Threonine 680 Phosphorylation of FLJ00018/PLEKHG2, a Rho Family-specific Guanine Nucleotide Exchange Factor,by Epidermal Growth Factor Receptor Signaling Regulates Cell Morphology of Neuro-2a Cells

FLJ00018/PLEKHG2 is a guanine nucleotide exchange factor for the small GTPases Rac and Cdc42 and has been shown to mediate the signaling pathways leading to actin cytoskeleton reorganization. The function of FLJ00018 is regulated by the interaction of heterotrimeric GTP-binding protein Gβγ subunits or cytosolic actin. However, the details underlying the molecular mechanisms of FLJ00018 activation have yet to be elucidated. In the present study we show that FLJ00018 is phosphorylated and activated by β₁-adrenergic receptor stimulation-induced EGF receptor (EGFR) transactivation in addition to Gβγ signaling. FLJ00018 is also phosphorylated and activated by direct EGFR stimulation. The phosphorylation of FLJ00018 by EGFR stimulation is mediated by the Ras/mitogen-activated protein kinase (MAPK) pathway. Through deletion and site-directed mutagenesis studies, we have identified Thr-680 as the major site of phosphorylation by EGFR stimulation. FLJ00018 T680A, in which the phosphorylation site is replaced by alanine, showed a limited response of the Neuro-2a cell morphology to EGF stimulation. Our results provide evidence that stimulation of the Ras/MAPK pathway by EGFR results in FLJ00018 phosphorylation at Thr-680, which in turn controls changes in cell shape.     

PLEKHG2/FLJ00018, a Rho family-specific guanine nucleotide exchange factor,is tyrosine phosphorylated via the EphB2/cSrc signaling pathway

PLEKHG2/FLJ00018, a Rho family-specific guanine nucleotide exchange factor (RhoGEF), is activated by heterotrimeric GTP-binding protein (G protein) Gβγ subunits, and in turn activates the small G protein Rac and Cdc42, which have been shown to mediate signaling pathways leading to actin cytoskeletal reorganization. In the present study, we show that co-expression of the constitutively active mutant of cSrc, a non-receptor tyrosine kinase, and PLEKHG2 induced the tyrosine phosphorylation of PLEKHG2 in HEK293 cells. Through deletion and base substitution mutagenesis we have identified Tyr489 of PLEKHG2 as the site phosphorylated by cSrc. Furthermore, using a high-throughput src homology 2 (SH2) domain binding assay, the SH2 domain of ABL1 and the PI 3-kinse regulator subunit (PIK3R3) were identified as candidates for the binding partner of tyrosine-phosphorylated PLEKHG2. The interaction between PLEKHG2 and the full-length of PIK3R3, but not ABL1, occurs in a tyrosine-phosphorylation-dependent manner. Furthermore, PLEKHG2 is tyrosine phosphorylated at Tyr489 by ephrinB2 receptor signaling via cSrc. Investigation of the physiological function of tyrosine phosphorylation at Tyr489 in PLEKHG2 remains a subject for future studies.     

Role of the head-to-tail overlap region in smooth and skeletal muscle beta-tropomyosin

Tropomyosin (Tm) is an alpha-helical, parallel, two-chain coiled coil which binds along the length of actin filaments in both muscle and non-muscle cells. Smooth and skeletal muscle Tms differ extensively at the C-terminus encoded by exon 9. Replacement of the striated muscle specific exon 9a-encoded C-terminus with that encoded by exon 9d expressed in smooth muscle and non-muscle cells increases the affinity of unacetylated alpha-SkTm for actin [Cho, Y. J., and Hitchcock-Degregori, S. E. (1991) Proc. Natl. Acad. Sci. U.S.A. 88, 10153-10157]. Here we show that swapping 10 amino acids at the C-terminus of beta-SkTm with the corresponding 10 amino acids of beta-SmTm had little effect on the regulation of S1 binding to actin, but Tm viscosity, Tm binding to actin, and troponin T1 binding to Tm all become like smooth rather than SkTm. beta-SkTm point mutations show that these properties are largely defined by the amino acids at two positions, 277 and 279. The N279L mutation reduces the viscosity of beta-SkTm to close to beta-SmTm values, while both residues contribute to the binding of TnT1. We also show that removing the first 11 N-terminal amino acids of beta-SmTm to make the mutant DeltaN-betaSmTm results in a 10-fold weakening in actin affinity compared to that of beta-SmTm. CD studies show no difference in thermal unfolding between beta-SmTm and DeltaN-betaSmTm; however, the viscosity of DeltaN-betaSmTm is much lower than that of the control. The results suggest that DeltaN-betaSmTm was unable to form filaments in solution but can form filaments on actin.     

The amino terminus of muscle tropomyosin is a major determinant for function

The amino-terminal region of muscle tropomyosin is highly conserved among muscle and 284-residue non-muscle tropomyosins. Analysis of fusion and nonfusion striated alpha-tropomyosins and a mutant in which residues 1-9 have been deleted has shown that the amino terminus is crucial for function. The presence of 80 amino acids of a nonstructural influenza virus protein (NS1) on the amino terminus of tropomyosin allows magnesium-independent binding of tropomyosin to actin. The fusion tropomyosin inhibits the actomyosin S1 ATPase at all myosin S1 concentrations tested, indicating that the presence of the fusion peptide prevents myosin S1 from switching the actin filament from the inhibited to the potentiated state. Nonfusion tropomyosin, an unacetylated form, has no effect on the actomyosin S1 ATPase, though it regulates normally with troponin. Deletion of residues 1-9, which are believed to overlap with the carboxyl-terminal end of tropomyosin in the thin filament, results in loss of tropomyosin function. The mutant is unable to bind to actin, in the presence and absence of troponin, and it has no regulatory function. The removal of the first 9 residues of tropomyosin is much more deleterious than removal of the last 11 by carboxypeptidase digestion. We suggest that the structure of the amino-terminal region and acetylation of the initial methionine are crucial for tropomyosin function.     

Structural and functional properties of the non-muscle tropomyosins

Summary The non-muscle tropomyosins (TMs), isolated from such tissues as platelets, brain and thyroid, are structurally very similar to the muscle TMs, being composed of two highly -helical subunits wound around each other to form a rod-like molecule. The non-muscle TMs are shorter than the muscle TMs; sequence analysis demonstrates that each subunit of equine platelet TM consists of 247 amino acids, 37 fewer than for skeletal muscle TM. The major differences in sequence between platelet and skeletal muscle TM are found near the amino and carboxyl terminal ends of the proteins. Probably as the result of such alterations, the non-muscle TMs aggregate in a linear end-to-end manner much more weakly than do the muscle TMs. Since end-to-end interactions are responsible for the highly cooperative manner in which TM binds to actin, the non-muscle TMs have a lower affinity for actin filaments than do the muscle TMs. However, the attachment of other proteins to actin (e.g. the Tn-I subunit of skeletal muscle troponin or the S-1 subfragment of skeletal muscle myosin) can increase the affinity of actin filaments for non-muscle TM. The non-muscle TMs interact functionally with the Tn-I component of skeletal muscle troponin to inhibit the ATPase activity of muscle actomyosin and with whole troponin to regulate the muscle actomyosin ATPase in a Ca⁺⁺-dependent manner, even though one of the binding sites for troponin on skeletal TM is missing in non-muscle TM. A novel actomyosin regulatory system can be produced using Tn-I, calmodulin and non-muscle TM; in this case inhibition is released when the non-muscle TM detaches from the actin filament in the presence of Ca⁺⁺. Although it has not yet been demonstrated that the non-muscle TMs participate in a Ca⁺⁺-dependent contractile regulatory system in vivo it does appear that they are associated with actin filaments in vivo.     

Characterization of a Drosophila cDNA clone that encodes a 252-amino acid non-muscle tropomyosin isoform

We report here the isolation and DNA sequence of a cDNA clone encoding a 252-amino acid non-muscle or cytoskeletal tropomyosin (cTm) isoform from Drosophila. The Drosophila cTm shows considerable homology with vertebrate cTm throughout the middle portion of the molecule. The amino-terminal end of the molecule, however, shows less homology and contains five more amino acids than the equine platelet and human tropomyosins. There is also a proline at position 6 in the Drosophila protein. The carboxyl-terminal 27 amino acids also show little homology with vertebrate non-muscle tropomyosins. This is a region of the molecule that shows considerably diversity among other Drosophila tropomyosins and vertebrate tropomyosins. A comparison of the DNA sequence of the cTm cDNA and a previously reported muscle tropomyosin II cDNA sequence shows regions of identical DNA sequence alternating with regions of nonidentical sequence, suggesting that both mRNAs are produced by alternate splicing of the same gene.     

Clostridium perfringens iota toxin ADP-ribosylates skeletal muscle actin in Arg-177

Clostridium perfringens iota toxin ADP-ribosylates actin. Substrates of C. perfringens toxin are both non-muscle beta/gamma-actin and skeletal muscle actin. This finding suggests that C. perfringens iota ADP-ribosylates the same amino acid in skeletal muscle and non-muscle actin as does C. botulinum C2 toxin in non-muscle actin. Protein chemical analysis involving thermolysin cleavage on [32P]ADP-ribosylated actin or tryptic digestion followed by a secondary thermolysin cleavage of the radiolabelled fragments showed one major site of ADP-ribosylation. From its amino acid composition and sequence, the radiolabelled peptide was identified as peptide 175-177, locating the acceptor ADP-ribosyl amino acid as Arg-177.     

Yersinia protein kinase YopO is activated by a novel G-actin binding process

Pathogenic bacteria of the genus Yersinia employ a type III secretion system to inject effector proteins (Yops) into host cells. The Yops down-regulate host cell functions through unique biochemical activities. YopO, a serine/threonine kinase required for Yersinia virulence, is activated by host cell actin via an unknown process. Here we show that YopO kinase is activated by formation of a 1:1 complex with monomeric (G) actin but is unresponsive to filamentous (F) actin. Two separate G-actin binding sites, one in the N-terminal kinase region (amino acids 89-440) and one in the C-terminal guanine nucleotide dissociation inhibitor-like region (amino acids 441-729) of YopO, were identified. Actin binding to both of these sites was necessary for effective autophosphorylation of YopO on amino acids Ser-90 and Ser-95. A S90A/S95A YopO mutant was strongly reduced in substrate phosphorylation, suggesting that autophosphorylation activates YopO kinase activity. In cells the kinase activity of YopO regulated rounding/arborization and was specifically required for inhibition of Yersinia YadA-dependent phagocytosis. Thus, YopO kinase is activated by a novel G-actin binding process, and this appears to be crucial for its anti-host cell functions.     

Binding of filamentous actin to anthrax toxin receptor 1 decreases its association with protective antigen

ANTXR1 is a type I membrane protein that binds the protective antigen (PA) component of anthrax toxin. The cytosolic domain of ANTXR1 has a novel actin-binding region that influences the interaction of the ectodomain with PA. Here, we have investigated features of the cytosolic domain of ANTXR1 that reduce the association of the receptor with PA. We mutated a stretch of conserved acidic amino acids adjacent to the actin-binding region and found that the mutation increased the affinity for monomeric actin in vitro. ANTXR1 bearing this mutation exhibited increased association with the cytoskeleton and bound less PA compared to the wild-type receptor, confirming the inverse correlation between the two interactions. To determine whether binding of actin is sufficient to regulate the ectodomain, we replaced the actin-binding region of ANTXR1 with that from the yeast protein abp140 and with the WH2 domain of WAVE2. Although both of these domains bound monomeric actin in vitro, only the sequence from abp140 reduced binding of PA to a hybrid receptor. The actin binding regions of ANTXR1 and abp140, but not the WH2 domain, colocalized with actin stress fibers, which suggested that filamentous actin regulates ANTXR1. Consistent with this notion, disruption of actin filaments using latrunculin A increased the amount of PA bound to cells. This work provides evidence that cytoskeletal dynamics regulate ANTXR1 function.     

Neuronal IP3 3-Kinase is an F-actin–bundling Protein: Role in Dendritic Targeting and Regulation of Spine Morphology

The actin microstructure in dendritic spines is involved in synaptic plasticity. Inositol trisphosphate 3-kinase A (ITPKA) terminates Ins(1,4,5)P₃ signals emanating from spines and also binds filamentous actin (F-actin) through its amino terminal region (amino acids 1-66, N66). Here we investigated how ITPKA, independent of its kinase activity, regulates dendritic spine F-actin microstructure. We show that the N66 region of the protein mediates F-actin bundling. An N66 fusion protein bundled F-actin in vitro, and the bundling involved N66 dimerization. By mutagenesis we identified a point mutation in a predicted helical region that eliminated both F-actin binding and bundling, rendering the enzyme cytosolic. A fusion protein containing a minimal helical region (amino acids 9-52, N9-52) bound F-actin in vitro and in cells, but had lower affinity. In hippocampal neurons, GFP-tagged N66 expression was highly polarized, with targeting of the enzyme predominantly to spines. By contrast, N9-52-GFP expression occurred in actin-rich structures in dendrites and growth cones. Expression of N66-GFP tripled the length of dendritic protrusions, induced longer dendritic spine necks, and induced polarized actin motility in time-lapse assays. These results suggest that, in addition to its ability to regulate intracellular Ca²⁺ via Ins(1,4,5)P₃ metabolism, ITPKA regulates structural plasticity.     

Distinction between smooth muscle,fibroblasts and endothelial cells in culture by the use of fluoresceinated antibodies against smooth muscle actin

Summary FITC-labelled antibodies against native actin from chicken gizzard smooth muscle (Gröschel-Stewart et al., 1976) have been used to stain cultures of guinea-pig vas deferens and taenia coli, rabbit thoracic aorta, rat ventricle and chick skeletal muscle. The I-band of myofibrils of cardiac muscle cells and skeletal muscle myotubes stains intensely. In isolated smooth muscle cells, the staining is located exclusively on long, straight, non-interrupted fibrils which almost fill the cell. Smooth muscle cells which have undergone morphological dedifferentiation to resemble fibroblasts with both phase-contrast microscopy and electronmicroscopy still stain intensely with the actin antibody. In those muscle cultures which contain some fibroblasts or endothelial cells, the non-muscle cells are not stained with the actin antibody even when the reactions are carried out at 37° C for 1 h or after glycerination. Prefusion skeletal muscle myoblasts also do not stain with this antibody.It is concluded that the actin antibody described in this report is directed against a particular sequence of amino acids in muscle actin which is not homologous with non-muscle actin. The usefulness of this antibody in determining the origin of cells in certain pathological conditions such as atherosclerosis is discussed.This work was supported by the Life Insurance Medical Research Fund of Australia and New Zealand, the National Heart Foundation of Australia, the Deutsche Forschungsgemeinschaft and the Wellcome Trust (London). We thank Janet D. McConnell for excellent technical assistance     

ADP-ribosylation of skeletal muscle and non-muscle actin by Clostridium perfringens iota toxin

The enzymatically active component ia of Clostridium perfringens iota toxin ADP-ribosylated actin in human platelet cytosol and purified platelet beta/gamma-actin, in a similar way to that been reported for component I of botulinum C2 toxin. ADP-ribosylation of cytosolic and purified actin by either toxin was inhibited by 0.1 mM phalloidin indicating that monomeric G-actin but not polymerized F-actin was the toxin substrate. Perfringens iota toxin and botulinum C2 toxin were not additive in ADP-ribosylation of platelet actin. Treatment of intact chicken embryo cells with botulinum C2 toxin decreased subsequent ADP-ribosylation of actin in cell lysates by perfringens iota or botulinum C2 toxin. In contrast to botulinum C2 toxin, perfringens iota toxin ADP-ribosylated skeletal muscle alpha-actin with a potency and efficiency similar to non-muscle actin. ADP-ribosylation of purified skeletal muscle and non-muscle actin by perfringens iota toxin led to a dose-dependent impairment of the ability of actin to polymerize.     

The Giardia lamblia actin gene and the phylogeny of eukaryotes

The single-copy actin gene of Giardia lamblia lacks introns; it has an average of 58% amino acid identity with the actin of other species; and 49 of its amino acids can be aligned with the amino acids of a consensus sequence of heat shock protein 70. Analysis of the potential RNA secondary structure in the transcribed region of the G. lamblia actin gene and of the single-copy actin gene of nine other species did not reveal any conserved structures. The G. lamblia actin sequence was used to root the phylogenetic trees based on 65 actin protein sequences from 43 species. This tree is congruent with small-subunit rRNA trees in that it shows that oomycetes are not related to higher fungi; that kinetoplatid protozoans, green plants, fungi and animals are monophyletic groups; and that the animal and fungal lineages share a more recent common ancestor than either does with the plant lineage. In contrast to smalls-ubunit rRNA trees, this tree shows that slime molds diverged after the plant lineage. The slower rate of evolution of actin genes of slime molds relative to those of plants, fungi, and animals species might be responsible for this incongruent branching. Correspondence to: G. Drouin     

Muscle-like Arrays in a Fibroblast Line

SEVERAL investigators have speculated that the basis for all cellular contractile activity resides in a common molecular mechanism involving an interaction between actin and myosin^1–4. Thin filaments resembling the actin filaments of muscle have indeed been widely observed^3–5 and the recent demonstrations of heavy meromyosin binding to thin filaments^4–6 suggest that these ubiquitous filaments are, in fact, actin. Although muscle-like thick filaments have not been observed in non-muscle cells, myosin thick filaments have been reconstituted from blood platelet preparations¹. To our knowledge, however, no one has presented evidence for the natural occurrence of ordered arrays of thick and thin filaments in non-muscle cells.     

Identification and characterization of GIV, a novel Galpha i/s-interacting protein found on COPI, endoplasmic reticulum-Golgi transport vesicles

In this report, we characterize GIV (Galpha-interacting vesicle-associated protein), a novel protein that binds members of the Galpha(i) and Galpha subfamilies of heterotrimeric G proteins. The Galpha(s) interaction site was mapped to an 83-amino acid region of GIV that is enriched in highly charged amino acids. BLAST searches revealed two additional mammalian family members, Daple and an uncharacterized protein, FLJ00354. These family members share the highest homology at the Galpha binding domain, are homologous at the N terminus and central coiled coil domain but diverge at the C terminus. Using affinity-purified IgG made against two different regions of the protein, we localized GIV to COPI, endoplasmic reticulum (ER)-Golgi transport vesicles concentrated in the Golgi region in GH3 pituitary cells and COS7 cells. Identification as COPI vesicles was based on colocalization with beta-COP, a marker for these vesicles. GIV also codistributes in the Golgi region with endogenous calnuc and the KDEL receptor, which are cis Golgi markers and with Galpha(i3)-yellow fluorescent protein expressed in COS7 cells. By immunoelectron microscopy, GIV colocalizes with beta-COP and Galpha(i3) on vesicles found in close proximity to ER exit sites and to cis Golgi cisternae. In cell fractions prepared from rat liver, GIV is concentrated in a carrier vesicle fraction (CV2) enriched in ER-Golgi transport vesicles. beta-COP and several Galpha subunits (Galpha(i1-3), Galpha(s)) are also most enriched in CV2. Our results demonstrate the existence of a novel Galpha-interacting protein associated with COPI transport vesicles that may play a role in Galpha-mediated effects on vesicle trafficking within the Golgi and/or between the ER and the Golgi.     

Evidence for a gelsolin-rich, labile F-actin pool in human polymorphonuclear leukocytes.

Filamentous (F) actin is a major cytoskeletal element in polymorphonuclear leukocytes (PMNs) and other non-muscle cells. Exposure of PMNs to agonists causes polymerization of monomeric (G) actin to F-actin and activates motile responses. In vitro, all purified F-actin is identical. However, in vivo, the presence of multiple, diverse actin regulatory and binding proteins suggests that all F-actin within cells may not be identical. Typically, F-actin in cells is measured by either NBDphallacidin binding or as cytoskeletal associated actin in Triton-extracted cells. To determine whether the two measures of F-actin in PMNs, NBDphallacidin binding and cytoskeletal associated actin, are equivalent, a qualitative and quantitative comparison of the F-actin in basal, non-adherent endotoxin-free PMNs measured by both techniques was performed. F-actin as NBDphallacidin binding and cytoskeletal associated actin was measured in cells fixed with formaldehyde prior to cell lysis and fluorescent staining (PreFix), or in cells lysed with Triton prior to fixation (PostFix). By both techniques, F-actin in PreFix cells is higher than in PostFix cells (54.25 +/- 3.77 vs. 23.5 +/- 3.7 measured as mean fluorescent channel by NBDphallacidin binding and 70.3 +/- 3.5% vs. 47.2 +/- 3.6% of total cellular actin measured as cytoskeletal associated actin). These results show that in PMNs, Triton exposure releases a labile F-actin pool from basal cells while a stable F-actin pool is resistant to Triton exposure. Further characterizations of the distinct labile and stable F-actin pools utilizing NBDphallacidin binding, ultracentrifugation, and electron microscopy demonstrate the actin released with the labile pool is lost as filament. The subcellular localization of F-actin in the two pools is documented by fluorescent microscopy, while the distribution of the actin regulatory protein gelsolin is characterized by immunoblots with anti-gelsolin. Our studies show that at least two distinct F-actin pools coexist in endotoxin-free, basal PMNs in suspension: 1) a stable F-actin pool which is a minority of total cellular F-actin, Triton insoluble, resistant to depolymerization at 4 degrees C, gelsolin-poor, and localized to submembranous areas of the cell; and 2) a labile F-actin pool which is the majority of total cellular F-actin, Triton soluble, depolymerizes at 4 degrees C, is gelsolin-rich, and distributed diffusely throughout the cell. The results suggest that the two pools may subserve unique cytoskeletal functions within PMNs, and should be carefully considered in efforts to elucidate the mechanisms which regulate actin polymerization and depolymerization in non-muscle cells.