It has been reported that growth factors activate Ras through a complex of an adaptor type SH2-containing molecule, Grb2, and a Ras guanine nucleotide-releasing protein (GNRP), mSos. We report on the involvement of another adaptor molecule, CRK, in the activation of Ras The Ras protein is regulated by three groups of proteins (4). First, there are proteins which accelerate the intrinsic GTPase activity of Ras, converting active GTP-bound Ras to the inactive GDP-bound state. One of these proteins, Ras GTPase-activating protein, contains two SH2 domains and is phosphorylated on tyrosine residues after stimulation by various growth factors (25). In PC12 cells, the GTPase-activating protein is activated after treatment with nerve growth factor (NGF) (18). However, activation of Ras does not always correlate with tyrosine phosphorylation of the GTPase-activating protein in hematopoietic cells (9), fibroblasts (25), and epithelial cells (12,29). The second group of proteins, guanine nucleotide-releasing proteins (GNRPs) for the Ras family, activate Ras by converting GDP-Ras to the GTP-bound state. It has been demonstrated that GDP-GTP exchange-stimulatory activity increases in PC12 cells treated with NGF (18). The third group of proteins, guanine nucleotide dissociation inhib-* Corresponding author. Mailing address:

Epidermal growth factor (EGF) receptor (EGFR) can induce cell growth and transformation in a ligand-dependent manner. To examine whether the autophosphorylation of EGFR correlates with the capacity of the activated EGFR to induce cell growth and transformation, we truncated the human EGFR just after residue 1011, removing all three major autophosphorylation sites (DEL1011). Further, a point mutation was introduced at another autophosphorylation site, Tyr-992 -* Phe (DEL1011+F992). The wild-type and mutant receptors were stably expressed in a NIH 3T3 variant cell line that expresses an extremely low level of endogenous EGFR and does not grow with EGF. As expected, DEL1011 and DEL1011+F992 were found to be severely impaired in EGF-induced autophosphorylation, due to the deletion of the appropriate target tyrosines. However, mutant receptors still could induce EGF-dependent DNA synthesis, morphological transformation, and anchorage-independent growth, although the extent of these was significantly reduced when compared with wild-type EGFR. EGF-induced tyrosine phosphorylation of Ras-GTPase activating protein-associated protein p62 and phospholipase C yl was dramatically reduced in the cells expressing DEL1011 and DEL1011+F992. On the other hand, tyrosine phosphorylation of Shc, complex formation of ShcGrb2/Ash, and activation of microtubule-associated protein kinase were still fully induced upon EGF stimulation without binding of Shc or Grb2/Ash to the mutant receptor. Thus, tyrosine phosphorylation of Shc may play a crucial role for activating Ras and generating mitotic signals by the activated EGFR mutant.

Radicicol, an inhibitor of protein-tyrosine kinase, was found to cause morphological reversion of v-Ha-ras-transformed NIH3T3 fibroblasts and T24 human urinary bladder carcinoma cells that contain an activated ras mutation. The network of actin stress fibers was restored during the treatment with radicicol. A similar morphological change was observed with another protein-tyrosine kinase inhibitor, herbimycin A. Radicicol did not cause any changes in the proportion of the active GTP binding form of p21ras or its subcellular localization. These results rule out the possibility that the morphological reversion by radicicol is due to direct or indirect inhibition of the p21ras function. Cycloheximide and actinomycin D inhibited the morphological change by radicicol, suggesting that the induced transcription of a gene(s) followed by de novo protein synthesis is required for suppression of the transformed phenotype in ras-transformed cells by tyrosine kinase inhibitors.

To elucidate the mechanism of transcription and replication of Sendai virus, we developed an efficient and faithful in vitro transcription system using purified virus particles. The in vitro RNA synthesis was almost entirely dependent on the addition of eukaryotic cell extracts, including those from various cultured mammalian cells, mammalian tissues, and even from plant cells. The RNA products were almost identical to authentic mRNA species synthesized in the infected cells, in their size distribution, the presence of 3'-poly(A) tail and the presence of methylated 5'-cap structure (m7GpppAm). Ribonuclease protection experiments after annealing the in vitro RNA with viral genomic RNA (vRNA) indicated that the virion-associated RNA-dependent RNA polymerase transcribes correct regions of the RNA genome in vitro. The active component(s) that is required for Sendai virus mRNA synthesis was partially purified from bovine brain and was separated into at least two complementary fractions, one of which could be replaced by highly purified cellular tubulin. When viral ribonucleoprotein complexes were used instead of virus particles in the in vitro transcription, only Sendai virus-infected cell extracts supported mRNA synthesis, and extracts from uninfected cells or cells infected with other viruses were found to be inert. These results suggest that, in addition to the general factors which are present ubiquitously in eukaryotic cells, a factor(s) specific to Sendai virus-infection is required for Sendai virus transcription.

No abstract