) TC4 (22). These proteins are ubiquitously expressed in mammals (5, 9, 49), except Rab-3 (brain specific [49]) and Rac2 (restricted to the hemopoietic lineage [56]). Although their precise function remains unclear, some of them have *

been shown to participate in cell proliferation and differentiation (H-Ras, K-Ras, and N-Ras), vesicular transport (YPT1 [4] and Rab-4 [64]), cell polarity and cytoskeleton integrity (Cdc42 [35]), scaffolding of actin microfilaments (RhoA [50] and RhoC [14]), and in vitro NADPH oxidase activation (Racl [1]). We prepared a cDNA library from hamster CCL39 fibroblasts treated for 5 h with serum and isolated a set of late-induced genes, among which an mRNA encoding a new small GTPase was characterized. Phylogenetic analysis including hamster and human sequences indicates that this new GTPase belongs to the rho gene family and suggests that it might have diverged early during evolution. RNA accumulation is in proportion with the strength of the mitogen used, as a result of transcriptional activation. Tissular distribution of the corresponding mRNA is positively correlated with the presence of blood vessels, including smooth muscle cells from which the CCL39 hamster fibroblastic cell line has been derived. MATERIALS AND METHODS Cell culture. CCL39 and HeLa cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum (FCS). Confluent cells were made quiescent by serum deprivation in Dulbecco's modified Eagle's medium for 20 h. Renewed growth was stimulated by addition of 10% FCS, ao-thrombin (1 U. ml-'), or fibroblast growth factor (FGF; 35 ng- ml-1) in combination with insulin (10 ,ug* ml-'). Concentrations were 100 nM for tumor promoter agent (TPA) and 1 mM for 8-bromo-cyclic AMP (8-bromo-cAMP). Protein synthesis and RNA transcription were inhibited by cycloheximide (CHX; 10 jig. ml-') and actinomycin D (ActD; 5 p.g. ml-'), respectively. RNA extraction and analysis. Rodent CCL39 and human HeLa cell RNA was extracted as previously described (26). Poly(A)+ RNA was purified by oligo(dT)-cellulose (Sigma) chromatography. Total RNA (10 ,ug) was fractionated on 2

Corresponding author. 3138

VOL. 12, 1992

GROWTH-REGULATED mRNA ENCODING A NEW 21-kDa G PROTEIN

M formaldehyde-containing 1% agarose gels, transferred, and bound to nylon membranes (Hybond N+; Amersham) as described by the supplier. Filters were hybridized for 12 to 24 h in a mixture containing 50% formamide, 5 x Denhardt's solution, 10 mM P04 buffer, 0.75 M NaCl, 0.1% sodium dodecyl sulfate (SDS), 10% dextran sulfate, and 100 ,ug of denatured salmon sperm DNA ml-' at 42°C by using 2p_ labeled nick-translated probes (106 cpm/ml). Filters were washed twice in 2x SSC-0.1% SDS (lx SSC is 0.15 M NaCl plus 0.015 M sodium citrate) at room temperature and once in 0.2x SSC-0.1% SDS at 65°C. Construction of hamster and human cDNA libraries. The cDNA cloning procedure was derived from a primer-adapted method (12a). Poly(A)+ RNA (0.5 ,ug) was primed with 600 ng of oligo(dT) linker, and first-strand synthesis was performed with 200 U of Moloney murine leukemia virus reverse transcriptase (Bethesda Research Laboratories). High-molecular-weight products were recovered after 0.1% cetyltrimethyl ammonium bromide precipitation (6). RNA was hydrolyzed in 0.3 N NaOH for 30 min at 65°C, and DNA was size fractionated on an S400 filtration column (Pharmacia LKB). Large-size cDNAs (>0.5 kb) were then dG tailed and annealed to dC-tailed, BamHI-cut pT3T718U phagemid DNA (Pharmacia LKB) (cDNA-vector ratio, 1:2 [wt/wt]) and to a fivefold excess (wt/wt) of adapter. DNA was ligated for 16 h at 16°C and then repaired with T4 DNA polymerase (Bethesda Research Laboratories) at 37°C for 30 min. After ethanol precipitation, DNA was electroporated (Bio-Rad) into TG1 bacteria. The yield was approximately 105 recombinant clones per ng of cDNA. Total libraries were grown for 5 h at 37°C in Luria-Bertani medium supplemented with 40 ,ug of ampicillin ml-' and then frozen in 30% glycerol at -800C. Differential screening procedure. The library derived from hamster cells was plated at low density (four to five colonies per cm2) on nylon membranes (Hybond N; Amersham) and screened by differential hybridization using a high-specificactivity 32P-labeled cDNA probe (1 x 109 to 2 x 109 dpm p.g-1) prepared from poly(A)+ RNA of quiescent or serum-stimulated cells. Only one round of hybridization was performed, using an activity of 2 x 106 to 3 to 106 dpm. ml-'. Autoradiograms were digitized with an imageprocessing workstation (Visage System; Millipore). BioImage two-dimensional-analysis software was used for spot recognition, signal quantitation, and automatic spot comparison. Differentially expressed clones were then isolated and grown overnight at 37°C. Phagemid DNA was then extracted and used to probe Northern (RNA) analysis membranes. Run-on analysis. Preparation of nuclei and elongation of nascent transcripts were done as described by Greenberg and Ziff (28). Dot spotting of DNA onto nitrocellulose and hybridization and washing conditions were as previously described (51). Before autoradiography, filters were RNase A treated in 2x SSC for 15 min at 25°C. Screening of HeLa cDNA library. A fraction of the library (40,000 clones) derived from exponentially growing HeLa cells was plated at low density (20 colonies per cm2) on nylon membranes (Hybond N; Amersham) and prehybridized for 10 h in a mixture containing 50% formamide, 5 x SSC, 0.75 M NaCl, 5 x Denhardt's solution, 10 mM phosphate buffer (pH 7.0), 200 ,ug of sonicated denatured salmon sperm DNA ml-1, and 1% SDS. Hybridization was performed in the same buffer with 4, 105 dpm of heat-denatured heterologous hamster 32P-labelled insert ml-1 (6 108 dpm/,ug). Filters were washed twice for 30 min in 2x SSC-0.1% SDS at room temperature and autoradiographied for 5 h. Phagemid DNA

3139

was extracted from pools of colonies isolated from the positive areas, restricted with BamHI and HindlIl restriction enzymes, size fractionated on a 1% agarose gel, and transferred onto nylon N+ membranes. Filters were hybridized with the hamster probe under the same conditions as described above, and plasmids exhibiting a positive insert of full-length size copy were chosen. The corresponding cultures were then plated at a lower density, and individual positive clones were selected after a second round of screening. Sequence determination and characterization of cDNA clones. Individual clones were grown in 2-ml overnight cultures, and plasmid DNA was prepared. Sequence determination was performed on crude double-stranded plasmid DNA by using the T7 sequencing kit (Pharmacia). The complete sequences of PlAl cDNA and its human counterpart were obtained on both strands from subclones that were generated by various internal deletions and from the use of specific primers. Data base searches and sequence analyses were worked out by using CITI2 facilities (19) and home-

made softwares. Phylogenetic analyses. Phylogenetic analyses were performed with the TREEALIGN program for nucleic and protein sequence alignment, distance computation, and construction of phylogenetic trees (30). The last was also performed by another method based on parsimony and maximum-likelihood procedures (PHYLIP package) (25). Nucleotide sequence accession numbers. The EMBL accession numbers for hamster and human rhoG cDNA sequences are X61588 and X61587, respectively. RESULTS Construction of CCL39 cDNA library and screening for late serum-stimulated RNAs. Previous works led to the isolation of a set of immediately and transiently expressed mRNAs upon growth stimulation (2, 36, 37). To increase the frequency of unstable sequences, libraries were derived from CHX-treated cells, thereby preventing the emergence of a second set of serum-induced RNAs, whose expression would have required translational products of the first set. In our case, total RNA was extracted from the fibroblastic CCL39 cell line, serum starved for 20 h and then refed for 5 h with only FCS. An oriented cDNA cloning strategy was used to construct in a phagemid vector a library containing approximately 5 105 recombinant clones. We then performed a screening procedure on a limited fraction of the library to determine the magnitude of the delayed response. To avoid false-positive clones, we sequentially probed a unique filter with single-stranded 32Plabeled cDNA derived from cells stimulated for 5 h (plus probe), then with cDNA derived from resting cells (minus probe), and finally with a 32P-labeled plasmidic vector to get a precise determination of colony position. Autoradiographic data were digitized, and images were processed for spot detection and quantification by using a bidimensional analysis software (Biolmage; Millipore). An automatic comparison was then performed to sort out spots exhibiting variations in signal intensity. A total of 2,000 clones were screened, among which about 1% displayed over- or underexpression between the stimulated and resting states. Identification of a serum-induced RNA encoding a small GTPase. To check the efficiency of the whole screening procedure, we arbitrarily chose a clone, referred to as PlAl, that displayed a signal sevenfold more intense with the FCS-stimulated cDNA probe than with the unstimulated

VINCENT ET AL.

3140

MOL. CELL. BIOL.

rhoA 28S z 2

PlAl

*i v#

w44

C

1.3kb

0.7 kb

racli

c) 3

sss@ P2G4 0.~ ~ ....

Q 1 3 5 8 10

-- m

__2___12_16_20_Q C 0 2 8 12 16 20 0

4

2Ti

z2

PlAl

f Er

Er

g

C Q 1 3 5

1- i-l

-'

8 10

Ha-ras

_

z 2

22 I:bru

C Q 1 3 5 8 10

C Q 1 3 5

8 10

FIG. 1. Kinetics of accumulation of PlAl mRNA during FCS stimulation of CCL39 resting cells. For the gel, 10 ,ug of total RNA from exponentially growing cells (lane C), resting cells (lanes Q), or cells serum stimulated for the indicated times (in hours) was electrophoresed and transferred to nylon membranes. Filters were sequentially probed for PlAl and internal-control P2G4 (see text). The position of 28S rRNA was determined before transfer by UV shadowing. Sizes of the hybridized transcripts are indicated. For the histograms, cells were serum stimulated for the indicated times (in hours) and RNA was analyzed as described above. Signals corresponding to PlAl, rhoA, racl, and Ha-ras were quantified by using a Biolmage Sun workstation (Millipore), corrected according to P2G4 variations, and plotted as relative mRNA levels. Comparable signals were obtained after an overnight exposure for rhoA and PlAl and 1-week exposure for Ha-ras and racl. one. This clone contained a 1.3-kb insert, which was purified and used to probe a Northern blot containing equal amounts of total RNA isolated from CCL39 cells stimulated to grow for various times with FCS. As shown in Fig. 1, a 1.4-kb RNA species that accumulates transiently, exhibiting a three- to fourfold accumulation 8 to 12 h after FCS addition and then returning to the prestimulation level after 20 h, is revealed. To monitor the FCS effect and RNA amounts, blots were probed with murine c-fos cDNA (65a), and glyceraldehyde-3-phosphate dehydrogenase (66) probes (not shown) and with the insert of the P2G4 clone isolated from the hamster library for its constant level of expression between the resting and proliferative states. Partial DNA sequencing followed by a search in the GenBank and EMBL data bases revealed a significant similarity with various ras-like nucleic sequences belonging to the rho family. To compare the pattern of PlAl mRNA expression with those of other ras-like RNAs, we probed the Northern blots for racl, rac2, rhoA, and Ha-ras. Autoradiograms were quantified with a Visage (Millipore) workstation, and results are plotted in Fig. 1. rac2 expression was undetectable in our cell system, while Ha-ras and racl RNA (2.4 and 1.1 kb), expressed at very low levels (1 week of exposure time was needed), displayed a twofold induction after 1 to 3 h of FCS treatment. rhoA was expressed as two RNA species that both exhibited a level similar to that of PlAl RNA and were twofold induced within the first hour of stimulation.

The DNA sequence of PlAl insert was then completely determined for both strands, revealing the presence of an open reading frame of 191 amino acids (aa) long. Nucleic and protein sequences are presented in Fig. 2. Each of the five canonical ras sequences (10, 11) is present (aa 5 to 20 [G-1], aa 32 to 39 [G-2], aa 53 to 62 [G-3], aa 111 to 118 [G-4], and aa 156 to 160 [G-5]), as is the signal Cys-aliphatic-aliphatic-X for posttraductional modifications signalling membrane anchorage (aa 188 to 191, CysIleLeuLeu). In addition, other amino acids (Phe-28, Arg-68, Tyr-72, and Pro-109) conserved in most ras-related genes (58) are found within the sequence. In vitro transcription and translation of the clone yielded a single 21,000-Da peptide (not shown), in agreement with sequencing data. All of these features led to the conclusion that this protein is a genuine low-molecular-mass GTPase. Characterization of the homologous human cDNA clone from HeLa cells. As no ras homolog sequence had hitherto been reported in the hamster, one could not rule out that PlAl would be the hamster counterpart to an already known human GTPase. We then constructed and screened 40,000 clones from a HeLa cDNA library with the PlAl insert. Two positive clones harboring inserts 1.3 and 0.6 kb long were selected. The complete nucleotide sequence of the longest insert, as well as the translated protein sequence, is shown in Fig. 2. Both nucleic sequences share 85% identity (91% for the coding region), while a single Ser-Gly amino acid substitution was observed in the deduced protein sequences. Therefore, PlAl and its human counterpart mRNAs encode a new protein sequence that exhibits all the characteristics of a low-molecular-weight G protein. Although not proved, both hamster and human cDNAs (1,269 and 1,284 nucleotides [nt], respectively) are likely to be full length, as the corresponding RNAs displayed an apparent mobility on denaturing agarose gels strictly identical to that of glyceraldehyde-3-phosphate dehydrogenase RNA, whose full-length cDNA is 1,266 bp (66). 5' and 3' noncoding region lengths are, respectively, 135 nt (hamster) and 137 nt (human) and 558 nt (hamster) and 574 nt (human). Both nucleic sequences exhibit no particular feature except an excess in dC content (32%), in particular in the first 60 bases, in which several dCpdG's are observed (8). Sequence comparison and phylogenetic analysis. First, we performed a general comparative protein analysis among the

different Ras-like families, that unambiguously branched the new sequence within the Ras homolog (Rho) family. For this reason, it will be referred to as RhoG (ras homolog growth related). We then investigated more precisely the relationships between rhoG and the other rho members by analyzing two sets of sequences. In the first set, we examined seven human nucleic coding sequences of the rho subfamily (rhoG, CDC42Hs, rhoA, rhoB, rhoC, racl, and rac2) and one external coding sequence from the ras subfamily (ralA [16]). In the second set, we analyzed 11 protein sequences, including eight human Rho members (RhoA, RhoB, and RhoC [15, 69]; Racl and Rac2 [20]; TC10 [22]; CDC42Hs [24, 47, 61]; and RhoG [this paper]) and yeast CDC42 (34) and RHOl and RHO2 (42). To avoid topological abnormalities due to homoplasy, we excluded from the comparison aa 1 to 4 and 176 to 191, as recommended by Valencia et al. (63). Sequences were first aligned with the TREEALIGN package (30). Parameters were carefully chosen to minimize the number of deletions and insertions and, in the case of nucleic sequences, to allow only deletions or additions that did not disrupt the reading frame. The alignment of complete protein

VOL. 12, 1992

GROWTH-REGULATED mRNA ENCODING A NEW 21-kDa G PROTEIN

3141

50

GCTTCTCGAGCCCGGAGCCGCTGCCGCCGCCCCCAGCTCCCCCGCCTCG*GGAG*GGGCACCAGGTCACTGCAGCCAGAGGGGTC *-T------A-------C----------------------AG---C--G-A-----------T---A-TT 100

.

.

.

CAGAAGAGAGAGGAGGCACTGCCTC*ACTACAGCAACTGCACCCACG ATG CAG --------G----------------C----C--G--AT--A---A--- --x QQ GGT GAT GGG GCT GTG

T--A---

--

150

AGC ATC AAG TGC GTG GTG GTG --- --- --- --T--- --- ---

I

x

C

V

V

V

200 GGC AAG ACG TGC CTG CTC ATC TGC TAC ACA ACT AAC GCT TTC CCC AAA --A --- --T --- --- --- --T --G --- --T -- --- --- --G X T C L L I C T F P X 0 Y T A N

--- --- --- --- --- --- --O

D

0

A

V

250

GAG TAC ATC CCC ACC GTG TTC GAC AAT TAC AGC GCG CAG AGC GCA GTT GAC GGG CGC ACA GTG --A --- --- --- --T --- --- --- --- --- --- --C --- --T --- --- --- --A --- --- --I I Y P T V F D Y 8 A Q A V D R T V 0

300

AAC CTG AAC CTG TGG GAC ACT GCG GGC CAG GAG GAG TAT GAC CGC CTC CGT ACA CTC TCC TAC --- --- --- --- --- --- --- --- --- --- --A --A --- --- --- --- --C --C --T --- --N N L L W D T A 0 Q Y D R L R T L Y

350

400

CCT CAG ACC AAC GTT TTC GTC ATC TGT TTC TCC ATT GCC AGT CCG CCG TCC TAT GAG AAC GTG --- --- --- --T --- --T --- --- --- --- --- --- --- --- --A --G --- --- --A --T --N Q P T V F V I C a I A s P P s Y B N V

450

CGG CAC AAG TGG CAT CCA GAG GTG TGC CAC CAC TGC CCT GAT GTG CCC ATC CTG CTG GTG GGC A-- --- --- --- --- --- --- --- --T --- --- --- --A --- --- --T --C --- --A --X W R P 8 V C C P D V P I L L V 0 ---

500 ACC AAG AAG GAC CTG AGA GCC CAG CCT GAC ACC CTA CGG CGC CTC AAG GAG CAG AGC CAG GCG --- --- --- --- --- --- -- T

T

X

D

R

L

A

--- --Q P

550

--T --D

T

--- -- A

L

R

--- --- --- --R L X

---

Q

G-T --A --Q 8 A

.

CCC ATC ACA CCG CAG CAG GGC CAG GCA CTC GCG AAA CAG ATC CAC GCT GTG CGC TAC CTC GAA --- --- --- --- --- --- --- --- --- --G --C --G --- --T --T --- --- --- --- --- --G R 0 P I T P Q Q Q A L A X Q I A V Y L

650

TGC TCA GCC CTG CAA CAG GAT GGT GTC AAG GAA GTG TTC GCC GAG GCT GTC CGG GCT GTG CTC --- --A --- --G- --C --C --- --- --G --- --T --T --- --- --- --- --A --- --0 Q C A L Q D V V F A X B A V R A V L 700 AAC CCC ACG CCG ATC AAG CGT GGG CGG TCC TGC ATC CTC TTG TGA --- --A --A --- --A --- --- --- --- --- --- --- --- --- --X R 0 R N I P T P I C L L

CCCTGGCACTTGGCTTGGAGGC ----------CA--------C-

750

800

11 ------I---------------

---

--

---

---

------------I

-A-----TGT*------ t ---------------- C--------T--T-C---------------------------------A .

850

TTCCAGGGG**TTCCCTGAATGCCTTTTTCTCTGAGTCCTTTTTCTCCTTAAGGAGGCCTGCAGAGAAAGGGGCTTTGGGCT -G-------GAC----------------AT*********ttt*t*t-------------ACA---G----------------C 900

.

.

.

.

950

CTGCCC*CTCTGGCTTG*GGAACA*C***TGGGTA*TTCTCATGAGCTCATCCAAGCCAAGGTTGGACCCCTCCCCAAGAGGC -AC---A----------A----T-T-CCA-A--C-C---*--A---G---C-T--------**-C[---- TT---t* 1000

CAACCCAGTGCCCCCTCCCATTTTCCGCTACTGACCAGTTCATCCAGCTTTCCACACAGTTGTTGCTGCCTATTGTGGTGCCG

-----------TG---T------C-T------------------------------A-------------------------T 1100

CCTC**AGGTTAGGGGCTCTCAGCCATCTCTAACCTCTGCCCTCGCTGCTCTTGGAATTGCGCCCCCAAGATGCTCTCTCCCT ----TC------A--------*********--------CTG-*-TT----------G-A-GCT-T-------AT--------1200 1150 . . . . TCTCCAATGAGG*GAGCCACAGAATCCTGAGAAGGTGAATGTACCCTAACCTGCTCCTCTGTGCCTAGGCCTTACGCATTTGC C-C----G-G--*--------A--C---------AAA----**-----------C--A*------C------G-TA-G-C---

1250

TGACTGACTC*AGCCCCCATGCTTCTGGGGACCTTTCCTACCCCC*ATCAGCATCAATAAAACCTCCTGTCTCC ----------G------*------------G--------------C----------------------------

FIG. 2. Sequence comparison of human (top line) and hamster (bottom line) rhoG full-length cDNAs and proteins. Sequences were worked out by using the CITI2 facilities (19). Identical residues are indicated by dashes. Deletions in nucleotide sequences are shown by asterisks.

sequences

is shown in Fig. 3A. Multiple sequence align-

ments were then used to compute identity and distance

between members of each pair (Table 1). These statistical properties were processed to trace the evolution of nucleic and protein sequences, presented in Fig. 3B. The phylogenetic tree confirms that the rho subfamily consists of two exclusive groups, one comprising yeast RHOJ and RHO2 genes and mammalian rhoA, rhoB, and rhoC genes scores

and the other including yeast CDC42 and human CDC42Hs, racl, rac2, and TC10, to which the new rhoG sequence belongs. To corroborate these results, we performed a second round of branching by using a parsimony analysis (25), which gave a strictly identical topology. An interesting observation is that despite the high identity score between RhoG and Rac (Racl and Rac2), these proteins are not topologically close to each other, which suggests that both

3142

MOL. CELL. BIOL.

VINCENT ET AL.

A Gl *

RHOG

RAC1 RAC2 CDC42Hs CDC42Sc TCl1 RHOA RHOB RHOC YRHOl YRHO2 RALA

*

*

*****

RAC1 RAC2 CDC42Hs CDC42Sc TCl0 RHOA RHOB RHOC YRHO2 YRHOl RALA

*

****

G3 **

*

*

******

*

**

***

**

*

*

**

MPGAGRSSMAHGPGALMLKCVVVGDGAVGKTCLLMSYANDAFPEEYVPTVFDHYAVSVTVGGKQYLLGLYDTAGQEDYDRLRPLSYPMTDVFLICFSVVNPASFQNVKEEW MA-----AIR-------KKLVIVGDGACGKTCLLIVFSKDQFPEVYVPTVFENYVADIEVDGKQVELALWDTAGQEDYDRLRPLSYPDTDVILMCFSIDSPDSLENIPEKW MA-----AIR----KKLVVVGDGACGKTCLLIVFSKDEFPEVYVPTVFENYVADIEVDGKQVELALWDTAGQEDYDRLRPLSYPDTDVILMCFSVDSPDSLENIPEKW MA-----AIR---KKLVIVGDGACGKTCLLIVFSKDQFPEVYVPTVFENYIADIEVDGKQVELALWDTAGQEDYDRLRPLSYPDTDVILMCFSIDSPDSLENIPEKW MSQQVGNSIR-------KLVIVGDGACGKTCLLIVFSKGQFPEVYVPTVFENYVADVEVDGRRVELALWDTAGQEDYDRLRPLSYPDSNVVLICFSIDLPDSLENVQEKW MSEK---AVR-------RKLVIIGDGACGKTSLLYVFTLGQFPEQYHPTVFENYVTDCRVDGIKVSLTLWDTAGQEEYERLRPFSYSKADIILIGFAVDNFESLINARTKW

MAAN---KPKGQNSLALHKVIMVGSGGVGKSALTLQFMYDEFVEDYEPTKADSYRKKVVLDGEEVQIDILDTAGQEDYAAIRDNYFRSGEGFLCVFSITEMESFAATADFR G5

_

CAAX

--------------NPTPI-KRGRSCILL HPEVCHH-CPDVPILLVGTKKDLRAQPDTLRRLKEQSQAPITPQQGQALAKQIHAVAYLECSALQQDGVKEVFAEAVRAVLYPEVRHH-CPNTPIILVGTKLDLRDDKDTIEKLKEKKLTPITYPQGLAMAKEIGAVKYLECSALTQRGLKTVFDEAIRAVL---------------CPPPVKKRKRKCLLL FPEVRHH-CPSTPIILVGTKLDLRDDKDTIEKLKEKKLAPITYPQGLALAKEIDSVKYLECSALTQRGLKTVFDEAIRAVL---------------CPQPTRQQKRACSLL VPEITHH-CPKTPFLLVGTQIDLRDDPSTIEKLAKNKQKPITPETAEKLARDLKAVKYVECSALTQRGLKNVFDEAILAAL----------------EPPETQPKRKCCIF FPEVHHH-CPGVPCLVVGTQIDLRDDKVIIEKLQRQRLRPITSEQGSRLARELKAVKYVECSALTQRGLKNVFDEAIVAAL----------------EPPVIKKSKKCTIL VPELKEY-APNVPFLLIGTQIDLRDDPKTLARLNDMKEKPICVEQGQKLAKEIGACCYVECSALTQKGLKTVFDEAIIAIL------TPKKHTVKKRIGS--RCINCCLIT TPEVKHF-CPNVPIILVGNKKDLRNDEHTRRELAKMKQEPVRPEEGRDMANRIGAFGYMECSAKTKDGVREVFEMATRAAL-------------QAR--- RGKKKSGCLVL VPEVKHF-CPNVPIILVANKKDLRSDEHVRTELARMKQEPVRTDDGRAMAVRIQAYDYLECSAKTKEGVREVFETATRAAL-------------QKRYGSQNGCINCCKVL TPEVKHF-CPNVPIILVGNKKDLRQDEHTRRELAKMKQEPVRSEEGRDMANRISAFGYLECSAKTKEGVREVFEMATRAGL-------------QVR--- KNKRRRGCPIL ADEALRY-CPDAPIVLVGLKKDLRQEAHFKE-NATDEMVPI--EDAKQVARAIGAKKYMECSALTGEGVDDVFEVATRTSL--------LMKKEPGA-----NCCIIL IAEVLHF-CQGVPIILVGCKVDLRNDPQTIEQLRQEGQQPVTSQEGQSVADQIGATGYYECSAKTGYGVREVFEAATRASL---MGKSKT-NGKAKKNTT-EKKKKKCVLL EQILRVKEDENVPFLLVGNKSDLEDK------------RQVSVEEAKNRAEQWN-VNYVETSAKTRANVDKVFFDLMREIRARKMEDSKEKNGKKKRKSLAKRIRERCCIL

FIG. 3. (A) Sequence comparison of Rho proteins. Protein sealigned by using the TREEALIGN software (30). Residues identical for all Rho proteins indicated by asterisks. Gl to G5 indicate Ras canonical boxes, and CAAX represents the signal for COOH processing. Deletions are indicated by dashes. YRHOl

3 RAC 1

6

B

**

**

MQS---------IKCVVVGDGAVGKTCLLICYTTNAFPKEYIPTVFDNYSAQSAVDGRTVNLNLWDTAGQEEYDRLRTLSYPQTNVFVICFSIASPPSYENVRHKW MQA---------IKCVVVGDGAVGKTCLLISYTTNAFPGEYIPTVFDNYSANVMVDGKPVNLGLWDTAGQEDYDRLRPLSYPQTDVFLICFSLVSPASFENVRAKW MQA----------IKCVVVGDGAVGKTCLLISYTTNAFPGEYIPTVFDNYSANVMVDSKPVNLGLWDTAGQEDYDRLRPLSYPQTDVFLICFSLVSPASYENVRAKW MQT---------IKCVVVGDGAVGKTCLLISYTTNKFPSEYVPTVFDNYAVTVMIGGEPYTLGLFDTAGQEDYDRLRPLSYPQTDVFLVCFSVVSPSSFENVKEKW MQT-----------LKCVVVGDGAVGKTCLLISYTTNQFPADYVPTVFDNYAVTVMIGDEPYTLGLFDTAGQEDYDRLRPLSYPSTDVFLVCFSVISPPSFENVKEKW

G4 RHOG

G2 ***

quences were

are

jlRAC2

1

4

6=

2

I-C

CDC42Hs

4

2

SCDC42Sc

16

and

19

YRHO2, yeast

RHOl

and

RHO2. (B)

Evolution

of

Rho

pro-

sequence alignment in panel A was used to elaborate an evolutionary tree by using the TREEALIGN (30) or PHYLIP (25) program. Numbers represent the computed branch length (in perteins. The

TClo

RHOG 26

RHo2

cent

divergence).

7

RHO 1

13

for

RHO

8

responses

to

various

mitogenic

and

nonmitogenic

agents. DNA synthesis can be induced in resting CCL39 cells by either ot-thrombin or FGF treatment, the effect of the

C

9

their

latter being enhanced by the addition of insulin. These

RHO B

the Rac and RhoG proteins have been subjected to a lower evolution rate than the other members of the family. Mitogenic induction of the rhoG mRNA level. To get information on the biochemical pathways involved in increased rhoG mRNA expression, we examined resting cells

factors follow distinct biochemical pathways for transducing their mitogenic trigger, either through receptor tyrosine kinase (FGF) or G-protein-coupled receptor (cx-thrombin) (43, 44, 54, 55). Total RNA, prepared from resting CCL39 cells stimulated for 3, 6, and 9 h with various mitogens, was analyzed by Northern blotting. Quantitation of the autoradiograms is plotted Fig. 4. ot-Thrombin (right panel) led to a

TABLE 1. Identity scores and computed distances among Rho proteins % Identity and proximity index with':

Protein RhoG

RhoG Racl Rac2 CDC42 TC10 RhoB RhoC RHOl RHO2

Racl

Rac2

CDC42

TC10

RhoB

RhoC

RHOl

RHO2

73

74

62 73 72

72 68 58 62 64 56

58 67 65 70

57 60 58 54 54

50

58 60 58 54 54 88

59

95

46 48 46 44 43 55 56 56

71 72 62 58

73 69

50

59

54 54 46

63 58 48

95

72 54 54 55

54 54 49 42

88 70 52

74 56

58 57 53

51 70 74 56

a Proteins were aligned, and the pairwise percent identity (upper triangle) and pairwise proximity index (lower triangle) were computed by using the TREEALIGN package (30).

VOL. 12, 1992

GROWTH-REGULATED mRNA ENCODING A NEW 21-kDa G PROTEIN

3143

0~~~~~~~~0 cN

0 03 > 3-

90 636

Time

Time~~~~~~~~~~~~~~~~~~~~~~~...

FIG. 4. Effect of purified growth factors on rhoG mRNA accumulation. Total RNA was extracted from cells stimulated for the indicated... times (in hours) with 100 nM TPA, 10 pg of insulin mV', 35 ngof~~~~~~... of FGF mP1, insulin plus FGF, 1 mM 8-bromo-cAMP (cAMP), 1 U ct-thrombin mP' (Thr), or 8-bromo-cAMP plus a-thrombin. RNA was analyzed by Northern blotting, and autoradiographic signals were~~~~~~.... quantified by using a Visage Sun workstation (Millipore).~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.......

large increase in rhoG mRNA (six to nine times, depending on the experiments) 9 h after stimulation. In contrast, stimulation with FGF (left panel) led to a twofold induction 6 to 9 h poststimulation, while no significant stimulation was observed with insulin. Combination of FGF and insulin led to a three- to fourfold induction, indicating a cooperative effect of both factors. However, even in the latter case, the

range of induction still remained lower than in cells stimulated with FCS. As FGF stimulation acts through protein kinase C (PKC) activation, we then analyzed the effect of TPA, which directly activates PKC (13). As expected, TPA appears as effective as FGF (2.5-fold induction after 9 h), suggesting that rhoG induction by FGF is mediated by PKC. The various agents used for the precedent stimulations

A

B -28S-

c-fos- X " s

c-fos

-2.2 -

-18S-28S-

rhoG -

-18S-

-

-1.4 -

P2G4 -

W

a o wo

-0.7 -

30' lh 3h 6h 8h FCS + CHX

lOh

I

0

rhoG

P2G4

3h 6h 6A 6B CHX

FIG. 5. Effect of protein synthesis inhibition on rhoG mRNA accumulation. (A) Total RNA was extracted from cells stimulated for the indicated times in the presence of 10% FCS and 10 1Lg of CHX ml-l. Northern blots were sequentially probed for rhoG, c-fos, and P2G4 (see text). (B) Total RNA was extracted from resting cells (lane 0), resting cells treated with CHX for 3 or 6 h, and resting cells treated for 1 (lane 6A) or 2 (lane 6B) h with 10% FCS and then for 5 (lane 6A) or 4 (lane 6B) h with 10% FCS and CHX.

3144

MOL. CELL. BIOL.

VINCENT ET AL.

RhoG 3

m

10

1

..

-

0

2

: m

7IHa 1HW nnfl JIL

4

JJL

0 2 4 6 10 24 0 3 6 0 3 6 FCS+CHX CHX FCS

Time (hrs) 0 2 4 6 10 24 0 3 6 0 3 6 FCS FCS+CHX CHX *

3.

la

4-

4-

y-actin

a) u

3-

3.

to

*6

2-

cu

1 V-

A:4

04 Time (hrs)

0* 0

2 4 6 10 24 0 3 6 0 3 6 FCS FCS+CHX CHX

0 2 4 6 10 24 0 3 6 0 3 6 FCS FCS+CHX CHX

FIG. 6. rhoG transcriptional activity in serum-stimulated cells. Cells were treated with FCS, CHX, or both for the indicated times. 32P-labeled nuclear run-on transcripts were hybridized to slot blots of plasmidic DNA containing various inserts. Filters were washed, RNase treated, and autoradiographed for 6 days. Signals were processed by using the Biolmage Whole Band analysis software, and integrated optical densities were plotted as transcriptional activities relative to those of unstimulated cells.

exhibit different mitogenic strengths: DNA synthesis is elicited in 10 to 20% of cells treated with FGF, in a slightly higher percentage of cells treated with FCS, and in 50 to 60% of cells treated with ao-thrombin (55). To ensure that rhoG mRNA accumulation is correlated to the mitogenic response induced by growth factors, we tested the antimitogenic effect of 8-bromo-cAMP on a-thrombin stimulation. As shown in Fig. 4 (right panel), accumulation of rhoG mRNA is 80% inhibited when 8-bromo-cAMP is added in conjunction with ao-thrombin, which strongly suggests that biochemical pathways involved in DNA synthesis are responsible for the increase in the rhoG RNA level. Effect of CHX on mRNA accumulation. A common property shared by early growth-regulated mRNAs is their ability to superaccumulate when protein synthesis is inhibited, mainly as a result of mRNA stabilization (2, 60, 67). We

examined the possibility that rhoG mRNA could behave in a similar way by Northern analysis of total RNA extracted from cells treated with combinations of FCS and CHX. As shown in Fig. 5A, a 10-h stimulation with FCS in the presence of CHX led to a 10-fold increase in rhoG mRNA level compared with that in resting cells. Nevertheless, CHX addition also provoked a 2- to 3-h lag period before the start of RNA accumulation (compare with Fig. 1). As an internal control, we reprobed the same blot for expression of c-fos, whose mRNA superaccumulated rapidly and then gradually disappeared, in agreement with what was previously reported (26, 55, 67). When cells were treated with CHX alone, only a slight increase in mRNA level was detected on overexposed films, even after a 6-h incubation (Fig. 5B), thereby indicating that serum stimulation is necessary for rhoG mRNA superaccumulation. The involvement of early

GROWTH-REGULATED mRNA ENCODING A NEW 21-kDa G PROTEIN

VOL. 12, 1992

3145

TABLE 2. rhoG mRNA t1,2 in resting and stimulated cells in the presence or absence of protein synthesisa Results with: ActD + CHX

ActD Cell treatment

Optical density at:

Optical density at:

Resting FCS 6h 12 h ae-Thrombin (12 h)

Oh

2h

4h

8h

0.42

0.34

0.19

ND

1.2 1.53 3.4

0.78 ND ND

0.47 0.33 0.82

ND 0.13 0.36

t,,2 (h)

4h

8h

t1/2(h)

3.5 ± 0.4

0.35

ND

15.2

2.9 ± 0.3 2.3 ± 0.5 2.6 ± 0.6

0.95 ND ND

ND 0.8 2.1

12 8.9 11.4

Resting cells were stimulated with FCS and with ca-thrombin. ActD (5 pLg ml-) or ActD plus 10 p,g of CHX ml-) were then added. Total RNA was extracted at various times after cell treatment, transferred to nylon membranes, and probed for rhoG and P2G4 (see text). Results were quantified by using a Visage Sun workstation (Millipore), and integrated optical densities were measured. ND, not determined.

expressed products in superaccumulation was tested by adding CHX 1 or 2 h after serum stimulation and then pursuing the incubation for 5 or 4 h, respectively (Fig. 5B, lanes 6A and 6B). Interestingly, the rhoG RNA level was increased when addition of CHX was delayed for 2 h rather than 1 h. This result is consistent with the lagging effect of CHX (see above) and suggests that protein synthesis, although not necessary, has a positive effect on rhoG gene activation during the first 2 h of serum stimulation. Regulation of rhoG mRNA expression. In order to determine the mechanisms involved in rhoG mRNA accumulation, transcriptional activity of the rhoG gene was analyzed by run-on experiments on nuclei isolated from cells treated for various times with FCS. In vitro 32P-labeled elongated RNA was hybridized to plasmidic DNA immobilized on nitrocellulose membranes, and signals were quantified by image analysis. As shown in Fig. 6, FCS addition elicits a transient increase in the transcriptional activity of rhoG, which exhibits a threefold induction after 4 h and then progressively returns to its unstimulated level after 24 h. This induction takes place 2 to 3 h before the increase in cytoplasmic RNA and can account for the range of mRNA accumulation (Fig. 1). Three other probes (isolated during the same screening procedure as rhoG and corresponding to thrombospondin, a-tubulin, and -y-actin) were also included in the experiment to monitor the extent of the stimulation (Fig. 6). No transcriptional activation is observed for the y-actin gene, while thrombospondin and a-tubulin display, respectively, eight- and threefold induction 2 h after serum addition, in agreement with published data (46, 48). We then tested the effect of CHX treatment on rhoG mRNA stability and gene transcription by ActD chase experiments and run-on analysis. Resting cells and cells stimulated with FCS or thrombin for 6 and 12 h were treated with combinations of ActD and CHX, and the level of rhoG mRNA at various times was estimated by Northern blotting experiments. Autoradiograms were quantitated, and results are presented in Table 2. rhoG mRNA exhibits a higher turnover rate in cells stimulated with FCS or thrombin for 12 h (half-life [t1l2] = 2.3-2.6 h) and 6 h (t412 = 2.9 h) than in resting cells (t1l2 = 3.5 h), and a combined treatment with CHX and ActD leads to a fourfold increase in mRNA stability in both the resting and stimulated states. rhoG transcriptional activity was measured by run-on analysis of cells treated with CHX alone or in combination with serum. The results presented in Fig. 6 led to several observations. First, transcription is slightly activated by FCS in the presence of CHX during the first 6 h, but it is not comparable

to the superinduction observed for c-fos (26) or thrombospondin genes. Second, an increase in transcription occurs later when FCS is added in combination with CHX, which is in agreement with the delayed accumulation of cytoplasmic RNA (Fig. 5A). These data therefore establish that rhoG transcriptional activation does not strictly require de novo protein synthesis and that RNA superaccumulation in the presence of CHX results from a stabilizing effect, as is the case for numerous primary response RNAs. rhoG mRNA expression in various human organs. Expression of several GTPases has been shown to be tissue specific (20, 22, 49). To determine whether rhoG mRNA is expressed in particular cell types, we probed a Northern membrane containing poly(A)+ RNA isolated from various human organs (Multiple Tissue Northern; Clontech). A single rhoG 1.4-kb mRNA species is present in RNAs from all organs but displays pronounced quantitative variations (Fig. 7): strong expression was found in placental and lung tissues and to a lesser extent in the heart. The other five organs displayed weaker expression in the following decreasing order: kidney, brain, skeletal muscle, liver, and pancreas. These variations do not reflect differences in amounts of loaded RNA, as a different pattern was obtained for glyceraldehyde-3-phosphate dehydrogenase mRNA, whose expression is particularly high in skeletal and cardiac muscles.

2.40 -

2.40

rhoG

-

1.35-

4W

1.35-

e-

GAPDH

M Li Lu P1 B H FIG. 7. Distribution of rhoG mRNA expression in human adult P

K

tissues. A multiple-tissue Northern blot (Clontech) containing 2 ,ug of poly(A)+ RNA per lane was probed for rhoG. Lane P, pancreas; lane K, kidney; lane M, skeletal muscle; lane Li, liver; lane Lu, lung; lane P1, placenta; lane B, brain; lane H, heart. Relevant RNA marker sizes (in kilobases) are indicated on the left. Autoradiography for the rhoG and glyceraldehyde-3-phosphate dehydrogenase probes was for 10 h at -80°C with an intensifying screen.

3146

VINCENT ET AL.

DISCUSSION To get a better understanding of the regulation of mammalian cell growth, several groups, including ours, have initiated an overall survey of mRNAs that accumulate lately upon mitogenic stimulation. These consist of primary-response RNA, which exhibits a low rate of accumulation (2, 32), and secondary-response RNA, whose accumulation requires de novo protein synthesis during the first hours following growth factor addition. In this study, we have isolated a subset of cDNA clones derived from CCL39 cells stimulated to grow for 5 h and identified one of these as originating from a gene encoding a protein which displayed all the features of a genuine and active small GTPase. First, in vitro transcription and translation produced a 21-kDa polypeptide. Then, the four domains involved in guanine nucleotide binding and hydrolysis are strongly similar to those of the Ras proteins, and finally, motives signalling posttranslational modifications for membrane anchorage are correctly located (10, 11). All of these features were also found in the protein sequence derived from the human homolog cDNA clone. Sequence comparison of hamster and human sequences with known Ras-like proteins revealed that this GTPase is a new member of the Rho family (referred to as RhoG). Although it shares substantial amino acid identity with Racl and Rac2, phylogenetic analysis strongly suggests that it is topologically distinct from Rac and has probably diverged first from the common stem shared by Rac, TC10, and CDC42. Two ras-like genes, rhoB (33) and rac2 (56), have been recently described as growth-regulated genes: rac2, which is the closest relative of rhoG, is specifically expressed in cells of the hemopoietic lineage and is strongly induced within 6 h in peripheral blood lymphocytes stimulated with phytohemagglutinin A and in T cells stimulated to grow. rhoB, which belongs to the sister subgroup of rhoG within the rho gene family and is an immediate-early response to epidermal growth factor, platelet-derived growth factor, and FCS and to the oncogenic v-Fps protein tyrosine kinase activation. Interestingly, this gene is not induced upon PKC activation or by signals transduced through other protein tyrosine kinases, such as insulin or FGF. We describe here a third case of a growth-regulated rho gene, rhoG, which belongs to the group of late primary response genes: its RNA accumulates gradually for 8 to 10 h after serum addition, as a result of transient transcriptional activation. Because of its moderate instability, rhoG cytoplasmic RNA then returns progressively to its prestimulation level 10 to 14 h later. CHX treatment induces a lag period before RNA transcription activation and elicits a superaccumulation due to an increase in RNA t1l2. However, the latter property, shared by many mRNAs involved in early response to growth factors (2) and by tubulin and histone mRNAs (reviewed in reference 12), is not specific enough to provide information on the nature and the location of the sequence(s) involved in rhoG mRNA degradation. In our cell system, rhoG displays some similarities to Ha-ras, as both genes are induced several hours after serum addition (40). However, if their regulations share qualitative properties, they differ quantitatively, as rhoG expression is at least 10 times higher than that of Ha-ras. Other rho genes, such as rhoA and rhoC, were previously shown to be uninducible in rat-2 and NRK cells (33). Our results show that rac2, the closest relative to rhoG, is not expressed in resting or stimulated CCL39 cells, while rhoA and racl are moderately induced but exhibit a pattern of induction distinct from that of rhoG.

MOL. CELL. BIOL.

An interesting correlation is observed between rhoG mRNA accumulation and the mitogenic strength of the purified factor used for the stimulation: ox-thrombin, which is the most potent mitogenic factor in CCL39 cells (55, 65), elicits an increase in the rhoG mRNA level much more important than that elicited by FGF alone or in combination with insulin. In addition, a poor induction is observed after TPA treatment, which activates PKC without any mitogenic effect in CCL39 cells. The correlation is further strengthened by the absence of a response to a-thrombin when the transduction pathway is inhibited by an increase in cAMP level (43, 45). Differences in the rhoG mRNA level in response to various growth factors may rely on the type of receptor and ligand involved in the stimulation and on the cell type on which the receptor is expressed. ao-Thrombin acts through receptors that are coupled through GTP-binding proteins to various effectors, including a phosphoinositide-specific (PIP2) phospholipase C (7, 54). FGF mitogenic signal is transduced through receptors with tyrosine kinase activity (68) but does not involve PIP2 breakdown and PKC activation in CCL39 cells, and elevation in the cAMP level inhibits PIP2 breakdown (45). Thus, rhoG induction by mitogens appears to be independent of PKC activation but rather correlates with PIP2 metabolism and more precisely with inositol-1,4,5-triphosphate accumulation. However, as PIP2 phospholipase C is only one of the effectors activated by ax-thrombin, additional data are necessary to determine whether inositol-1,4,5-triphosphate production is sufficient to promote rhoG gene activation in this cell system. What could be the physiological meaning of rhoG induction? Numerous primary-response genes have already been characterized, isolated from a wide range of cell systems (e.g., murine NIH 3T3, BALB/c, and Swiss 3T3 cells; BHK and hamster lung CCL39 cells; and rat PC12 nerve cells) (2, 18, 32, 48) subjected to different external stimuli (e.g., platelet-derived growth factor, epidermal growth factor, FCS, TPA, and nerve growth factor). Some of them appear to be ligand or cell specific, for instance, cytokines JE and KC, which are preferentially induced by platelet-derived growth factor (18), and thrombospondin, an extracellular matrix protein whose expression is restricted to cell types such as vascular cells or neurons (reviewed in reference 27). But the majority of them were found to be activated in a variety of cellular contexts, suggesting that cell and ligand specificities rely on changes in the relative amounts of multiple products rather than synthesis of specific ones (31). RhoG is likely to represent a cell-specific induced response for several reasons. First, its mRNA is expressed at different levels in various cell systems (strong expression in stimulated CCL39 cells and much lower expression in HeLa and LTk- cells [not shown]). Second, tissue distribution of rhoG mRNA reveals a high level of expression in lung, placenta, and to a lesser extent, heart tissue, all of them involving a high proportion of vascular cells. Interestingly, CCL39 cells, which were derived from smooth muscle cells, exhibit a high level of rhoG mRNA in response to ot-thrombin, a hormone directly involved in angiogenesis and wound healing. One can therefore speculate that this new GTPase is specific for vascular cells and participates in their proliferation by acting either directly on the pathway specifying DNA synthesis or indirectly during the setup of specialized functions. Both possibilities are consistent with what is known about this class of small GTPases. First, various reports have demonstrated their implication in the regulation of cell proliferation: (i) amplification of normal rhoA results in colonies that exhibit reduced dependence on serum for

VOL. 12, 1992

GROWTH-REGULATED mRNA ENCODING A NEW 21-kDa G PROTEIN

growth and that are tumorigenic in nude mice (3), (ii) the translocation breakpoint in Philadelphia chromosome-positive chronic myeloid leukemias prevents the expression of bcr, which encodes a GAP protein which down-regulates the active GTP-bound form of p2lracI (21), and (iii) the protooncogene dbl encodes a protein that promotes the catalysis of guanine nucleotide exchange on p21CDC42Hs (29). In addition, rhoB was found to be an immediate-early gene inducible by various growth factors (33), while the rac2 mRNA level is strongly increased in activated T cells (56). Second, these proteins are likely to be involved in cytoskeleton organization: (i) microinjection of recombinant p2lrho induces rapid changes in cell morphology (50), (ii) mutations in the yeast CDC42 gene affect cell polarity, either preventing the correct location of buds on the mother cell or impeding the budding process (34), and (iii) ADP ribosylation of RhoC in Vero cells results in the disappearance of microfilaments (14). All of these features suggest that rho proteins are implicated in the modifications of the overall cell structure required for progressing towards the S phase, perhaps by the docking of proteins involved in cytoskeletal reorganization or of extracellular matrix proteins controlling cell adhesion. Along this line, rhoG, which is closely related to rac and CDC42, could be under the control of the same regulatory proteins and implicated in similar pathways. In conclusion, a new RNA sequence encoding a G protein was found to be induced during the late G1 phase of renewed fibroblastic growth. According to our observations, two non-mutually-exclusive hypotheses can be formulated: (i) RhoG participates in late events that regulate the cellular growth of a limited set of cell types, in particular, vascular endothelial or smooth muscle cells, and (ii) it is implicated in the establishment of cell- or tissue-specific functions, such as docking of structural proteins or cytokines. We are currently investigating the subcellular location of the RhoG protein, as well as the effects of its deregulated expression by transfection and microinjection experiments. ACKNOWLEDGMENTS We thank D. Caput and A. Minty (ESBR, Labege) for valuable advice in cDNA cloning and screening procedures; C. Theillet, P. Boquet, and 0. Dorseuil for the gifts of human Ha-Ras, RhoA, and Rac2 plasmid DNAs; and A. Vie for help with tissue cultures. This work was supported by institutional grants from CNRS, INSERM, and the University of Montpellier II and contracts from the Association pour la Recherche contre le Cancer, the Caisse Regionale d'Assurance Maladie, the Ligue Nationale contre le Cancer, and the Foundation pour la Recherche M6dicale. REFERENCES 1. Abo, A., E. Pick, A. Hall, N. Totty, C. G. Teahan, and A. W. Segal. 1991. Activation of the NADPH oxidase involves the small GTP-binding protein p2lracl. Nature (London) 353:668-670. 2. Almendral, J. M., D. Sommer, J. Perera, J. Burckardt, H. MacDonald-Bravo, and R. Bravo. 1988. Complexity of early genetic response to growth factors in mouse fibroblasts. Mol. Cell. Biol. 8:2140-2148. 3. Avraham, H., and R. A. Weinberg. 1989. Characterization and expression of the human rhoH12 gene product. Mol. Cell. Biol. 9:2058-2066. 4. Bacon, R. A., A. Salminen, H. Ruhohala, P. Novick, and S. Ferro-Novick. 1989. The GTP-binding protein YPT1 is required for transport in vitro: the Golgi apparatus is defective in yptl mutants. J. Cell Biol. 109:1015-1022. 5. Barbacid, M. 1987. ras genes. Annu. Rev. Biochem. 56:779-828. 6. Belyavsky, A., T. Vinogradova, and K. Rajewsky. 1989. PCRbased cDNA library construction: general cDNA libraries at the level of a few cells. Nucleic Acids Res. 17:2919-2932.

3147

7. Berridge, M. J., and R. F. Irvine. 1984. Inositol trisphosphate, a novel second messenger in signal transduction. Nature (London) 312:315-321. 8. Bird, A. P. 1987. CpG-rich islands and the function of hemimethylation. Nature (London) 321:209-213. 9. Bos, J. L. 1988. The ras family and human carcinogenesis. Mutat. Res. 195:255-262. 10. Bourne, H. R., D. A. Sanders, and F. McCormiclk 1990. The GTPase superfamily: a conserved switch for diverse cell functions. Nature (London) 348:125-132. 11. Bourne, H. R., D. A. Sanders, and F. McCormick. 1991. The GTPase superfamily: conserved structure and molecular mechanism. Nature (London) 349:117-123. 12. Brawerman, G. 1989. mRNA decay: finding the right targets. Cell 57:9-10. 12a.Caput, D. Personal communication. 13. Castagna, M., Y. Takai, K. Kaibuchi, K. Sano, U. Kikkawa, and Y. Nishizuka. 1982. Direct activation of calcium activated phospholipid dependent protein kinase by tumor promoting phorbolesters. J. Biol. Chem. 257:7847. 14. Chardin, P., P. Boquet, P. Madaule, M. R. Popoff, E. J. Rubin, and D. M. Gill. 1989. The mammalian G protein rhoC is ADP-ribosylated by Clostridium botulinum exoenzyme C3 and affects actin microfilaments in Vero cells. EMBO J. 8:10871092. 15. Chardin, P., P. Madaule, and A. Tavitian. 1988. Coding sequences of human rho cDNAs clone 6 and clone 9. Nucleic Acids Res. 16:2717. 16. Chardin, P., and A. Tavitian. 1986. The ral gene: a new ras-related gene isolated by the use of a synthetic probe. EMBO J. 5:2203-2208. 17. Ciccarelli, C., L. Philipson, and V. Sorrentino. 1989. Regulation of expression of growth arrest-specific genes in mouse fibroblasts. Mol. Cell. Biol. 10:1525-1529. 18. Cochran, B. H., A. C. Reffel, and C. D. Stiles. 1983. Molecular cloning of gene sequences regulated by platelet-derived growth factor. Cell 33:939-947. 19. Dessen, P., C. Fondrat, C. Valencien, and C. Mugnier. 1990. BISANCE: a French service for access to biomolecular sequence databases. Cabios 6:355-356. 20. Didsbury, J., R. F. Weber, G. M. Bokoch, T. Evans, and R. Synderman. 1989. Rac, a novel Ras-related family of proteins that are botulinum toxin substrates. J. Biol. Chem. 264:1637816382. 21. Dieckmann, D., S. Brill, M. D. Garrett, N. Totty, J. Hsuan, C. Monfries, C. Hall, L. Lim, and A. Hall. 1991. Bcr encodes a GTPase-activating protein for p2lrac. Nature (London) 351:400402. 22. Drivas, G. T., A. Shih, E. Coutavas, M. G. Rush, and P. D'Eustachio. 1990. Characterization of four novel ras-like genes expressed in a human teratocarcinoma cell line. Mol. Cell. Biol. 10:1793-1798. 23. Ellis, R. W., D. De Feo, T. Y. Shih, M. A. Gonda, H. A. Young, N. Tsuchida, D. Lowy, and E. Scolnick. 1981. The p21 src genes of Harvey and Kirsten sarcoma viruses originate from divergent members of a family of normal vertebrate genes. Nature (London) 292:506-511. 24. Evans, T., M. L. Brown, E. D. Fraser, and J. K. Northup. 1986. Purification of the major GTP-binding proteins from human placental membranes. J. Biol. Chem. 261:7052-7059. 25. Felsenstein, J. 1981. Evolutionary trees from DNA sequences: a maximum likelihood approach. J. Mol. Evol. 17:368-376. 26. Fort, P., J. Rech, A. Vie, M. Piechaczyk, A. Bonnieu, P. Jeanteur, and J.-M. Blanchard. 1987. Regulation of c-fos gene expression in hamster fibroblasts: initiation and elongation of transcription and mRNA degradation. Nucleic Acids Res. 15: 5657-5667. 27. Frazier, W. A. 1991. Thromspondins. Curr. Opin. Cell Biol. 3:792-799. 28. Greenberg, M. E., and E. B. Ziff. 1984. Stimulation of 3T3 cells induces transcription of the c-fos proto-oncogene. Nature (London) 311:433-438. 29. Hart, M. J., A. Eva, T. Evans, S. A. Aaronson, and R. A.

3148

30. 31. 32.

33.

34. 35.

36.

37.

38. 39.

40.

41. 42. 43.

44.

45.

VINCENT ET AL.

Cerione. 1991. Catalysis of guanine nucleotide exchange on the CDC42Hs protein by the dbl oncogene product. Nature (London) 354:311-314. Hein, J. L. 1989. A unified approach to alignments and phylogenies. Methods Enzymol. 183:626-645. Herschmann, H. R. 1991. Primary response genes induced by growth factors and tumor promoters. Annu. Rev. Biochem. 60:281-319. Hirschhorn, R. R., P. Aller, Z. A. Yuan, C. W. Gibson, and R. Baserga. 1984. Cell-cycle-specific cDNAs from mammalian cells temperature sensitive for growth. Proc. Natl. Acad. Sci. USA 81:6004-6008. Jahner, D., and T. Hunter. 1991. The ras-related gene rhoB is an immediate-early gene inducible by v-Fps, epidermal growth factor, and platelet-derived growth factor in rat fibroblasts. Mol. Cell. Biol. 11:3682-3690. Johnson, D. I., and J. R. Pringle. 1990. Molecular characterization of CDC42, a Saccharomyces cerevisiae gene involved in the development of cell polarity. J. Cell Biol. 111:143-152. Kikuchi, A., T. Yamashita, M. Kawata, K. Yamamoto, K. Ikeda, T. Taminoto, and Y. Takai. 1988. Purification and characterization of a novel GTP-binding protein with a molecular weight of 24,000 from bovine brain membranes. J. Biol. Chem. 263:2897-2904. Lau, L. F., and D. Nathans. 1985. Identification of a set of genes expressed during the Go/Gl transition of cultured mouse cells. EMBO J. 4:3145-3151. Lau, L. F., and D. Nathans. 1987. Expression of a set of growth-related immediate early genes in BALB/c 3T3 cells: coordinate regulation with c-fos or c-myc. Proc. Natl. Acad. Sci. USA 84:1182-1186. Leonard, D. G., A. B. Ziff, and L. A. Greene. 1987. Identification and characterization of mRNAs regulated by nerve growth factor in PC12 cells. Mol. Cell. Biol. 7:3156-3167. Lowe, D. G., D. J. Capon, E. Delwart, A. Y. Sakaguchi, S. L. Naylor, and D. V. Goeddel. 1987. Structure of the human and murine R-ras genes, novel genes closely related to ras protooncogenes. Cell 48:137-146. Lu, K., R. A. Levine, and J. Campisi. 1989. c-ras-Ha gene expression is regulated by insulin or insulinlike growth factor and by epidermal growth factor in murine fibroblasts. Mol. Cell. Biol. 9:3411-3417. Madaule, P., and R. Axel. 1985. A novel ras-related gene family. Cell 41:31-40. Madaule, P., R. Axel, and A. Myers. 1987. Characterization of two members of the rho gene family from the yeast Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 84:779-783. Magnaldo, I., G. L'Allemain, J. C. Chambard, M. Moenner, D. Barritault, and J. Pouyssegur. 1986. The mitogenic signalling pathway of fibroblast growth factor is not mediated through polyphosphoinositide hydrolysis and protein kinase C activation in hamster fibroblasts. J. Biol. Chem. 261:16916-16922. Magnaldo, I., J. Pouyssegur, and S. Paris. 1988. Thrombin exerts a dual effect on stimulated adenylate cyclase in hamster fibroblasts: an induction via a GTP-binding protein and a potentiation via activation of protein kinase C. Biochem. J. 253:711-719. Magnaldo, I., J. Pouyssegur, and S. Paris. 1989. Cyclic AMP inhibits mitogen-induced DNA synthesis in hamster fibroblasts, regardless of the signalling pathway involved. FEBS Lett.

245:65-69. 46. Majack, R. A., J. Milbrandt, and V. M. Dixit. 1987. Induction of thrombospondin messenger RNA levels occurs as an immediate primary response to platelet-derived growth factor. J. Biol. Chem. 262:8821-8825. 47. Munemitsu, S., M. A. Innis, R. Clark, F. McCormick, A. Ullrich, and P. Polakis. 1990. Molecular cloning and expression of a G25K cDNA, the human homolog of the yeast cell cycle gene CDC42. Mol. Cell. Biol. 10:5977-5982. 48. Nikaido, T., D. W. Bradley, and A. Pardee. 1991. Molecular cloning of transcripts that accumulate during the late Gl phase in cultured mouse cells. Exp. Cell Res. 192:102-109. 49. Olofsson, B., P. Chardin, N. Touchot, A. Zarahoui, and A. Tavitian. 1988. Expression of the ras-related raLA, rhol2 and rab

MOL. CELL. BIOL.

genes in adult mouse tissues. Oncogene 3:231-234. 50. Paterson, H., A. J. Self, M. D. Garrett, I. Just, K. Aktories, and A. Hall. 1990. Microinjection of recombinant p2lrho induces rapid changes in cell morphology. J. Cell Biol. 111:1001-1007. 51. Piechaczyk, M., J. M. Blanchard, L. Marty, C. Dani, F. Panabieres, S. El Sabrouty, P. Fort, and P. Jeanteur. 1984. Glyceraldehyde-3-phosphate dehydrogenase gene expression is regulated at the post-transcriptional level in rat tissues. Nucleic Acid Res. 12:6951-6963. 52. Pizon, V., P. Chardin, I. Lerosey, B. Olofsson, and A. Tavitian. 1988. Human cDNAs rapl and rap2 homologous to the Drosophila gene Dras3 encode proteins closely related to Ras in the effector region. Oncogene 3:201-204. 53. Polakis, P. G., R. F. Weber, B. Nevins, J. R. Didsbury, T. Evans, and R. Snyderman. 1989. Identification of the ral and racl gene products, low molecular GTP-binding proteins from human platelets. J. Biol. Chem. 28:16383-16389. 54. Pouyssegur, J. 1990. G-proteins in growth factors action, p. 555-570. In L. Birnbaumer and R. Iyengar (ed.), G-proteins. Academic Press, Inc., New York. 55. Pouyssegur, J., C. Kahan, and K. Seuwen. 1990. Transmembrane signalling pathways initiating cell growth. Role of G-protein coupled receptors, p. 80-100. In S. Weber (ed.), Growth factors: from genes to clinical application. Raven Press, New York. 56. Reibel, L., 0. Dorseuil, R. Stancou, J. Bertoglio, and G. Gacon. 1991. A hemopoietic specific gene encoding a small GTP binding protein is overexpressed during T cell activation. Biochem. Biophys. Res. Commun. 175:451-458. 57. Ruther, U., E. F. Wagner, and R. Muller. 1985. Analysis of the differentiation-promoting potential of inducible c-fos genes introduced into embryonal carcinoma cells. EMBO J. 4:1775-1781. 58. Santos, E., and R. Nebreda. 1989. Structural and functional properties of Ras proteins. FASEB J. 3:2151-2163. 59. Schneider, C., R. M. King, and L. Philipson. 1988. Genes specifically expressed at growth arrest of mammalian cells. Cell 54:787-793. 60. Shaw, G., and R. Kamen. 1986. A conserved AU sequence from the 3' untranslated region of GM-CSF mRNA mediates selective mRNA degradation. Cell 46:659-667. 61. Shinjo, K., J. G. Koland, M. J. Hart, V. Narasimhan, D. I. Johnson, T. Evans, and R. A. Cerione. 1990. Molecular cloning of the gene for the human placental GTP-binding protein Gp (G25K): identification of this GTP-binding protein as the human homolog of the yeast cell-division cycle protein CDC42. Proc. Natl. Acad. Sci. USA 87:9853-9857. 62. Touchot, N., P. Chardin, and A. Tavitian. 1987. Four additional members of the ras gene superfamily isolated by oligonucleotide strategy: molecular cloning or YPT-related cDNAs from a rat brain library. Proc. Natl. Acad. Sci. USA 84:8210-8214. 63. Valencia, A., P. Chardin, A. Wittinghofer, and C. Sander. 1991. The Ras family: evolutionary tree and role of conserved amino acids. Biochemistry 19:4637-4648. 64. Van der Sluis, P., M. Hull, A. Zarahoui, A. Tavitian, B. Goud, and I. Mellman. 1991. The small GTP-binding protein Rab4 is associated with early endosomes. Proc. Natl. Acad. Sci. USA 88:6313-6317. 65. Van Obberghem-Schilling, E., J. C. Chambard, S. Paris, G. L'Allemain, and J. Pouyssegur. 1983. a-Thrombin-induced early mitogenic signalling events and Go to S-phase transition of fibroblasts require continuous external stimulation. EMBO J. 4:2927-2932. 65a.Vincent, S. Unpublished results. 66. Vincent, S., and P. Fort. 1990. Nucleotide sequence of hamster glyceraldehyde-3-phosphate dehydrogenase mRNA. Nucleic Acids Res. 10:3054. 67. Wilson, T., and R. Treisman. 1988. Removal of poly(A) and consequent degradation of c-fos mRNA facilitated by 3' AUrich sequences. Nature (London) 336:396-399. 68. Yarden, Y., and A. UlIrich. 1988. Growth factor receptor tyrosine kinases. Annu. Rev. Biochem. 57:453-478. 69. Yeramian, P., P. Chardin, P. Madaule, and A. Tavitian. 1987. Nucleotide sequence of human rho cDNA clone 12. Nucleic Acids Res. 15:189.