Ca2+-dependent monomer and dimer formation switches CAPRI Protein between Ras GTPase-activating protein (GAP) and RapGAP activities

CAPRI is a member of the GAP1 family of GTPase-activating proteins (GAPs) for small G proteins. It is known to function as an amplitude sensor for intracellular Ca(2+) levels stimulated by extracellular signals and has a catalytic domain with dual RasGAP and RapGAP activities. Here, we have investigated the mechanism that switches CAPRI between its two GAP activities. We demonstrate that CAPRI forms homodimers in vitro and in vivo in a Ca(2+)-dependent manner. The site required for dimerization was pinpointed by deletion and point mutations to a helix motif that forms a hydrophobic face in the extreme C-terminal tail of the CAPRI protein. Deletion of this helix motif abolished dimer formation but did not affect translocation of CAPRI to the plasma membrane upon cell stimulation with histamine. We found that dimeric and monomeric CAPRI coexist in cells and that the ratio of dimeric to monomeric CAPRI increases upon cell stimulation with histamine. Free Ca(2+) at physiologically relevant concentrations was both necessary and sufficient for dimer formation. Importantly, the monomeric and dimeric forms of CAPRI exhibited differential GAP activities in vivo; the wild-type form of CAPRI had stronger RapGAP activity than RasGAP activity, whereas a monomeric CAPRI mutant showed stronger RasGAP than RapGAP activity. These results demonstrate that CAPRI switches between its dual GAP roles by forming monomers or homodimers through a process regulated by Ca(2+). We propose that Ca(2+)-dependent dimerization of CAPRI may serve to coordinate Ras and Rap1 signaling pathways.     

Modulation of the eukaryotic initiation factor 2 alpha-subunit kinase PERK by tyrosine phosphorylation

The endoplasmic reticulum (ER)-resident protein kinase PERK attenuates protein synthesis in response to ER stress through the phosphorylation of translation initiation factor eIF2alpha at serine 51. ER stress induces PERK autophosphorylation at several serine/threonine residues, a process that is required for kinase activation and phosphorylation of eIF2alpha. Herein, we demonstrate that PERK also possesses tyrosine kinase activity. Specifically, we show that PERK is capable of autophosphorylating on tyrosine residues in vitro and in vivo. We further show that tyrosine 615, which is embedded in a highly conserved region of the kinase domain of PERK, is essential for autocatalytic activity. That is, mutation of Tyr-615 to phenylalanine compromises the autophosphorylation capacity of PERK and the phosphorylation of eIF2alpha in vitro and in vivo. The Y615F mutation also impairs the ability of PERK to induce translation of ATF4. Immunoblot analyses with a phosphospecific antibody confirm the phosphorylation of PERK at Tyr-615 both in vitro and in vivo. Thus, our data classify PERK as a dual specificity kinase whose regulation by tyrosine phosphorylation contributes to its optimal activation in response to ER stress.     

RNA polymerase of influenza virus. III. Isolation of RNA polymerase-RNA complexes from influenza virus PR8

Ribonucleoprotein (RNP) cores with RNA-synthesizing activity were prepared in two fractions, M protein-free and M protein-associated, from detergent-treated influenza virus PR8 by centrifugation through a discontinuous triple gradient of cesium sulfate, glycerol, and NP-40. The M-free RNP was fractionated by phosphocellulose column chromatography into two major RNP forms, A and B, which differed in the content of P proteins, while the M-associated RNP gave only the low P-content Form-B RNP. Starting from the high P-content Form-A RNP, an RNA-P proteins complex virtually free from NP protein was isolated by cesium sulfate equilibrium centrifugation. The complex, containing only three P proteins (P1, P2, and P3), was still active in catalyzing RNA synthesis in vitro without addition of exogenous template, indicating that NP protein is not required for the catalysis of RNA synthesis. RNA synthesis by the isolated RNA-P proteins complex was dependent on either ApG or capped RNA primers, and required four ribonucleoside triphosphates as substrates. The RNA product in this reaction was hybridizable to viral RNA. A complex of one each of the three P proteins was separated from RNA by glycerol gradient centrifugation after ribonuclease treatment or cesium chloride equilibrium centrifugation.     

Adenovirus VAI RNA complexes with the 68 000 Mr protein kinase to regulate its autophosphorylation and activity.

We have investigated the interaction of VAI RNA with the interferon-induced, double-stranded (ds) RNA-activated protein kinase, P68, both of which regulate protein synthesis in adenovirus-infected cells. Previous work has shown that during infection by the VAI RNA-negative mutant, dl331, both viral and cellular protein synthesis are inhibited due to phosphorylation of the alpha-subunit of the eukaryotic initiation factor, eIF-2, by the P68 protein kinase. Utilizing monoclonal antibodies specific for P68, we demonstrated that the physical levels of P68 in dl331-infected, wild-type Ad2-infected and uninfected cells were all comparable suggesting that the elevated kinase activity detected during mutant infection was not due to increased P68 synthesis. To examine the basis of the increased activity of P68, the protein kinase was purified from infected-cell extracts using the monoclonal antibody. We found that P68 was heavily autophosphorylated during dl331 infection but not during wild-type or mock infection. The extent of autophosphorylation correlated with elevated P68 activity and the loss of the dsRNA requirements to phosphorylate the exogenous substrates, eIF-1 alpha and histones. We also analyzed VAI RNA function in vitro and present evidence that purified VAI RNA can block the autophosphorylation of P68 in the ribosomal salt wash fraction of interferon-treated cells. Finally we suggest VAI RNA functions through a direct interaction with the P68 protein kinase, since we demonstrated that VAI RNA forms a complex with P68 both in vitro and in vivo.     

In vivo modifications of the maize mitochondrial small heat stress protein, HSP22

A maize (Zea mays L.) small heat shock protein (HSP), HSP22, was previously shown to accumulate to high levels in mitochondria during heat stress. Here we have purified native HSP22 and resolved the protein into three peaks using reverse phase high performance liquid chromatography. Mass spectrometry (MS) of the first two peaks revealed the presence of two HSP22 forms in each peak which differed in mass by 80 daltons (Da), indicative of a monophosphorylation. Phosphorylation of HSP22 by [gamma-(32)P]ATP was also observed in mitochondria labeled in vitro, but not when purified native HSP22 was similarly used, demonstrating that HSP22 does not autophosphorylate, implicating a kinase involvement in vivo. Collisionally induced dissociation tandem MS (CID MS/MS) identified Ser(59) as the phosphorylated residue. We have also observed forms of HSP22 that result from alternative intron splicing. The two HSP22 proteins in the first peak were approximately 57 Da larger than the two HSP22 proteins in the second peak. MS analysis revealed that the +57-Da forms have an additional Gly residue directly N-terminal of the expected Asp(84), which had been converted to an Asn residue. These results are the first demonstrations of phosphorylation and alternative intron splicing of a plant small HSP.     

Evidence for autocrine activation of a tyrosine kinase in a human gastric carcinoma cell line

Phosphotyrosine (P-Tyr) antibodies have been used to identify the phosphorylated forms of growth factor receptors and oncogene-coded tyrosine kinases. Western blot analysis of a gastric carcinoma cell line with P-Tyr antibodies revealed a tyrosine-phosphorylated protein of Mr 145,000 (P145). In addition, in vitro phosphorylation with (gamma-32P)ATP or P-Tyr immunoprecipitates of the same cells resulted in labelling of this protein on tyrosine. P145 appears to be a transmembrane glycoprotein, with features suggestive of a growth factor receptor. However, the in vivo or in vitro addition of known growth factors did not affect P145 tyrosine phosphorylation. We now report that P145 is rapidly dephosphorylated in vivo when cells are exposed to low pH, a condition known to dissociate ligands from their receptors. The addition of serum-free medium, conditioned by the gastric carcinoma cells, fully restores the tyrosine phosphorylation lost with acid treatment. These data suggest that the activity responsible for P145 phosphorylation on tyrosine, whether intrinsic to P145 itself or due to an associated kinase, is stimulated by a factor secreted by the tumor cells themselves.     

Development of an oxazolopyridine series of dual thrombin/factor Xa inhibitors via structure-guided lead optimization

Thrombin-inhibitor X-ray crystal structures, in combination with the installation of binding elements optimized within the pyrazinone series of thrombin inhibitors, were utilized to transform a weak triazolopyrimidine lead into a series of potent oxazolopyridines. A modification intended to attenuate plasma protein binding (i.e., conversion of the P3 pyridine to a piperidine) conferred significant factor Xa activity to this series. Ultimately, these dual thrombin/factor Xa inhibitors demonstrated excellent in vitro and in vivo anticoagulant efficacy.     

Purification, primary structure characterization, and cellular distribution of two forms of cellular retinol-binding protein, type II from adult rat small intestine

Cellular retinol-binding protein type II (CRBP(II)) is a major protein in the small intestine, accounting for more than 1% of the soluble protein recovered from rat jejunal mucosa. Two forms of the protein, called CRBP(II)A and CRBP(II)B, were purified from rat small intestine using a three-column procedure. The two forms were present in equal abundance. The primary structures of CRBP(II)A and CRBP(II)B were determined using a combination of techniques including amino acid composition and sequence analyses, and fast atom bombardment and gas chromatography-electron impact mass spectrometry. The primary structures of both proteins were found to be identical, but they differed in their NH2-terminal processing. CRBP(II)B was acetylated at its NH2 terminus, while CRBP(II)A was not. The results also confirmed the amino acid sequence of CRBP(II)A that was deduced from the cDNA sequence by Li et al. (Li, E., Demmer, L. A., Sweetser, D. A., Ong, D. E., and Gordon, J. I. (1986) Proc. Natl. Acad. Sci. U.S.A. 83, 5770-5783). Antibodies capable of distinguishing between the two forms of CRBP(II) were used for immunohistochemical studies which indicated that the organ and cellular distributions of the two forms were identical. The 50% acetylation observed here in vivo fits the pattern predicted by recent in vitro studies which described the effect of NH2-terminal sequence on cotranslational NH2-terminal processing of cytosolic proteins (Boissel, J. P., Kasper, T. J., and Bunn, H. F. (1988) J. Biol. Chem. 263, 8443-8449). Our results provide a basis for investigating the possibility of different roles of CRBP(II)A and CRBP(II)B within cells, as well as the importance of acetylation of the amino terminus for these biological functions.     

The association of insulin-elicited phosphotyrosine proteins with src homology 2 domains.

The interactions of the phosphotyrosine (Tyr(P))-containing proteins in basal and insulin-stimulated 3T3-L1 adipocytes with src homology 2 (SH2) domains from phosphatidylinositol 3-kinase (PI3K), ras GTPase-activating protein (GAP), and phospholipase C gamma have been examined. The Tyr(P) forms of the insulin receptor and its 160-kDa substrate protein (pp160) associated with fusion proteins containing either or both the SH2 domains of PI3K, but not with fusion proteins containing the two SH2 domains of GAP or phospholipase C gamma. These results demonstrate a specificity for the association of the Tyr(P) form of the insulin receptor and pp160 with SH2 domains that parallels the reported effects of insulin on PI3K, GAP, and phospholipase C gamma in vivo. Immunoprecipitates of pp160 from the cytosol of insulin-treated, but not basal, 3T3-L1 adipocytes contained PI3K activity. Moreover, the Tyr(P) form of pp160 with associated PI3K activity migrated at 10 S on a sucrose velocity gradient, whereas the Tyr(P) form without associated activity migrated at 6 S. These findings indicate that the Tyr(P) form of pp160 associates directly with PI3K in vivo.     

A series of annexins from human placenta and their characterization by use of an endogenous phospholipase A2.

Membranes from human placenta contain proteins which inhibit the activity of phospholipases A2 by binding to phospholipid thus impeding substrate availability. We used unilamellar mixed liposomes and a partially purified cytosolic phospholipase A2 from placenta for characterizing this substrate-depleting activity. A major portion of these inhibitory proteins was released by extracting washed membranes with a Ca+(+)-chelator. Biochemical fractionation and systematic analysis resulted in the unequivocal identification of a series of annexin proteins. We describe a straightforward procedure which allows to obtain 8 annexins from placenta either in pure form or as a mixture of two annexins. One of them was obtained in two forms which had the same molecular mass of 68 kDa but differed in charge. We also present suggestive evidence for a novel annexin I-related polypeptide of Mr 45,000 which is an excellent in vitro substrate for protein kinase C. We estimate that about 2% of the total placental membrane proteins are annexins. For achieving half inhibition of phospholipase A2 activity with pure annexins, up to a 6.5-fold difference in the amounts of protein was observed when calculated on a molar basis. This suggests specificity of individual annexin species.     

Homeostatic plasticity studied using in vivo hippocampal activity-blockade: synaptic scaling, intrinsic plasticity and age-dependence

Homeostatic plasticity is thought to be important in preventing neuronal circuits from becoming hyper- or hypoactive. However, there is little information concerning homeostatic mechanisms following in vivo manipulations of activity levels. We investigated synaptic scaling and intrinsic plasticity in CA1 pyramidal cells following 2 days of activity-blockade in vivo in adult (postnatal day 30; P30) and juvenile (P15) rats. Chronic activity-blockade in vivo was achieved using the sustained release of the sodium channel blocker tetrodotoxin (TTX) from the plastic polymer Elvax 40W implanted directly above the hippocampus, followed by electrophysiological assessment in slices in vitro. Three sets of results were in general agreement with previous studies on homeostatic responses to in vitro manipulations of activity. First, Schaffer collateral stimulation-evoked field responses were enhanced after 2 days of in vivo TTX application. Second, miniature excitatory postsynaptic current (mEPSC) amplitudes were potentiated. However, the increase in mEPSC amplitudes occurred only in juveniles, and not in adults, indicating age-dependent effects. Third, intrinsic neuronal excitability increased. In contrast, three sets of results sharply differed from previous reports on homeostatic responses to in vitro manipulations of activity. First, miniature inhibitory postsynaptic current (mIPSC) amplitudes were invariably enhanced. Second, multiplicative scaling of mEPSC and mIPSC amplitudes was absent. Third, the frequencies of adult and juvenile mEPSCs and adult mIPSCs were increased, indicating presynaptic alterations. These results provide new insights into in vivo homeostatic plasticity mechanisms with relevance to memory storage, activity-dependent development and neurological diseases.     

In Vitro Incorporation of Molybdate into Demolybdoproteins in Escherichia coli

When Escherichia coli was grown in the presence of tungstate, inactive forms of two molybdoenzymes, nitrate reductase and formate dehydrogenase, accumulated and were converted to their active forms upon incubation of cell suspensions with molybdate and chloramphenicol. The conversion to the active enzymes did not occur in cell extracts. When incubated with [(99)Mo]molybdate and chloramphenicol, the tungstate-grown cells incorporated (99)Mo into protein components which were released from membranes by procedures used to release nitrate reductase and formate dehydrogenase and which migrated with these activities on polyacrylamide gels. Although neither activity was formed during incubation of the crude extract with molybdate, (99)Mo was incorporated into protein components which were released from the membrane fraction under the same conditions and were similar to the active enzymes in their electrophoretic properties. The in vitro incorporation of (99)Mo occurred specifically into these components and was equal to or greater than the amount incorporated in vivo under the same conditions. Molybdenum in preformed, active nitrate reductase and formate dehydrogenase did not exchange with [(99)Mo]molybdate, demonstrating that the observed incorporation depended on the demolybdo forms of the enzymes. We conclude that molybdate may be incorporated into the demolybdo forms both in vivo and in vitro; some unknown additional factor or step, required for active enzyme formation, occurs in vivo but not in vitro under the conditions employed.     

Point and deletion mutations eliminate one or both methyl group transfers catalysed by the yeast TRM1 encoded tRNA (m22G26)dimethyltransferase.

Guanosine at position 26 in eukaryotic tRNAs is usually modified to N2 , N2 -dimethylguanosine (m22G26). In Saccharomyces cerevisiae , this reaction is catalysed by the TRM1 encoded tRNA (m22G26)dimethyltransferase. As a prerequisite for future studies, the yeast TRM1 gene was expressed in Escherichia coli and the His-tagged Trm1 protein (rTrm1p) was extensively purified. rTrm1p catalysed both the mono- and dimethylation of G26 in vivo in Escherichia coli tRNA and in vitro in yeast trm1 mutant tRNA. The TRM1 gene from two independent wild-type yeast strains differed at 14 base positions causing two amino acid exchanges . Exchange of the original Ser467 for Leu caused a complete loss of enzyme activity in vitro against trm1 yeast tRNA. Comparatively short N- or C-terminal deletions from the 570 amino acid long Trm1 polypeptide decreased or eliminated the enzyme activity, as did some point mutations within these regions. This indicated that the protein is not a two domain peptide with the enzyme activity localised to one of the domains, but rather that both ends of the polypeptide seem to interact to influence the conformation of those parts that make up the RNA-binding site and/or the active site of the enzyme.     

A membrane-associated, carbohydrate-modified form of the v-abl protein that cannot be phosphorylated in vivo or in vitro.

Abelson murine leukemia virus encodes a transforming protein which contains tyrosine kinase activity and is phosphorylated in vivo and in vitro. We found that P160 and P160-derived virus strains expressed an additional, altered v-abl protein which could not be phosphorylated. The altered v-abl protein (L-v-abl) differed from the phosphorylated form (K-v-abl) in that it was glycosylated and localized exclusively to the membrane fraction. Tunicamycin inhibition of N-linked carbohydrate addition did not restore phosphorylation. It did, however, reveal that L-v-abl had additional sequences relative to K-v-abl. The coding sequences required for this region and for the expression of L-v-abl were identified by replacing sequences in the P120 virus genome, which did not express L-v-abl, with sequences from the P160 virus genome. The necessary sequences were localized to the Moloney murine leukemia virus-derived gag gene. Comparison between the in vitro altered P120 and wild-type P120 virus strains indicated that expression of L-v-abl did not increase the efficiency of lymphoid transformation. Although the biological role of L-v-abl is not clear, our analyses have revealed that a specific amino terminal gag sequence can prevent v-abl from acting as a kinase substrate and can alter the cellular localization and modification of v-abl. These properties distinguish L-v-abl from previously reported v-abl proteins.     

Opposite displacement of helix F in attractant and repellent signaling by sensory rhodopsin-Htr complexes

Two forms of the phototaxis receptor sensory rhodopsin I distinguished by differences in its photoactive site have been shown to be directly correlated with attractant and repellent signaling by the dual-signaling protein. In prior studies, differences in the photoactive site defined the two forms, namely the direction of light-induced proton transfer from the chromophore and the pK(a) of an Asp counterion to the protonated chromophore. Here, we show by both in vivo and in vitro measurements that the two forms are distinct protein conformers with structural similarities to two conformers seen in the light-driven proton transport cycle of the related protein bacteriorhodopsin. Measurements of spontaneous cell motility reversal frequencies, an in vivo measure of histidine kinase activity in the phototaxis system, indicate that the two forms are a photointerconvertible pair, with one conformer activating and the other inhibiting the kinase. Protein conformational changes in these photoconversions monitored by site-directed spin labeling show that opposite structural changes in helix F, distant from the photoactive site, correspond to the opposite phototaxis signals. The results provide the first direct evidence that displacements of helix F are directly correlated with signaling and impact our understanding of the sensory rhodopsin I signaling mechanism and the evolution of diverse functionality in this protein family.