The kelch proteins Gpb1 and Gpb2 inhibit Ras activity via association with the yeast RasGAP neurofibromin homologs Ira1 and Ira2

The G protein-coupled receptor Gpr1 and associated Galpha subunit Gpa2 govern dimorphic transitions in response to extracellular nutrients by signaling coordinately with Ras to activate adenylyl cyclase in the yeast Saccharomyces cerevisiae. Gpa2 forms a protein complex with the kelch Gbeta mimic subunits Gpb1/2, and previous studies demonstrate that Gpb1/2 negatively control cAMP-PKA signaling via Gpa2 and an unknown second target. Here, we define these targets of Gpb1/2 as the yeast neurofibromin homologs Ira1 and Ira2, which function as GTPase activating proteins of Ras. Gpb1/2 bind to a conserved C-terminal domain of Ira1/2, and loss of Gpb1/2 results in a destabilization of Ira1 and Ira2, leading to elevated levels of Ras2-GTP and unbridled cAMP-PKA signaling. Because the Gpb1/2 binding domain on Ira1/2 is conserved in the human neurofibromin protein, an analogous signaling network may contribute to the neoplastic development of neurofibromatosis type 1.     

Trap RACK1 with Ras to mobilize Src signaling at syndecan-2/p120-GAP upon transformation with oncogenic ras

HiTrap-syndecan-2/p120-GAP and HiTrap-syndecan-2/RACK1 affinity columns were applied to reveal that Src tyrosine kinase was highly expressed in BALB/3T3 cells transfected with plasmids pcDNA3.1-[S-ras(Q(61)K)] of shrimp Penaeus japonicus. Both columns were effective to isolate Src tyrosine kinase. The selective molecular affinity for Src was found to be stronger with HiTrap-syndecan-2/RACK1, as revealed with competitive RACK1 to dislodge Src from HiTrap-syndecan-2/p120-GAP. We thus challenged the syndecan-2/p120-GAP and syndecan-2/RACK1 with GTP-K(B)-Ras(Q(61)K). The reaction between RACK1 and syndecan-2 was sustained in the presence of mutant Ras proteins, but not the reaction between p120-GAP and syndecan-2. In the presence of syndecan-2, GTP-K(B)-Ras(Q(61)K) exhibited sufficient reactivity with p120-GAP to discontinue the reaction between p120-GAP and syndecan-2. But the interference of mutant Ras disappeared when Src tyrosine kinase was introduced to stabilize the syndecan-2/p120-GAP complex. On the other hand, in the absence of syndecan-2, GTP-K(B)-Ras(Q(61)K) was found to react with RACK1. The reaction between GTP-K(B)-Ras(Q(61)K) and RACK1 could provide a mechanism to deprive RACK1 for the organization of syndecan-2/RACK1 complex and to facilitate the formation of syndecan-2/p120-GAP complex, as well as to provide docking sites for Src signaling upon transformation with oncogenic ras.     

Differential actions of p60c-Src and Lck kinases on the Ras regulators p120-GAP and GDP/GTP exchange factor CDC25Mm.

It is known that the human Ras GTPase activating protein (GAP) p120-GAP can be phosphorylated by different members of the Src kinase family and recently phosphorylation of the GDP/GTP exchange factor (GEF) CDC25Mm/GRF1 by proteins of the Src kinase family has been revealed in vivo [Kiyono, M., Kaziro, Y. & Satoh, T. (2000) J. Biol. Chem. 275, 5441-5446]. As it still remains unclear how these phosphorylations can influence the Ras pathway we have analyzed the ability of p60c-Src and Lck to phosphorylate these two Ras regulators and have compared the activity of the phosphorylated and unphosphorylated forms. Both kinases were found to phosphorylate full-length or truncated forms of GAP and GEF. The use of the catalytic domain of p60c-Src showed that its SH3/SH2 domains are not required for the interaction and the phosphorylation of both regulators. Remarkably, the phosphorylations by the two kinases were accompanied by different functional effects. The phosphorylation of p120-GAP by p60c-Src inhibited its ability to stimulate the Ha-Ras-GTPase activity, whereas phosphorylation by Lck did not display any effect. A different picture became evident with CDC25Mm; phosphorylation by Lck increased its capacity to stimulate the GDP/GTP exchange on Ha-Ras, whereas its phosphorylation by p60c-Src was ineffective. Our results suggest that phosphorylation by p60c-Src and Lck is a selective process that can modulate the activity of p120-GAP and CDC25Mm towards Ras proteins.     

Tfs1p, a member of the PEBP family, inhibits the Ira2p but not the Ira1p Ras GTPase-activating protein in Saccharomyces cerevisiae

Ras proteins are guanine nucleotide-binding proteins that are highly conserved among eukaryotes. They are involved in signal transduction pathways and are tightly regulated by two sets of antagonistic proteins: GTPase-activating proteins (GAPs) inhibit Ras proteins, whereas guanine exchange factors activate them. In this work, we describe Tfs1p, the first physiological inhibitor of a Ras GAP, Ira2p, in Saccharomyces cerevisiae. TFS1 is a multicopy suppressor of the cdc25-1 mutation in yeast and corresponds to the so-called Ic CPY cytoplasmic inhibitor. Moreover, Tfs1p belongs to the phosphatidylethanolamine-binding protein (PEBP) family, one member of which is RKIP, a kinase and serine protease inhibitor and a metastasis inhibitor in prostate cancer. In this work, the results of (i) a two-hybrid screen of a yeast genomic library, (ii) glutathione S-transferase pulldown experiments, (iii) multicopy suppressor tests of cdc25-1 mutants, and (iv) stress resistance tests to evaluate the activation level of Ras demonstrate that Tfs1p interacts with and inhibits Ira2p. We further show that the conserved ligand-binding pocket of Tfs1-the hallmark of the PEBP family-is important for its inhibitory activity.     

The C-terminal domains of human neurofibromin and its budding yeast homologs Ira1 and Ira2 regulate the metaphase to anaphase transition

The human tumor suppressor neurofibromin contains a cysteine and serine-rich domain/Ras-GTPase activating protein domain (CSRD/RasGAP) and a C-terminal domain (CTD). Domain studies of neurofibromin suggest it has other functions in addition to being a RasGAP, but the mechanisms underlying its tumor suppressor activity are not well understood. The budding yeast Saccharomyces cerevisiae is a good model system for studying neurofibromin function because it possesses Ira1 and Ira2, which are homologous to human neurofibromin in both sequence and function. We found that overexpression of CTD or a neurofibromin CTD-homologous domain (CHD) of Ira1/2 in budding yeast delayed degradation of the securin protein Pds1, whereas overexpression of CSRD/RasGAP did not affect Pds1 degradation. We also found that when CTD or CHD was overexpressed, the number of cells in metaphase was higher than in the control. These results demonstrate that CTD and CHD function in the metaphase to anaphase transition. In addition, Δira1Δira2 cells bypassed mitotic arrest in response to spindle damage, indicating that Ira1 and Ira2 may be involved in the spindle assembly checkpoint (SAC). However, Δira1Δira2Δmad2 cells are more sensitive to spindle damage than Δmad2 or Δira1Δira2 cells are, suggesting that Ira1/2 and Mad2 function in different pathways. Overexpression of CTD but not CSRD/RasGAP partially rescued the hypersensitivity of Δira1Δira2Δmad2 cells to microtubule-destabilizing drugs, indicating a role for CTD in the SAC pathway. Taken together, independently of RasGAP activity, the C-terminal domains of neurofibromin, Ira1, and Ira2 regulate the metaphase to anaphase transition in a Mad2-independent fashion.     

Aberrant Ras regulation and reduced p190 tyrosine phosphorylation in cells lacking p120-Gap.

The Ras guanine nucleotide-binding protein functions as a molecular switch in signalling downstream of protein-tyrosine kinases. Ras is activated by exchange of GDP for GTP and is turned off by hydrolysis of bound GTP to GDP. Ras itself has a low intrinsic GTPase activity that can be stimulated by GTPase-activating proteins (GAPs), including p120-Gap and neurofibromin. These GAPs possess a common catalytic domain but contain distinct regulatory elements that may couple different external signals to control of the Ras pathway. p120-Gap, for example, has two N-terminal SH2 domains that directly recognize phosphotyrosine motifs on activated growth factor receptors and cytoplasmic phosphoproteins. To analyze the role of p120-Gap in Ras regulation in vivo, we have used fibroblasts derived from mouse embryos with a null mutation in the gene for p120-Gap (Gap). Platelet-derived growth factor stimulation of Gap-/- cells led to an abnormally large increase in the level of Ras-GTP and in the duration of mitogen-activated protein (MAP) kinase activation compared with wild-type cells, suggesting that p120-Gap is specifically activated following growth factor stimulation. Induction of DNA synthesis in response to platelet-derived growth factor and morphological transformation by the v-src and EJ-ras oncogenes were not significantly affected by the absence of p120-Gap. However, we found that normal tyrosine phosphorylation of p190-rhoGap, a cytoplasmic protein that associates with the p120-Gap SH2 domains, was dependent on the presence of p120-Gap. Our results suggest that p120-Gap has specific functions in downregulating the Ras/MAP kinase pathway following growth factor stimulation, and in modulating the phosphorylation of p190-rhoGap, but is not required for mitogenic signalling.     

The membrane localization of Ras2p and the association between Cdc25p and Ras2-GTP are regulated by protein kinase A (PKA) in the yeast Saccharomyces cerevisiae

The Saccharomyces cerevisiae Ras2p has been suggested to be a target in the feedback regulation of Ras-cAMP pathway. This work proves that the Ras2p localization is regulated by PKA activity, and that PKA down-regulates Ras2p activity and the protein association between Cdc25p and Ras2-GTP, which is due to a reduced Ras2-GEF Cdc25p activity. These results suggest that Ras2p localization and Ras2-GEF activity of Cdc25p play roles in the feedback regulation of Ras2p in the Ras-cAMP pathway.     

Feedback regulation of Ras2 guanine nucleotide exchange factor (Ras2-GEF) activity of Cdc25p by Cdc25p phosphorylation in the yeast Saccharomyces cerevisiae

Ras-GEF Cdc25p has been found to be hyperphosphorylated upon glucose addition. This work provides evidence indicating that PKA activity positively regulates the degree of Cdc25p phosphorylation, and that the intracellular association of Cdc25p and Ras2p is independent of PKA activity. In vitro experiments revealed that the Ras2-GEF activity of Cdc25p is inhibited by Cdc25p phosphorylation. These data suggest a negative feedback mechanism by which intracellular cAMP synthesis is inhibited by PKA through Cdc25p phosphorylation.

Structured summary

MINT-8053016: CDC25p (uniprotkb:P04821) physically interacts (MI:0915) with ras2p (uniprotkb:P01120) by anti tag co-immunoprecipitation (MI:0007)MINT-8053030: ras2p (uniprotkb:P01120) physically interacts (MI:0915) with CDC25p (uniprotkb:P04821) by anti bait co-immunoprecipitation (MI:0006)     

High sensitivity of cultured human trophoblasts to ribosome-inactivating proteins.

Many plant proteins possessing abortifacient activities were identified as ribosome-inactivating proteins (RIPs). The effect of several ribosome-inactivating proteins (saporin 6, dianthin 32, pokeweed antiviral protein from seeds, gelonin, bryodin-R, and momordin) on primary cultures of human trophoblasts and human embryonal fibroblasts and on choriocarcinoma (JAR and BeWo) and ovarian carcinoma (TG) cell lines was studied. Protein synthesis of human trophoblasts and BeWo cells was lowered by RIPs more than that of other cells. The trophoblastic receptors for estradiol were not affected by treatment of the cells with momordin. The binding and uptake of saporin 6 and momordin by BeWo and HeLa cells were not correlated to cell toxicity.     

Generation of mice with a conditional allele of the p120 Ras GTPase-activating protein

p120 Ras GTPase-activating protein (RasGAP) encoded by the rasa1 gene in mice is a prototypical member of the RasGAP family of proteins involved in negative-regulation of the p21 Ras proto-oncogene. RasGAP has been implicated in signal transduction through a number of cell surface receptors. In humans, inactivating mutations in the coding region of the RASA1 gene cause capillary malformation arteriovenous malformation. In mice, generalized disruption of the rasa1 gene results in early embryonic lethality associated with defective vasculogenesis and increased apoptosis of neuronal cells. The early lethality in this mouse model precludes its use to further study the importance of RasGAP as a regulator of cell function. Therefore, to circumvent this problem, we have generated a conditional rasa1 knockout mouse. In this mouse, an exon that encodes a part of the RasGAP protein essential for catalytic activity has been flanked by loxP recognition sites. With the use of different constitutive and inducible Cre transgenic mouse lines, we show that deletion of this exon from the rasa1 locus results in effective loss of expression of catalytically-active RasGAP from a variety of adult tissues. The conditional rasa1 mouse will be useful for the analysis of the role of RasGAP in mature cell types.     

Tissue-specific functions of the Caenorhabditis elegans p120 Ras GTPase activating protein GAP-3

All metazoan genomes encode multiple RAS GTPase activating proteins (RasGAPs) that negatively regulate the conserved RAS/MAPK signaling pathway. In mammals, several RasGAPs exhibit tumor suppressor activity by preventing excess RAS signal transduction. We have identified gap-3 as the to date missing Caenorhabditiselegans member of the p120 RasGAP family. By studying the genetic interaction of gap-3 with the two previously identified RasGAPs gap-1 and gap-2, we find that different combinations of RasGAPs are used to repress LET-60 RAS signaling depending on the cellular context. GAP-3 is the predominant negative regulator of RAS during meiotic progression of the germ cells, while GAP-1 is the key inhibitor of RAS during vulval induction. In other tissues such as the sex myoblasts or the chemosensory neurons, all three RasGAPs act in concert. The C. elegans RasGAPs have thus undergone partial specialization after gene duplication to allow the differential regulation of the RAS/MAPK signaling pathway in different cell types. A similar tissue specialization of the human tumor suppressor genes may explain the strong bias in the type of cancer they promote when mutated.     

Oligomerization of the Mg2+-transport proteins Alr1p and Alr2p in yeast plasma membrane

Alr1p is an integral plasma membrane protein essential for uptake of Mg(2+) into yeast cells. Homologs of Alr1p are restricted to fungi and some protozoa. Alr1-type proteins are distant relatives of the mitochondrial and bacterial Mg(2+)-transport proteins, Mrs2p and CorA, respectively, with which they have two adjacent TM domains and a short Mg(2+) signature motif in common. The yeast genome encodes a close homolog of Alr1p, named Alr2p. Both proteins are shown here to be present in the plasma membrane. Alr2p contributes poorly to Mg(2+) uptake. Substitution of a single arginine with a glutamic acid residue in the loop connecting the two TM domains at the cell surface greatly improves its function. Both proteins are shown to form homo-oligomers as well as hetero-oligomers. Wild-type Alr2p and mutant Alr1 proteins can have dominant-negative effects on wild-type Alr1p activity, presumably through oligomerization of low-function with full-function proteins. Chemical cross-linking indicates the presence of Alr1 oligomers, and split-ubiquitin assays reveal Alr1p-Alr1p, Alr2p-Alr2p, and Alr1p-Alr2p interactions. These assays also show that both the N-terminus and C-terminus of Alr1p and Alr2p are exposed to the inner side of the plasma membrane.     

Identification of a lumenal sequence specifying the assembly of Emp24p into p24 complexes in the yeast secretory pathway

The p24 proteins are transmembrane proteins of the endomembrane system that play a poorly defined role in vesicle traffic between the endoplasmic reticulum and the Golgi apparatus. Various lines of evidence indicate that p24 proteins fall into four subfamilies (alpha, beta, gamma, and delta) and that tetramers are assembled containing one representative from each subfamily; however, the nature of the protein-protein interactions within these hetero-oligomers is unknown. We have identified a lumenal segment of yeast p24beta (Emp24p) that is necessary for its assembly into p24 complexes. Replacement of 52 C-terminal residues of Emp24p with the corresponding sequence from Erv25p (p24delta) generates a chimeric protein able to replace Emp24p in p24 complexes that retain partial function in vivo, ruling out a role for the transmembrane and cytosolic domains in specifying p24 interactions. Substitution of a further 50 residues, encompassing a heptad repeat region, abolishes the ability of the chimera to replace Emp24p but instead creates a protein that resembles its Erv25p parent in its requirement for stabilization by Emp24p. These data point to a role for coiled-coil interactions in directing subfamily-specific assembly of p24 oligomers that project into the lumen of transport vesicles, where they may act to exclude secretory cargo from coat protein complex type I-coated retrograde transport vesicles.     

Structural fingerprints of the Ras-GTPase activating proteins neurofibromin and p120GAP

Ras specific GTPase activating proteins (GAPs), neurofibromin and p120GAP, bind GTP bound Ras and efficiently complement its active site. Here we present comparative data from mutations and fluorescence-based assays of the catalytic domains of both RasGAPs and interpret them using the crystal structures. Three prominent regions in RasGAPs, the arginine-finger loop, the phenylalanine-leucine-arginine (FLR) region and alpha7/variable loop contain structural fingerprints governing the GAP function. The finger loop is crucial for the stabilization of the transition state of the GTPase reaction. This function is controlled by residues proximal to the catalytic arginine, which are strikingly different between the two RasGAPs. These residues specifically determine the orientation and therefore the positioning of the arginine finger in the Ras active site. The invariant FLR region, a hallmark for RasGAPs, indirectly contributes to GTPase stimulation by forming a scaffold, which stabilizes Ras switch regions. We show that a long hydrophobic side-chain in the FLR region is crucial for this function. The alpha7/variable loop uses several conserved residues including two lysine residues, which are involved in numerous interactions with the switch I region of Ras. This region determines the specificity of the Ras-RasGAP interaction.     

Interactions between the bud emergence proteins Bem1p and Bem2p and Rho- type GTPases in yeast

The SH3 domain-containing protein Bem1p is needed for normal bud emergence and mating projection formation, two processes that require asymmetric reorganizations of the cortical cytoskeleton in Saccharomyces cerevisiae. To identify proteins that functionally and/or physically interact with Bem1p, we screened for mutations that display synthetic lethality with a mutant allele of the BEM1 gene and for genes whose products display two-hybrid interactions with the Bem1 protein. CDC24, which is required for bud emergence and encodes a GEF (guanine- nucleotide exchange factor) for the essential Rho-type GTPase Cdc42p, was identified during both screens. The COOH-terminal 75 amino acids of Cdc24p, outside of the GEF domain, can interact with a portion of Bem1p that lacks both SH3 domains. Bacterially expressed Cdc24p and Bem1p bind to each other in vitro, indicating that no other yeast proteins are required for this interaction. The most frequently identified gene that arose from the bem1 synthetic-lethal screen was the bud-emergence gene BEM2 (Bender and Pringle. 1991. Mol. Cell Biol. 11:1295-1395), which is allelic with IPL2 (increase in ploidy; Chan and Botstein, 1993. Genetics. 135:677-691). Here we show that Bem2p contains a GAP (GTPase-activating protein) domain for Rho-type GTPases, and that this portion of Bem2p can stimulate in vitro the GTPase activity of Rho1p, a second essential yeast Rho-type GTPase. Cells deleted for BEM2 become large and multinucleate. These and other genetic, two-hybrid, biochemical, and phenotypic data suggest that multiple Rho-type GTPases control the reorganization of the cortical cytoskeleton in yeast and that the functions of these GTPases are tightly coupled. Also, these findings raise the possibility that Bem1p may regulate or be a target of action of one or more of these GTPases.     

Microcell-mediated cotransfer of genes specifying methotrexate resistance, emetine sensitivity, and chromate sensitivity with Chinese hamster chromosome 2.

Many selectable mutants of somatic Chinese hamster cells have been described, but very few of the mutations have been mapped to specific chromosomes. We have utilized the microcell-mediated gene transfer technique to establish the location of three selectable genetic markers on chromosome 2 of Chinese hamster. Microcells were prepared from the methotrexate-resistant MtxRIII line of Flintoff et al. (Somatic Cell Genet. 2:245-261, 1976) and fused to wild-type CHO cells, and microcell hybrids (transferants) were selected in medium containing methotrexate. All transferants were karyotyped and found to contain a marker chromosome from the donor MtxRIII line. This marker chromosome, called 2p-, consisted of a chromosome 2 with a reduced short arm resulting from a reciprocal translocation between 2p and 5q. In experiments utilizing emetine-resistant (Emtr) or chromate-resistant (Chrr) recipient cells it was found that the emt+ and chr+ wild-type genes were cotransferred with the 2p- chromosomes. Karyotype analysis of several transferants with rearranged or broken 2p- markers allowed regional localization of the emt and chr loci to the proximal third of the long arm and localization of the gene or genes conferring methotrexate resistance to the short arm. These results confirm our earlier assignment of the emt and chr loci to chromosome 2 in Chinese hamster.