Regulation of ras p21 by GTPase activating proteins

We propose a model for dual effector functions of the known ras GAPs p120-GAP and NF1-GAP. This model is consistent with known biological and biochemical effects of GAPs in mammalian cells, but it is clearly not a proven hypothesis, and several difficulties remain in making this model convincing. One is the apparent difference between mammalian cells and yeasts, in which GAPs do not have a demonstrable effector function. The other is the difficulty of eliminating the possibility that other effectors exist that do not have GAP activity and do not bind ras p21 sufficiently tightly to allow detection through physical association. We hope that further analysis of GAP function will clarify the roles of these proteins, allowing at least a partial description of ras action in normal and malignant mammalian cells.

Purification of a plasma membrane-associated GTPase-activating protein specific for rap1/Krev-1 from HL60 cells

rap1/Krev-1 is a p21ras-related GTP-binding protein that has been implicated in the reversion of the ras-transformed cell phenotype. We have identified a GTPase-activating protein (GAP) specific for rap in plasma membranes isolated from differentiated HL60 cells. The rap GAP activity remained quantitatively associated with the membrane following washes with buffered 1 M LiCl containing 20 mM EDTA but was solubilized with the detergents Nonidet P-40 and deoxycholate. On the basis of size-exclusion chromatography, the membrane-associated rap GAP (rap GAPm) appeared distinct from the rap GAP detected in the cytosolic fraction from HL60 cells. The molecular sizes of the membrane and cytosolic forms were estimated to be 36 and 54 A, respectively. rap GAPm was solubilized and purified to near homogeneity by successive column chromatographies in the presence of detergent. The rap GAPm activity corresponded to a single polypeptide that migrated with a molecular mass of approximately 88 kDa on SDS/polyacrylamide gels. The purified rap GAPm was inactive toward the GTP-bound forms of p21ras, rho, G25K, and rac-1 and did not stimulate dissociation of guanine nucleotide from rap.

Purification, characterization, and western blot analysis of human GTPase-activating protein from native and recombinant sources

Human ras GTPase-activating protein (GAP) is a cytoplasmic factor that stimulates the GTPase activity of normal N-ras p21 while having no stimulatory effect on the GTPase activity of oncogenic variants of N-ras p21. We have purified two forms of native ras GAP from human placental tissue. In addition to the Mr = 120,000 type I GAP reported previously (1), an equivalent amount of an Mr = 95,000 molecule with GAP activity was recovered and shown to have the N-terminal sequence expected for type II GAP. The two GAP forms in placental extracts were resolved by molecular sieve chromatography and appeared to have a monomeric native structure. Human recombinant type I GAP was produced intracellularly in Sf9 insect cells using a baculovirus expression vector, and 10-mg quantities were purified to homogeneity in three steps. Comparison of the purified native and recombinant GAP molecules revealed that all three displayed similar biological specific activities in an in vitro GAP assay. A polyclonal antibody to purified recombinant GAP was prepared and shown to neutralize the activity of both native and recombinant GAPs. The antibody was also highly specific for the detection of native GAP by Western blot. Type I and II GAP species were detected in approximately equal amounts in cytoplasmic extracts of human placenta, but only type I GAP was observed when other human tissues were examined.

Clinical characteristics of acromegalic patients whose pituitary tumors contain mutant Gs protein

Activating mutations in the gene for the alpha-chain of Gs, the stimulatory regulator of adenylyl cyclase, have been identified in human GH-secreting pituitary tumors. Using the polymerase chain reaction and allele-specific oligonucleotide hybridization, we screened 25 GH-secreting tumors for the presence of the activating mutations. We also reviewed the clinical charts of the patients from whom the tumors were removed. Of 25 tumors, 10 (40%) contained activating mutations. Patients in the mutation-positive group came to surgery with smaller tumors and had lower GH levels. The activating mutations identify a subgroup of GH-secreting pituitary tumors that probably arise from a shared oncogenic mechanism.

The GTPase superfamily: a conserved switch for diverse cell functions

Proteins that bind and hydrolyse GTP are being discovered at a rapidly increasing rate. Each of these many GTPases acts as a molecular switch whose 'on' and 'off' states are triggered by binding and hydrolysis of GTP. Conserved structure and mechanism in myriad versions of the switch--in bacteria, yeast, flies and vertebrates--suggest that all derive from a single primordial protein, repeatedly modified in the course of evolution to perform a dazzling variety of functions.

Molecular cloning and expression of a G25K cDNA, the human homolog of the yeast cell cycle gene CDC42

G25K is a low-molecular-mass GTP-binding protein with a broad distribution in mammalian tissues. A cDNA clone was isolated by using oligonucleotides corresponding to the partial amino acid sequence of purified human G25K. The cDNA encodes an 191-amino-acid polypeptide containing GTP-binding consensus sequences and a putative farnesylation site at the C terminus. The sequence exhibits 50 and 70% identities to the mammalian rho and rac proteins, respectively, and an 80% identity to the Saccharomyces cerevisiae CDC42 gene product. Insect Sf9 cells infected with recombinant baculovirus vectors expressing the G25K cDNA produced a 25-kDa protein that bound GTP and was recognized by antibodies specifically reactive to G25K. G25K appears to be the human homolog of the CDC42 gene product, since expression of the G25K cDNA in S. cerevisiae suppressed both cdc42-1 and cdc24-4 temperature-sensitive lethal mutations.

Oncogenes in multiple myeloma: point mutation of N-ras

Alterations of ras, c-myc and bcl-1 have been described in hematologic malignancies of lymphoid origin. We investigated the structure of these genes and evaluated the frequency of point mutations involving H-, K- or N-ras in bone marrow samples from patients with multiple myeloma. No abnormalities were detected in the c-myc and bcl-1 genes, but two of 17 patients were found to have N-ras mutations by differential oligonucleotide hybridization and dideoxynucleotide sequencing following amplification by polymerase chain reaction. Bone marrow DNA from both patients had identical missense mutations of N-ras codon 61 changing CAA to AAA, resulting in a substitution of lysine for glutamine in the encoded protein. Multiple myeloma is the first mature B cell neoplasm found to harbor ras mutations.

Suppression of c-ras transformation by GTPase-activating protein

The ras genes are required for normal cell growth and mediate transformation by oncogenes encoding protein tyrosine kinases. Normal ras can transform cells in vitro and in vivo, but mutationally activated ras does so much more efficiently, and highly transforming mutant versions of ras have been isolated from a variety of human and animal tumours. The ras genes encode membrane-associated, guanine nucleotide-binding proteins that are active when GTP is bound and inactive when GDP is bound. The slow intrinsic GTPase activity of normal mammalian Ras proteins can be greatly accelerated by the GTPase-activating protein (GAP), which is predominantly cytoplasmic. This activity of GAP, which can increase with cell density in contact-inhibited cells, suggests that it functions as a negative, upstream regulator of ras. Other studies, however, show that GAP interacts with a region of ras-encoded protein implicated in ras effector function, which raises the possibility that GAP might also be a downstream target of ras. Mutationally activated ras-encoded proteins also interact with GAP, although they are resistant to its catalytic activity. In an attempt to define the role of GAP in ras-mediated transformation, we examined the effects on transformation of normal or mutant ras when cells overexpress GAP. We found that GAP suppresses transformation of NIH 3T3 cells by normal Ha-ras (c-ras) but does not inhibit transformation by activated Ha-ras (v-ras). These results support the hypothesis that GAP functions as a negative regulator of normal ras and make it unlikely that GAP alone is the ras target.

Inhibition of GTPase activating protein stimulation of Ras-p21 GTPase by the Krev-1 gene product

Krev-1 is known to suppress transformation by ras. However, the mechanism of the suppression is unclear. The protein product of Krev-1, Rap1A-p21, is identical to Ras-p21 proteins in the region where interaction with guanosine triphosphatase (GTPase) activating protein (GAP) is believed to occur. Therefore, the ability of GAP to interact with Rap1A-p21 was tested. Rap1A-p21 was not activated by GAP but bound tightly to GAP and was an effective competitive inhibitor of GAP-mediated Ras-GTPase activity. Binding of GAP to Rap1A-p21 was strictly guanosine triphosphate (GTP)-dependent. The ability of Rap1A-p21 to bind tightly to GAP may account for Krev-1 suppression of transformation by ras. This may occur by preventing interaction of GAP with Ras-p21 or with other cellular proteins necessary for GAP-mediated Ras GTPase activity.

Biochemical characterization of baculovirus-expressed rap1A/Krev-1 and its regulation by GTPase-activating proteins

Normal human rap1A and 35A rap1A (which encodes a protein with a Thr-35----Ala mutation) were cloned into a baculovirus transfer vector and expressed in Sf9 insect cells. The resulting proteins were purified, and their nucleotide binding, GTPase activities, and responsiveness to GTPase-activating proteins (GAPs) were characterized and compared with those of Rap1 purified from human neutrophils. Recombinant wild-type Rap1A bound GTP gamma S, GTP, and GDP with affinities similar to those observed for neutrophil Rap1 protein. The rate of exchange of GTP by Rap1 without Mg2+ was much slower than that by Ras. The basal GTPase activities by both recombinant proteins were lower than that observed with the neutrophil Rap1, but the GTPase activity of the neutrophil and wild-type recombinant Rap1 proteins could be stimulated to similar levels by Rap-GAP activity in neutrophil cytosol. In contrast to wild-type Rap1A, the GTPase activity of 35A Rap was unresponsive to Rap-GAP stimulation. Neither recombinant Rap1A nor neutrophil Rap1 protein GTPase activity could be stimulated by recombinant Ras-GAP at a concentration 25-fold higher than that required to hydrolyze 50% of H-Ras-bound GTP under similar conditions. These results suggest that the putative effector domains (amino acids 32 to 40) shared between Rap1 and Ras are functionally similar and interact with their respective GAPs. However, although Rap1 and Ras are identical in this region, secondary structure or additional regions must confer the ability to respond to GAPs.

ras p21 and GAP inhibit coupling of muscarinic receptors to atrial K+ channels

The signal-transducing G protein Gk couples muscarinic receptors to K+ (K+[ACh]) channels in atrial cells. Recombinant human ras p21 GAP (GTPase activating protein) at subnanomolar concentrations inhibited GTP-dependent channel opening in isolated atrial cell membranes. This inhibition depended on interaction of GAP with ras p21 in the isolated membranes. In addition, recombinant ras p21 proteins blocked the currents; this effect could be blocked by prior incubation of membranes with specific anti-GAP antibodies. We therefore propose that ras p21 GTP complexed with GAP (ras p21-GAP) blocks K+[ACh] currents. The channel block could be overcome by GTP gamma S activation of endogenous Gk; this indicates that ras p21-GAP does not interfere with interaction of Gk with the K+[ACh] channel directly, but prevents coupling of the muscarinic receptor to endogenous Gk.

PDGF beta-receptor stimulates tyrosine phosphorylation of GAP and association of GAP with a signaling complex

Platelet-derived growth factor (PDGF) stimulated the tyrosine phosphorylation of the GTPase activating protein (GAP) in 3T3 cells and in CHO cells expressing wild-type PDGF receptors, but not in several CHO cell lines expressing mutant receptors defective in transmitting mitogenic signals. Following PDGF treatment of cells, GAP physically associated with the PDGF receptor and with Raf-1, phospholipase c-gamma, and PI-3 kinase, suggesting that PDGF induced the formation of complexes of signaling molecules. The association of GAP with the PDGF receptor and the phosphorylation of GAP with the PDGF receptor and the phosphorylation of GAP were reconstituted in vitro using purified protein and in insect cells expressing murine PDGF receptor and human GAP. However, in cells transformed by activated c-Ha-ras, which are defective in certain responses to PDGF, GAP failed to associate with the PDGF receptor or increase its phosphotyrosine content in response to PDGF. The association of GAP with ligand-activated PDGF receptors may directly link PDGF and ras signaling pathways.

Retrospective analysis of ras gene activation in myeloid leukemic cells

Ras genes are activated by point mutations at critical sites of their coding regions. Activated N-ras genes with transforming ability have been detected in patients with myelodysplastic syndromes (MDS), acute myelogenous leukemia (AML) and in human myeloid cell lines. We used polymerase chain reaction (PCR), differential oligonucleotide hybridization and direct DNA sequencing to retrospectively analyze the N-ras gene of blast cells from the same patient (a) at time of diagnosis of MDS, (b) after the patient had developed AML. Two types of archival tissue samples served as a source of cells. Different passages of the KG-1 myeloid cell line which had been established from leukemic blasts of this patient were also analyzed. We found that native blast cells isolated at either of the two disease stages did not carry an N-ras mutation, and neither did early passage KG-1 cells. However, direct DNA sequencing of PCR-amplified DNA from nude mice transformants induced by DNA from late passage of the KG-1 cell line revealed two linked mutations involving both the second nucleotide of codon 12 and the third nucleotide of codon 15 of N-ras. The nucleotide substitution at codon 15 did not result in an amino acid substitution (silent mutation). The mutations probably occurred during prolonged passaging of the KG-1 cells and might have been overlooked by oligonucleotide hybridization assay.

RAS mutations in patients following cytotoxic therapy for lymphoma

The occurrence of Myelodysplastic Syndrome (MDS) and Acute Myeloblastic Leukaemia (AML) following cytotoxic therapy for neoplastic disease is well recognised. RAS mutations are common in patients with MDS and AML. To determine whether these lesions are found as early markers of secondary disease, we have studied the incidence of RAS mutations in the peripheral blood of 70 patients in complete remission from lymphoma. Patients were treated by standard chemotherapy regimes and/or localised radiotherapy. Treatment had been given 6 months to 14 1/2 years previously and no patient showed any sign of residual disease. Genomic DNA from peripheral blood leukocytes was amplified in vitro at target codons of N, K and H RAS genes, and mutations detected by hybridisation with oligonucleotide probes. RAS mutations were detected in 9 subjects. One patient with an N12 valine (Val) substitution had been in complete remission from Hodgkin's disease (HD) for 9 years. DNA from this patient registered in a nude mouse tumorigenicity assay (NMT). The N12 Val mutation was not detected in the original tumour tissue from the same patient. A second patient in remission from HD showed evidence of co-existent N12 cysteine (Cys) and N13 valine (Val) substitutions which were not detected in presentation material or unaffected tissues. All patients are currently haematologically normal, indicating that clones of mutant RAS bearing cells may be detected prior to any overt sign of disease.

Phosphorylation of GAP and GAP-associated proteins by transforming and mitogenic tyrosine kinases

The critical pathways through which protein-tyrosine kinases induce cellular proliferation and malignant transformation are not well defined. As microinjection of antibodies against p21ras can block the biological effects of both normal and oncogenic tyrosine kinases, it is likely that they require functional p21ras to transmit their mitogenic signals. No biochemical link has been established, however, between tyrosine kinases and p21ras. We have identified a non-catalytic domain of cytoplasmic tyrosine kinases, SH2, that regulates the activity and specificity of the kinase domain. The presence of two adjacent SH2 domains in the p21ras GTPase-activating protein (GAP) indicates that GAP might interact directly with tyrosine kinases. Here we show that GAP, and two co-precipitating proteins of relative molecular masses 62,000 and 190,000 (p62 and p190) are phosphorylated on tyrosine in cells that have been transformed by cytoplasmic and receptor-like tyrosine kinases. The phosphorylation of these polypeptides correlates with transformation in cells expressing inducible forms of the v-src or v-fps encoded tyrosine kinases. Furthermore, GAP, p62 and p190 are also rapidly phosphorylated on tyrosine in fibroblasts stimulated with epidermal growth factor. Our results suggest a mechanism by which tyrosine kinases might modify p21ras function, and implicate GAP and its associated proteins as targets of both oncoproteins and normal growth factor receptors with tyrosine kinase activity. These data support the idea that SH2 sequences direct the interactions of cytoplasmic proteins involved in signal transduction.

Genetic analysis of mammalian GAP expressed in yeast

We have designed a vector to express the mammalian GAP protein in the yeast S. cerevisiae. When expressed in yeast, GAP inhibits the function of the human H-rasgly12 protein, but not that of the H-rasval12 protein, and complements the loss of IRA1. IRA1 is a yeast gene that encodes a protein with homology to GAP and acts upstream of RAS. Mammalian GAP can therefore function in yeast and interact with yeast RAS. Because expression of GAP complements ira1-mutants, we propose that GAP shares some biochemical functions with IRA1. Other studies indicate that IRA1 controls the level of RAS activity, presumably by regulating GTP hydrolysis. By analogy, we propose that GAP may play a similar role.

Production of and responsiveness to transforming growth factor-beta in normal and oncogene-transformed human mammary epithelial cells

Transforming growth factors-beta (TGF beta) are a family of closely related, ubiquitously expressed growth factors with the common properties of induction of growth inhibition and expression of differentiation-related markers in epithelial cells. We investigated the role of TGF beta 1 in growth regulation of normal human mammary epithelial cells and in benzo(a)pyrene immortalized sublines further transformed by oncogenes in retroviral vectors. The normal cells were markedly growth inhibited by TGF beta 1, produced TGF beta in a latent form, and expressed TGF beta receptors. In the immortalized cells, both TGF beta-induced growth inhibition and TGF beta receptor binding were reduced. With the single oncogenes v-Ha-ras, v-mos, and SV40 T, growth sensitivity to TGF beta 1 increased, but TGF beta production or TGF beta receptor expression was not altered. Transformation to full malignancy by both SV40 T and v-Ha-ras led to escape from growth inhibition by TGF beta under anchorage-independent, but not anchorage-dependent, conditions without affecting TGF beta production or receptor characteristics. Thus, modulation of TGF beta growth responsiveness in these normal and oncogene transformed human mammary epithelial cells apparently occurs at a level distal to TGF beta receptor binding and is not solely correlated to expression of transforming oncogenes. Further, modulation of TGF beta production is not an indicator of malignant transformation in this system.

The polymerase chain reaction and cancer diagnosis

The introduction of polymerase chain reaction (PCR) technology, and its continual refinement, have had an enormous impact on many areas of molecular biology, including cancer research. This review examines several types of PCR assays that have been designed to investigate the molecular basis of cancer, in particular, those assays that have potential diagnostic applications.

Vimentin rather than keratin expression in some hormone-independent breast cancer cell lines and in oncogene-transformed mammary epithelial cells

To characterize differences in gene expression between hormone-dependent and hormone-independent mammary carcinoma, we cloned complementary DNAs of genes expressed in a hormone-independent breast carcinoma cell line that were not expressed in a hormone-dependent line. One clone, which was isolated in many copies, coded for the intermediate filament protein vimentin. A complementary DNA clone 1.8 kilobases long included the entire protein-coding region for vimentin. Vimentin was expressed by more than one-half of the hormone-independent breast carcinoma cell lines tested but not by the hormone-dependent cell lines. The cell lines which expressed vimentin expressed only low levels of cytokeratins. The correlation between vimentin expression and more advanced stages of mammary cell transformation was tested in a model system in which immortal, nontumorigenic human mammary epithelial cells or derivative lines transformed with v-ras-H or SV40 T-antigen were found not to express vimentin, whereas a derivative highly tumorigenic cell line transformed by both v-ras-H and T-antigen did express vimentin. Analysis of several other kinds of epithelial carcinoma cell lines showed only rare examples of vimentin expression.