F-actin Sequesters Elongation Factor from Interaction with Aminoacyl-tRNA in a pH-dependent Reaction

The machinery of eukaryotic protein synthesis is found in association with the actin cytoskeleton. A major component of this translational apparatus, which is involved in the shuttling of aa-tRNA, is the actinbinding protein elongation factor let (EF-let). To investigate the consequences for translation of the interaction of EF-let with F-actin, we have studied the effect of F-actin on the ability of EF-let to bind to aa-tRNA. We demonstrate that binding of EF-let:GTP to aatRNA is not pH sensitive with a constant binding affinity of ~0.2 ~M over the physiological range of pH. However, the sharp pH dependence of binding of EF-let to F-actin is sufficient to shift the binding of EFlet from F-actin to aa-tRNA as pH increases. The ability of EF-let to bind either F-actin or aa-tRNA in competition binding experiments is also consistent with the observation that EF-let's binding to F-actin and aatRNA is mutually exclusive. Two pH-sensitive actinbinding sequences in EF-let are identified and are predicted to overlap with the aa-tRNA-binding sites. Our results suggest that pH-regulated recruitment and release of EF-let from actin filaments in vivo will supply a high local concentration of EF-let to facilitate polypeptide elongation by the F-actin-associated translational apparatus. I N the current model of the elongation cycle of eukaryotic protein translation, elongation factor let (EF-let) t plays a role in transporting aminoacyl-tRNA to the ribosome during protein synthesis. Binding of EF-let with the nucleotide exchange factors EF-113~/leads to the replacement of GDP with GTP, which switches on the ability of EF-let to interact with aminoacyl-tRNA. Subsequently, the binding of EF-let:GTP:aa-tRNA ternary complex with the ribosome triggers the GTPase activity on EF-let and the resultant EF-let:GDP dissociates from the ribosome, ready for the next cycle (Riis et al., 1990). EF-let is a ubiquitous protein with homologues (EF-Tu) in prokaryotic systems. It is a very abundant protein that constitutes about 1-2% of the total protein in normal growing cells. Large increases in mRNA levels for EF-let are observed in rapidly proliferating cultured cells, embryos, and a variety of human tumors, suggesting a correlation of EF-let expression level with the rate of cell growth and proliferation (for review see Condeelis, 1995). The first evidence that EF-let is an actin-binding protein Address all correspondence to Gang Liu, Department of Anatomy and Structural Biology, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, NY 10461. Tel.: (718) 430o4113. Fax: (718) 430-8996. 1. Abbreviat ions used in this paper. EF-let, elongation factor la; EtBr, ethidium bromide; GST, glutathione-S-transferase; 3-D, three-dimensional. was obtained in Dictyostelium (Demma et al., 1990; Yang et al., 1990). Subsequently, EF-let has been shown to colocalize with actin filaments and this colocalization changes with chemoattractant stimulation in Dictyostelium and adenocarcinoma cells (Dharmawardhane et al., 1991; Okazaki and Yumura, 1995; Edmonds et al., 1996). In fibroblasts, EF-let is found to colocalize at actin filament junctions and EF-let from carrot root cells bundles actin filaments (Yang et al., 1993; Bassell et al., 1994). Owen et al. (1992) demonstrated that EF-let cross-links F-actin into bundles with a unique cross-bridge bonding rule that would tend to exclude other actin cross-linking proteins. This unique cross-bridge structure may represent a special property of EF-let that is important in the stability of the cytoskeleton and the transport, anchorage, and translation of mRNA (Condeelis, 1995). Apart from binding to actin filaments, binding to calmodulin, bundling, and/or severing of microtubules by EF-let from carrot, Trypanosome, Xenopus, rabbit liver, and human (recombinant) have been reported (Durso and Cyr, 1994a; Kaur and Ruben, 1994; Shiina et al., 1994). In addition to EF-let, an increasing number of protein synthesis components have been observed to associate with the cytoskeleton. The association of mRNA with the cytoskeleton has been well documented (for review see St Johnston, 1995), and there is correlation between this association and protein synthesis (see Nielsen et al., 1983; Singer, 1993). In addition, ribosomes and initiation factor 2 (elF-2) have been shown to associate with the cytoskele© The Rockefeller University Press, 0021-9525/96/11/953/11 $2.00 The Journal of Cell Biology, Volume 135, Number 4, November 1996 953-963 953 on N ovem er 7, 2017 jcb.rress.org D ow nladed fom ton (Howe and Hershey, 1984; Gavrilova et al., 1987; Zambetti et al., 1990; Hamill et al., 1994; Hesketh et al., 1991). Interestingly, the other elongation factor (EF-2) has been demonstrated to bind directly to actin filaments (Bektas et al., 1994). Colocalization of these components with the cytoskeleton supports the speculation that protein synthesis in vivo is channeled, i.e., the components are organized in a high degree of spatial order and intermediates are transferred from one enzyme to another without mixing with the surrounding cytoplasm (Stapulionis and Deutscher, 1995). A correlation between cytoplasmic alkalinization and increases in protein synthesis has been observed in a number of different cell types (for review see Grinstein et al., 1989). In sea urchins, elevation of intracellular pH serves as a primary signal in the activation of protein synthesis at fertilization (Winkler et al., 1980). Measurements of cytosolic pH in sea urchin eggs before fertilization indicate that protein synthesis is restricted below pH 6.8 but not at pH 7.1 (Rees et al., 1995). In fibroblasts, intracellular pH may play a determinant role in the control of cell division by controlling the rate of protein synthesis (Chambard and Pouyssegur, 1986). In Dictyostelium, stimulation of cells with cAMP induces cytoplasmic alkalinization, and artificially raising the intracellular pH can trigger a severalfold increase in the rate of DNA and protein synthesis (Aerts et al., 1985, 1987). The interaction of Dictyostelium EF-Iot with F-actin is pH-dependent with a transition from tight to loose bundling between pH 6.7 and 7.6 (Edmonds et al., 1995). It has been proposed that pH may regulate the association of EFla with the cytoskeleton in such a way as to regulate, both spatially and temporally, its activity as an elongation factor (Liu et al., 1996). This is potentially important for developing organisms like Dictyostelium, in which, during early development, the mean cytoplasmic pH can range from 6.0 to 7.2 (Furukawa et al., 1990). To understand the physiological significance of the interaction of EF-let with actin filaments, we investigated the interaction of EF-la with F-actin and aa-tRNA in vitro. We demonstrate that the abilities of EF-let to bundle and bind to F-actin are blocked by the formation of EF-let: GTP:Phe-tRNA ternary complex in a pH-dependent manner. To understand the mechanism of the blockade, we chose to map the F-actin-binding sites on EF-lec Using truncated recombinants of EF-let, we have identified two F-actin-binding domains that exhibit different pH sensitivities for F-actin binding. Structural comparison by using EF-Tu:GTP:Phe-tRNA complex as a model (Nissen et al., 1995) suggests that the proposed F-actin-binding domains on EF-let may overlap with those for the EF-la/Phe-tRNA interaction. These observations provide clues in explaining how pH may, by modulating the interaction of EFla with F-actin, influence the dynamics of the cytoskeleton and the rate of protein translation in the cells. Materials and Methods Construction of Expression Vectors for Glutathione-STransferase ( GST)-EFl a Fusion Proteins Full-length Dictyostelium EF-la cDNA sequence (Yang et al., 1990) was subcloned into pGEX-KG vector (Guan and Dixon, 1991) at NcoI and XhoI sites. Construct pGEX-Dd-dmI, encoding amino acids 1-221, was generated by PCR from Dictyostelium EF-lct eDNA with primers GGC GGA ATT CTA ATG GAA t t t CCG GAA TCC GAA AAA ACA CAT and GCG AAG CTT ATF CTA ATA AAG TTG GAC C-TIT and inserting the PCR product into pGEX-KG vector at EcoRI and HindIII sites. Similarly, construct pGEX-Dd-dmII (encoding amino acids 222-320 of Dictyostelium EF-lct) was generated with primers CGC GGA ATT CTA GCC CTC GAT GCC ATC GTC GAA and CGCA AGC TFA GGC GTC ACC AGC GAC CAT, and construct pGEX-Dd-dmIII (encoding amino acids 321-456 of Dictyostelium EF-let) was generated with primers AGC GGA ATT CTA AAA AAC GAT CCA CCA CAA GAA and CG CGA AGC TTA T I T CTT CTT TGA TGG AGC AGC. A construct of mouse EF-lct (Lu and Werner, 1989) in pGEX-KG vector was a gift from Dr. E. Richard Stanley (Albert Einstein College of Medicine). Construct pGEX-mouse~dmI (encoding amino acids 1-230 of mouse EFla) was made by taking advantage of a HindIII site on the mouse EF-la sequence near residue 230 to remove the coding sequence for amino acids 231-461 from the construct pGEX-mouse-EF-lct and religating the rest of the construct. All the constructs were validated by DNA sequencing and Western blotting with antibodies against EF-hx. Protein Purification Dictyostelium EF-lu was purified as previously described (Edmonds et al., 1995). Rabbit skeletal muscle actin was prepared from acetone powder by the method of Spudich and Watt (1971) and further purified by G-150 gel filtration (Bresnick et al., 1990). Dictyostelium actin was isolated and purified by the method of Bresnick and Condeelis (1990). The GST fusion proteins of EF-lcts and their truncates were expressed and purified by the method modified from Smith and Johnson (1988). In brief, host cells (XL1-Blue or JM109) containing the desired construct were allowed to grow overnight in LB medium with 100 p~g/ml ampicillin at 30°C. When cell density reached OD600 = 1, the expression of fusion protein was induced by addition of 0.1~).5 mM of IPTG for 4 to 6 h at 30°C (depending on which fusion protein was induced). At the end of induction, the cells were harvested and cell pellet was washed once with wash buffer (10 mM Tris, 1 mM DTT, pH 7.5) and then resuspended with lysis buffer (20 mM NaPO4, 150 mM NaCI, 20 ~g/ml leupeptin, 20 ~g/ml pepstatin A, 20 p~g/ml chymostatin, 3% [vol/vol] aprotinin, 1 mM DTT and 1 mM EDTA, 1% Triton X-100, pH 8.0). After sonication and centrifugation at 50,000 g for 30 min, the supernatant was incubated with glutathione (GSH)-conjugated beads at room temperature for 30 min. The beads were washed with PBS (pH 8.0) and bound GST fusion proteins were eluted with elution buffer (10 mM GSH, 200 mM NaCI, 120 mM Tris, pH 9.0). Right Angle Light Scattering to Study EF-l a Cross-linking of F-actin The loading of GTP to EF-lct was performed by incubating 1 I~M EF-I~ with 1 mM GTP for 30 min at room temperature. Nucleotide binding was confirmed by nitrocellulose filtration assay (Nagata et al., 1976) or MantGTP fluorescence. Right angle light scattering was used to study the EFltx-mediated formation of F-actin bundles. In an assay buffer containing 20 mM Pipes, 50 mM KCI, 5 mM MgC12, 2 mM EGTA, 1 mM DTF, 1 mM ATP, and 15% glycerol, preformed Dictyostelium F-actin (3 p~M) was mixed with Dictyostelium EF-I~ (1 ~M) that was incubated with 1 mM GTP for 30 min and then 1 ~M [3H]Phe-tRNA for an additional 20-30 min to form ternary complex at room temperature. The assays were performed by using a fluorescence spectrophotometer (model F-2000; Hitachi Sci. Instrs., Mountain View, CA) with 600-nm excitation and emission wavelength at a band pass of 5 nm. Data were collected and analyzed by using the computer software SpectraCalc and GRAMS/386 (Galactic Industries Corp., Salem, NH). After light scattering analysis, the reaction mixtures were collected for actin cosedimentation assay. Actin Cosedimentation Assay Actin cosedimentation assay was used to test the actin-binding activity of the fusion proteins of EF-lcc Each fusion protein was mixed with G-actin in sedimentation buffer (20 mM Pipes, 50 mM KC1, 2 mM EGTA, 2 mM MgC12, 1 mM DTT, 1 mM ATP) at preset pH and then incubated at 0-4°C for 18-20 h. This buffer contains physiological concentrations of monovalent salts that have been measured in amebae as ~50 mM (Marin and Rothman, 1980). The reaction mixture was centrifuged and samples of suThe Journal of Cell Biology, Volume 135, 1996 954 on N ovem er 7, 2017 jcb.rress.org D ow nladed fom pernatants and pellets were analyzed by SDS-PAGE and densitometry. Fusion proteins were soluble under these assay conditions in the absence of F-actin. A differential actin sedimentation assay was applied to study the effect of Phe-tRNA on EF-let bundling and binding to F-actin. Samples collected after light scattering assays were allowed to incubate at room temperature for 2 h. The samples then were centrifuged by using an airfuge at 50,000 g for 2.5 min to pellet F-actin bundles (low speed pellet), and the supernatants were transferred and further centrifuged at 130,000 g for 40 rain to pellet single actin filaments (high speed pellet) as demonstrated previously (Demma et al., 1990; Edmonds et al., 1995). Aliquots of reaction mixture, supernatants, and pellets were quantified by SDS-PAGE for protein contents, ethidium bromide (EtBr) fluorescence for tRNA contents, and liquid scintillation counting for [3H]Phe-tRNA.