Subunit Assembly and Guanine Nucleotide Exchange Activity of Eukaryotic Initiation Factor-2B Expressed in Sf9 Cells*

(Received for publication, January 3, 1997)

John R. Fabian , Scot R. Kimball , Nina K. Heinzinger and Leonard S. Jefferson Dagger

From the Department of Cellular and Molecular Physiology, Pennsylvania State University College of Medicine, Hershey, Pennsylvania 17033

ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENT
REFERENCES


ABSTRACT

Eukaryotic initiation factor-2B (eIF-2B) is a guanine nucleotide exchange factor (GEF) that plays a key role in the regulation of protein synthesis. In this study, we have used the baculovirus-infected Sf9 insect cell system to express and characterize the five dissimilar subunits of rat eIF-2B. GEF activity was detected in extracts of Sf9 cells expressing the epsilon -subunit alone and was greatly increased when all five subunits were coexpressed. In addition, high GEF activity was observed in extracts containing a four-subunit complex lacking the alpha -subunit. Assembly of an eIF-2B holoprotein was confirmed by coimmunoprecipitation of all five subunits. Gel filtration chromatography revealed that recombinant eIF-2B had the same molecular mass as eIF-2B purified from rat liver and that it did indeed possess GEF activity. Phosphorylation of the substrate eIF-2 inhibited the GEF activity of the five-subunit eIF-2B; this inhibition required the eIF-2B alpha -subunit. The results demonstrate that eIF-2Balpha functions as a regulatory subunit that is not required for GEF activity, but instead mediates the regulation of eIF-2B by substrate phosphorylation. Furthermore, eIF-2Bepsilon is necessary and is perhaps sufficient for GEF activity in vitro.


INTRODUCTION

Protein synthesis in eukaryotic cells is regulated in response to growth factors, hormones, and changes in the external environment (reviewed in Refs. 1 and 2). One of the major control points for protein synthesis is the initiation of mRNA translation, which is mediated by a number of eukaryotic initiation factors (eIFs)1 (recently reviewed in Ref. 3). Translation of mRNA begins with the binding of initiator Met-tRNAi to the 40 S ribosomal subunit and is mediated by eIF-2 as part of the eIF-2·GTP·Met-tRNAi ternary complex. During the initiation process, the GTP bound to eIF-2 is hydrolyzed, and a binary complex consisting of eIF-2 and GDP is released from the 80 S initiation complex. Since eIF-2 has a 100-400-fold higher affinity for GDP than for GTP, the guanine nucleotide exchange factor (GEF) known as eIF-2B is required to regenerate the GTP-bound form of eIF-2, which can then participate in another cycle of translation initiation.

In cells subjected to a variety of stresses, e.g. heat shock, serum and amino acid deprivation, and heme deprivation in reticulocytes, protein synthesis is inhibited by a mechanism involving phosphorylation of the alpha -subunit of eIF-2 (reviewed in Ref. 4). Phosphorylated eIF-2 efficiently binds to eIF-2B, but cannot undergo nucleotide exchange and therefore competitively inhibits the GEF function of eIF-2B, which is present in limiting amounts in all cell types examined (5-7). In addition, eIF-2B may be regulated by other mechanisms since its activity is modulated under certain conditions in the absence of a change in phosphorylation of eIF-2alpha (8-12). The epsilon -subunit of eIF-2B is phosphorylated in vitro by glycogen synthase kinase-3, casein kinase-1, and casein kinase-2, which may alter the GEF activity of the holoprotein (13-16). In addition, the activity of eIF-2B appears to be regulated by allosteric effectors, e.g. NADP+, NAD+, and ATP inhibit guanine nucleotide exchange, whereas NADPH, NADH, and polyamines activate exchange (17-20).

Most GEFs, such as those stimulating GDP dissociation from the Ras superfamily of proteins, exist as single subunit proteins (21, 22). In contrast, eIF-2B is a complex protein consisting of five dissimilar subunits. The cDNAs for all five subunits of eIF-2B have been cloned from both mammals (23-26) and yeast (27). Alignment of the coding regions of mammalian and yeast subunits indicates sequence homology for the mammalian alpha -, beta -, gamma -, delta -, and epsilon -subunits with the yeast GCN3, GCD7, GCD1, GCD2, and GCD6 subunits, respectively. Genetic studies in yeast indicate that deletion of the alpha -subunit of eIF-2B (GCN3) does not affect viability and therefore is not an essential component of the GCD complex under normal growth conditions (28). However, there is controversy over the requirement of the alpha -subunit for eIF-2B function since a recent report has suggested that purified rabbit eIF-2B lacking the alpha -subunit does not have GEF activity (29).

The current model for translation initiation in eukaryotic cells is formulated from biochemical studies of cell-free systems derived from rabbit reticulocytes and from genetic experiments conducted in yeast. In the study reported here, we set out to determine if the baculovirus expression system could be used as a new approach to study the structure/function relationship of translation initiation factors, many of which are composed of multiple subunits with unknown functions. For example, despite the fact that the mammalian eIF-2B complex has been available in purified form for 15 years and all five yeast and mammalian eIF-2B subunits have now been cloned, relatively little is known about the function of individual subunits or the mechanism of nucleotide exchange. Here we report our studies on the expression and coexpression of rat eIF-2B subunits in Sf9 insect cells using recombinant baculoviruses and the characterization of GEF activity in extracts of cells expressing these proteins.


MATERIALS AND METHODS

Antibodies

Mouse monoclonal antibodies against the gamma - and epsilon -subunits of eIF-2B were produced using purified protein for antigen as described previously (30). For immunoblot analysis, a monoclonal antibody against the gamma -subunit of eIF-2B was used in conjunction with a rabbit anti-FLAG polyclonal antibody (Santa Cruz Biotechnology) or a mouse anti-FLAG M2 monoclonal antibody (Kodak Scientific Imaging Systems) to detect the eIF-2B gamma -subunit and the alpha -, beta -, delta -, and epsilon -subunits of eIF-2B harboring an amino-terminal FLAG marker octapeptide (DYKDDDDK), respectively.

Construction of Baculovirus Transfer Vectors

The full-length clone for the rat eIF-2Bbeta (GenBankTM/EMBL accession number U83914[GenBank]) used in this study was obtained by screening the Stratagene rat brain Uni ZAPXR cDNA library (catalogue number 936515) using the rabbit eIF-2Bbeta cDNA (31) as a probe (kindly provided by Dr. C. G. Proud, University of Kent, Canterbury, United Kingdom). The complementary DNA clones encoding rat eIF-2B alpha -, gamma -, delta -, and epsilon -subunits were previously described (23-26), and their GenBankTM/EMBL accession numbers are U05821[GenBank], Z48225[GenBank], U38253[GenBank], and U19511[GenBank], respectively.

To facilitate identification and quantitation of rat eIF-2B subunits in Sf9 cells, sequences encoding a common 8-amino acid FLAG epitope (DYKDDDDK) were added to the cDNAs encoding the alpha -, beta -, delta -, and epsilon -subunits by polymerase chain reaction using the Pfu DNA polymerase (Stratagene). The polymerase chain reaction products were cloned into the SrfI site of the plasmid pCR-Script Cam SK(+) according to instructions supplied by the manufacturer (Stratagene). The eIF-2Bgamma cDNA had an internal BamHI site that was removed using the QuikChange site-directed mutagenesis kit (Stratagene). The primers used to remove the BamHI site resulted in a single nucleotide substitution that did not alter the amino acid sequence of the protein. The identities of all clones were confirmed by sequence analysis of the 5'- and 3'-regions using the reverse primer and T7 primers, which bind to the polylinker region of the pCR-Script plasmid. The full-length eIF-2Balpha , eIF-2Bbeta , eIF-2Bgamma , and eIF-2Bdelta cDNAs were next subcloned out of the pCR-Script plasmid and into the baculovirus transfer vector pAcUW51 (Pharmingen) containing the baculovirus p10 and polyhedrin promoters. Specifically, the cDNAs for the alpha - and delta -subunits containing the FLAG epitope were sequentially subcloned into the BamHI and BglII sites, respectively, of pAcUW51. The cDNAs for the gamma - and delta -subunits were sequentially subcloned into the BamHI and BglII sites, respectively, of pAcUW51. A baculovirus transfer vector encoding the FLAG-tagged eIF-2Bepsilon cDNA was constructed by isolating an NcoI/Acc65I fragment encoding the full-length cDNA fragment from pCR-Script and then ligating the fragment into the corresponding sites downstream of the polyhedrin promoter of the pAcSG2 transfer vector (Pharmingen). A baculovirus transfer vector that expressed the FLAG-tagged delta -subunit alone was constructed by subcloning the eIF-2Bdelta cDNA into the BglII site of pAcUW51.

Expression of eIF-2B Subunits in Sf9 Cells

Sf9 insect cells (American Type Culture Collection CRL-1711) were placed in culture using Grace's insect medium (Life Technologies, Inc.) supplemented with 10% fetal calf serum (Hyclone Laboratories), 0.1% Pluronic F-68, 2 mM glutamine, and 20 µg/ml gentamicin. To generate recombinant baculoviruses expressing the combinations of eIF-2B subunits described above, BaculoGold Autographa californica nuclear polyhedrosis virus DNA (BaculoGold AcNPV, Pharmingen) and each of the recombinant transfer vectors were cotransfected into Sf9 cells, and recombinant baculoviruses were isolated by plaque purification as described previously (32). Recombinant plaques expressing eIF-2B subunits were identified by infecting 2 × 106 cells with the plaque supernatant. Cell lysates were prepared 72 h later and screened for protein production by immunoblot analysis.

For expression of eIF-2B subunits, 6 × 106 Sf9 insect cells were plated in 100-mm dishes. The Sf9 cells were then infected with recombinant baculovirus(es) at a multiplicity of infection of 10 in a reduced volume of 3.0 ml for 1 h, fed with 11 ml of medium, and then lysed at 60 h post-infection. As a control, Sf9 insect cells were also infected with wild-type AcNPV baculovirus.

Immunoprecipitation and Immunoblotting

For immunoprecipitation assays, 6 × 106 infected cells were washed once in ice-cold phosphate-buffered saline and then lysed in 700 µl of lysis buffer consisting of 1% Nonidet P-40, 20 mM Tris, pH 8.0, 137 mM NaCl, 10% glycerol, 2 mM EDTA, 1 mM phenymethylsulfonyl fluoride, 0.15 unit/ml aprotinin, and 20 mM leupeptin. Cells were vortexed in lysis buffer and incubated on ice for 15 min. Insoluble material was removed by centrifugation at 4 °C for 10 min at 16,000 × g. Immunoprecipitation assays were performed by incubating lysates with 250 µl of a monoclonal antibody against eIF-2Bepsilon at 4 °C overnight. Goat anti-mouse Biomag IgG beads (PerSeptive Diagnostics) were used to collect the antigen-antibody complexes. Prior to use, the beads were washed in 1% nonfat dry milk in Buffer A (20 mM Tris-HCl, pH 7.4, 5 mM EDTA, 0.04% beta -mercaptoethanol, 0.5% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, 50 mM NaF, 50 mM beta -glycerophosphate, 0.1 mM phenymethylsulfonyl fluoride, 1 mM benzamidine, and 0.5 mM sodium vanadate) containing 150 mM NaCl. The beads were then captured using a magnetic stand and washed twice with Buffer A containing 50 mM NaCl and once with Buffer A containing 150 mM NaCl. Protein bound to the beads was eluted in 100 µl of SDS sample buffer, and the sample was incubated at 100 °C for 5 min. The beads were collected by centrifugation, and proteins in the sample were separated by electrophoresis as described below.

Immunoprecipitates of cell lysates prepared as described above were resolved by SDS-polyacrylamide gel electrophoresis (PAGE) and were electrophoretically transferred to 0.2-µm pore size nitrocellulose membranes (Bio-Rad). Membranes were then blocked with 2% bovine serum albumin (Fraction V) in Tris-buffered saline, pH 8.0, for 1 h; washed in TBST (Tris-buffered saline, pH 8.0, containing 0.2% Tween 20); and probed overnight at 4 °C with primary antibody diluted in TBST. The rabbit and mouse anti-FLAG antibodies were diluted 1:2500 (0.04 µg/ml final concentration) and 1:2000 (1.5 µg/ml final concentration), respectively. The filters were then washed in TBST, probed for 1 h with a horseradish peroxidase-coupled secondary antibody (Amersham Corp.) diluted 1:20,000 in TBST, washed again in TBST, and developed using enhanced chemiluminescence (Amersham Corp.). To detect the eIF-2B gamma -subunit, the blots were reprobed with a monoclonal antibody against rat eIF-2Bgamma diluted 1:100, reprobed with secondary antibody, and developed as described above.

Measurement of eIF-2B Activity

Infected Sf9 cells (6 × 106) were washed once in ice-cold phosphate-buffered saline and lysed in 1.0 ml of ice-cold extraction buffer (45 mM HEPES, pH 7.4, 0.375 mM magnesium acetate, 0.075 mM EDTA, 95 mM potassium acetate, 2.5 mg/ml digitonin, and 10% (v/v) glycerol). The lysates were clarified by centrifugation at 10,000 × g for 10 min at 4 °C. The resulting supernatants were immediately assayed for eIF-2B activity as determined by the exchange of [3H]GDP bound to eIF-2 for unlabeled GDP as described previously (33). The labeled binary complex eIF-2·[3H]GDP was prepared by incubating tubes containing rat liver eIF-2 (~95% pure (34)) and [3H]GDP (2.5 mM, 10.9 Ci/mmol) in 80 µl of assay buffer (62.5 mM MOPS, pH 7.4, 125 mM KCl, 1.25 mM dithiothreitol, and 0.2 mg/ml bovine serum albumin) at 30 °C for 10 min. The Mg2+ concentration was adjusted to 2 mM, and the binary complex was stored on ice before use.

To measure eIF-2B activity, assay buffer containing a 100-fold excess of GDP, 1.25-40 µl of cell lysate, and 2 mM Mg2+ was added to a tube, followed by 1-2 pmol of labeled binary complex. The mixture was then incubated at 30 °C for 0, 2, 4, or 6 min. The exchange reaction was measured as a decrease in the eIF-2-mediated binding of [3H]GDP to nitrocellulose filters with time.

eIF-2alpha Phosphorylation and GEF Activity

For the experiments shown in Fig. 6, eIF-2 was phosphorylated using the eIF-2alpha kinase HCR by incubating 7.5 µg of pure eIF-2 in 37.5 µl of HCR kinase buffer (20 mM Tris-HCl, pH 7.4, 100 mM KCl, 10% (v/v) glycerol, 0.1 mM EDTA, 1 mM dithiothreitol, 2.5 mM magnesium chloride, and 0.1 mM ATP) with 7.5 µl (~0.5 µg) of purified rabbit reticulocyte HCR (35). The mixture was incubated at 37 °C for 30 min, and the reaction was terminated by placing the tube on ice. An equal volume of kinase assay buffer containing the phosphatase inhibitors NaF (100 µM final concentration) and microcystin (2 µM final concentration) was added to each tube. A similar reaction containing eIF-2 without HCR was processed identically for use as the unphosphorylated eIF-2 control. The phosphorylation state of eIF-2alpha was monitored by isoelectric focusing gel electrophoresis and immunoblot analysis. To measure eIF-2B activity, assay buffer containing phosphorylated or unphosphorylated eIF-2, a 100-fold excess of GDP, 1.25-40 µl of cell lysate, and 2 mM Mg2+ was added to a plastic tube and incubated at 30 °C for 2 min. The eIF-2·[3H]GDP complex (1-2 pmol) was then added to each tube, and the mixture was incubated at 30 °C. Aliquots of the reaction mixture were removed from the tubes at 0, 2, 4, or 6 min and filtered through nitrocellulose filters as described above.


Fig. 6. Effect of eIF-2alpha phosphorylation on eIF-2B activity. eIF-2 purified from rat liver was phosphorylated using the eIF-2alpha kinase HCR in an in vitro reaction. A similar reaction containing eIF-2 without HCR was processed identically for use as the unphosphorylated eIF-2 control. The phosphorylated (open symbols) and control (closed symbols) eIF-2 samples were added to eIF-2B assay buffer containing extracts of cells expressing all five subunits of eIF-2B (bullet , open circle ); cells expressing the beta -, gamma -, delta -, and epsilon -subunits (black-square, square ); or cells infected with wild-type virus (black-triangle, triangle ). The mixture was incubated at 30 °C for 2 min. The reaction was then initiated by addition of the eIF-2·[3H]GDP binary complex. Expression of eIF-2B subunits was monitored by immunoblot analysis as described in the legend to Fig. 1 and is shown in the inset. The positions of the subunits of eIF-2B are noted to the right of the inset. Lane 1, extract of Sf9 cells infected with wild-type AcNPV baculovirus; lane 2, extract of Sf9 cells coexpressing the beta -, gamma -, delta -, and epsilon -subunits of eIF-2B; lane 3, extract of cells expressing all five eIF-2B subunits.
[View Larger Version of this Image (28K GIF file)]

Gel Filtration of eIF-2B Complexes

Infected Sf9 cells (6 × 106) were washed once with ice-cold phosphate-buffered saline and then lysed in 0.25 ml of ice-cold buffer consisting of 45 mM HEPES, pH 7.4, 0.375 mM magnesium acetate, 0.075 mM EDTA, 95 mM potassium acetate, 2.5 mg/ml digitonin, and 10% (v/v) glycerol. The extracts were clarified by centrifugation at 10,000 × g for 10 min at 4 °C and loaded onto a Superose 6 gel filtration column (Pharmacia Biotech Inc.) pre-equilibrated with buffer consisting of 20 mM HEPES, pH 7.6, 0.1 mM EDTA, 1 mM dithiothreitol, 10% glycerol, and 150 mM KCl. Fractions (0.5 ml) were collected, and the absorbance at 280 nm was measured. Aliquots (0.02 ml) of each fraction containing protein were assayed for GEF activity as described above. To confirm the presence of eIF-2B subunits in fractions, 25 µl of 2 × SDS sample buffer were added to 25-µl aliquots of each fraction, and 15 µl of the mixture were resolved by SDS-PAGE and visualized by immunoblotting with anti-FLAG and anti-eIF-2Bgamma antibodies as described above.


RESULTS

Expression of Rat eIF-2B Subunits in Sf9 Cells

To facilitate detection, a common epitope consisting of the FLAG octapeptide (DYKDDDDK) was added to the N terminus of the eIF-2B alpha -, beta -, delta -, and epsilon -subunits. Recombinant baculoviruses for each of the FLAG-tagged eIF-2B subunits were constructed so as to express the individual subunit cDNAs under the control of either the polyhedrin promoter or the p10 promoter. To reduce the number of baculoviruses needed to express all five subunits of eIF-2B in Sf9 cells, single baculoviruses encoding either the alpha - and delta - or beta - and gamma -subunits were generated using a transfer vector designed for multigene expression. Expression of the largest eIF-2B subunit (eIF-2Bepsilon ) was achieved using a transfer vector designed for single gene expression.

To investigate the relative expression/coexpression of eIF-2B subunits, Sf9 cells were (i) singly infected with baculoviruses encoding individual eIF-2B subunits or (ii) triply infected with baculoviruses encoding all five eIF-2B subunits. The recombinant eIF-2B proteins expressed in Sf9 cells 60 h post-infection were resolved by SDS-PAGE and identified by immunoblot analysis using anti-FLAG and anti-eIF-2Bgamma monoclonal antibodies. Similar amounts of expressed eIF-2B alpha -, beta -, delta -, and epsilon -subunit proteins were observed for Sf9 cells singly infected with baculoviruses encoding the epsilon -subunit (Fig. 1A, lane 1), the alpha - and delta -subunits (lane 2), and the beta - and gamma -subunits (lane 3). For Sf9 cells infected with all three viruses, coexpression of proteins for each of the five eIF-2B subunits was observed by immunoblot analysis (Fig. 1A, lane 4). However, in these experiments, the amount of eIF-2Bgamma produced could not be directly compared with the amount of the other four subunits because a different antibody was used to detect eIF-2Bgamma than was used to detect the other four subunits. To our knowledge, this is the first time a complex multisubunit translation initiation factor has been overexpressed using the baculovirus-infected Sf9 cell system.


Fig. 1. GEF activity in lysates from Sf9 cells expressing rat eIF-2B subunits. A, Sf9 cells were infected with baculoviruses encoding the epsilon -subunit (lane 1), the alpha - and delta -subunits (lane 2), and the beta - and gamma -subunits (lane 3) or were infected with a combination of viruses encoding all five subunits of eIF-2B (lane 4) as described under "Materials and Methods." Infected cells were lysed 60 h post-infection in buffer containing digitonin. Lysates were fractionated by SDS-PAGE on a 12.5% gel, transferred to nitrocellulose, and visualized by immunoblotting with a mouse monoclonal antibody against the FLAG epitope, followed by reprobing with a monoclonal antibody against the eIF-2B gamma -subunit. The positions of the individual subunits of eIF-2B are noted to the right of A. B, GEF activity was analyzed in lysates from the infected Sf9 cells described for A. Aliquots (40 µl) of lysates containing the epsilon -subunit (square ), the alpha - and delta -subunits (bullet ), the beta - and gamma -subunits (black-down-triangle ), or all five subunits of eIF-2B (black-square) were assayed for GEF activity, which was measured as a decrease in the eIF-2-mediated binding of [3H]GDP to nitrocellulose filters with time. GEF activity was also measured in extracts of cells infected with wild-type AcNPV baculovirus (open circle ) as a control. The results are expressed as a percentage of [3H]GDP bound to eIF-2 (eIF-2·[3H]GDP) at time 0. All infections were done in duplicate, and GEF activity is represented as an average value for three experiments.
[View Larger Version of this Image (23K GIF file)]

GEF Activity of Rat eIF-2B Subunits Expressed in Sf9 Cells

It was important to establish whether or not rat eIF-2B subunits expressed in Sf9 cells exhibited GEF activity. The GDP dissociation assay is a direct measurement of the GEF activity of eIF-2B since, in the presence of Mg2+, the rate-limiting step of the nucleotide exchange reaction is GDP dissociation (36). Cell lysates were prepared from Sf9 cells infected with viruses encoding pairs of subunits or the three viruses encoding all five subunits of eIF-2B. The extracts were then assayed for GEF activity by determining the percentage of [3H]GDP released from eIF-2·[3H]GDP binary complexes over a 6-min interval (Fig. 1B). A low basal level of GEF activity was detected in extracts of cells infected with wild-type AcNPV baculovirus or in extracts of cells infected with baculoviruses encoding either the alpha - and delta - or beta - and gamma -subunits of eIF-2B (Fig. 1B). Significantly more GEF activity was detected in lysates from cells expressing the eIF-2B epsilon -subunit (i.e. 40% of the bound [3H]GDP was released from eIF-2 over 6 min) compared with lysates from cells infected with wild-type AcNPV baculovirus. The amount of GEF activity detected in extracts of triply infected cells coexpressing all five eIF-2B subunits was markedly greater (~90% of [3H]GDP bound to eIF-2 was released) than that in extracts of cells expressing eIF-2Bepsilon alone, suggesting that a functional, five-subunit eIF-2B complex was formed in the cells.

As mentioned above, extracts of cells expressing the eIF-2B epsilon -subunit alone exhibited enhanced GEF activity compared with extracts of cells infected with wild-type virus. Therefore, we next determined if the epsilon -subunit was required for the GEF activity observed in extracts of cells expressing all five subunits of eIF-2B. When cells were doubly infected with viruses expressing a four-subunit complex lacking eIF-2Bepsilon , no GEF activity was observed (Fig. 2B). In contrast, the rate of GDP exchange observed in extracts of cells expressing a four-subunit complex lacking eIF-2Balpha was greatly enhanced compared with control extracts. These results demonstrate that the eIF-2B epsilon -subunit is required for the GEF activity of eIF-2B, whereas the eIF-2B alpha -subunit is not.


Fig. 2. GEF activity in lysates from Sf9 cells expressing four- or five-subunit rat eIF-2B. Shown are the immunoblot analysis (A) and GEF activity (B) of lysates from Sf9 cells coinfected with baculoviruses expressing four subunits of rat eIF-2B lacking the epsilon -subunit (lane 1; triangle ), four subunits of eIF-2B lacking the alpha -subunit (lane 2; square ), and all five subunits of rat eIF-2B (lane 3; black-triangle). Infected cells were lysed and analyzed by immunoblotting as described in the legend to Fig. 1. The positions of the subunits of eIF-2B are noted to the right of A. Aliquots (40 µl) of the cell extracts were assayed for GEF activity as described under "Materials and Methods," and the results are expressed as a percentage of the amount of [3H]GDP bound to eIF-2 (eIF-2·[3H]GDP) at time 0. In this experiment, the control cells were uninfected Sf9 cells (none). All infections were done in duplicate, and GEF activity is represented as an average value of three experiments.
[View Larger Version of this Image (17K GIF file)]

GEF Activity in Extracts of Cells Expressing the eIF-2B epsilon -Subunit Alone Relative to Extracts of Cells Expressing Multisubunit eIF-2B

The activity observed for the five-subunit eIF-2B complex in Figs. 1B and 2B was out of the linear range of the assay, and therefore, it was necessary to titrate lysates to determine the activity of the epsilon -subunit alone relative to the highly active five- and four-subunit eIF-2B complexes. The results show that 1.25 µl of lysate from cells expressing all five subunits of eIF-2B or only the beta -, gamma -, delta -, and epsilon -subunits yielded a rate of exchange similar to that observed for 40 µl of lysate from cells expressing the epsilon -subunit alone (Fig. 3). Because the amount of epsilon -subunit expressed in each case was similar, the results indicate that the five-subunit and four-subunit (beta , gamma , delta , and epsilon ) eIF-2B complexes were ~40-fold more active than the eIF-2B epsilon -subunit alone.


Fig. 3. Titration of GEF activity in lysates from Sf9 cells expressing all five subunits of rat eIF-2B, four subunits of eIF-2B lacking eIF-2Balpha , or the eIF-2B epsilon -subunit alone. GEF activity was measured in extracts of Sf9 cells coinfected with baculoviruses expressing all five rat eIF-2B subunits (A), four subunits of eIF-2B lacking the alpha -subunit (B), or the epsilon -subunit of eIF-2B alone (C). Infected cells were lysed 60 h post-infection, and GEF activity in 10 (open circle ), 5 (square ), 2.5 (triangle ), and 1.25 (down-triangle) µl of extract was quantitated as described under "Materials and Methods." The results are expressed as a percentage of [3H]GDP bound to eIF-2 (eIF-2·[3H]GDP) at time 0. The different rates of GDP exchange observed for the titrated extracts shown in A and B were then compared with the GDP exchange activity observed for 40 µl of extract of cells expressing the eIF-2B epsilon -subunit alone (C, bullet ). The control for each experiment was 40 µl of extract of cells infected with wild-type AcNPV virus (black-square). Expression of eIF-2B subunits was monitored by immunoblot analysis as described in the legend to Fig. 1. Results of typical blots are shown in the insets. The positions of the subunits of eIF-2B are noted to the left of each inset. All infections were done in duplicate, and GEF activity is represented as an average value for the duplicate plates.
[View Larger Version of this Image (25K GIF file)]

Formation of the eIF-2B Holoprotein in Sf9 Cells

The enhanced rate of GDP exchange observed in extracts of cells expressing all five subunits of eIF-2B suggested that the coexpressed subunits were interacting to form a functional multisubunit complex. To examine eIF-2B complex formation in insect cells, we utilized a monoclonal antibody against the epsilon -subunit of rat eIF-2B previously shown to immunoprecipitate the five-subunit eIF-2B complex isolated from rat liver (30). In these studies, extracts of Sf9 cells expressing all five subunits of eIF-2B were immunoprecipitated using the anti-eIF-2Bepsilon monoclonal antibody, and the resulting immunoprecipitates were resolved by SDS-PAGE. The proteins in the gel were visualized by immunoblotting with a rabbit polyclonal antibody against the FLAG epitope and reprobed with a monoclonal antibody against eIF-2Bgamma . All five subunits were effectively coimmunoprecipitated with the anti-eIF-2Bepsilon monoclonal antibody (Fig. 4B, lane 1), indicating that the subunits interact to form a complex in insect cells. As a control for antibody specificity, Sf9 cells were doubly infected so as to express the alpha -, beta -, gamma -, and delta -subunits in the absence of the epsilon -subunit (Fig. 4B, lane 2). In the absence of the epsilon -subunit, none of the eIF-2B subunits were immunoprecipitated by the anti-eIF-2Bepsilon monoclonal antibody. In addition, in the absence of the anti-eIF-2Bepsilon monoclonal antibody, none of the subunits were detected in the precipitate. Thus, in Sf9 cell extracts containing all five rat eIF-2B subunits, the subunits associate to form a stable complex that can be detected by coimmunoprecipitation.


Fig. 4. Analysis of holoprotein formation in lysates from Sf9 cells expressing rat eIF-2B subunits by coimmunoprecipitation using a monoclonal antibody against rat eIF-2Bepsilon . Sf9 cells were infected with (lanes 1 and 3) or without (lane 2) virus expressing the epsilon -subunit of rat eIF-2B in combination with the alpha -, beta -, gamma -, and delta -subunits and lysed 60 h post-infection as described under "Materials and Methods." A, aliquots (15 µl) of cell extract were subjected to protein immunoblot analysis as described in the legend to Fig. 1. B, the remainder of the extract was immunoprecipitated using a monoclonal antibody against the epsilon -subunit of eIF-2B (eIF-2Bepsilon MAB; lanes 1 and 2). Goat anti-mouse Biomag IgG beads were used to collect the antigen-antibody complexes. As a control, one sample was incubated without antibody (lane 3) prior to addition of the Biomag beads. Immune complexes were washed, and proteins in the immunocomplexes were resolved by SDS-PAGE. Proteins in the gel were transferred to nitrocellulose and visualized by immunoblotting with a rabbit polyclonal antibody against the FLAG epitope, followed by reprobing using a monoclonal antibody against the eIF-2B gamma -subunit. The positions of the subunits of eIF-2B are noted to the left of each panel.
[View Larger Version of this Image (40K GIF file)]

Extracts of triply infected Sf9 cells expressing either four (beta , gamma , delta , and epsilon ) or all five rat eIF-2B subunits were chromatographed on a Superose 6 gel filtration column to examine the molecular mass and activity of the recombinant multisubunit complexes. As shown in Fig. 5, GEF activity cofractionated with the multisubunit complex and exhibited an apparent molecular mass similar to that observed for eIF-2B isolated from rat liver. In contrast, when extracts of cells expressing eIF-2Bepsilon alone were fractionated, GEF activity eluted at a much lower apparent molecular mass than observed for the protein purified from rat liver. The latter result suggests that the GEF activity observed in cells expressing eIF-2Bepsilon alone is not the result of rat eIF-2Bepsilon binding to the endogenous Sf9 cell eIF-2B alpha -, beta -, gamma -, and/or delta -subunit to form an active complex. Rather, the results suggest that the eIF-2B epsilon -subunit alone exhibits GEF activity.


Fig. 5. Fractionation of eIF-2B complexes expressed in Sf9 cells by gel filtration chromatography. Sf9 cells were infected with combinations of baculoviruses encoding all five rat eIF-2B subunits (square ); the beta -, gamma -, delta -, and epsilon -subunits of eIF-2B (bullet ); or the epsilon -subunit alone (black-triangle). The cells were lysed 60 h post-infection, and extracts were prepared and fractionated using a Superose 6 gel filtration column as described under "Materials and Methods." For extracts of cells expressing the four- or five-subunit eIF-2B complex, aliquots (0.02 ml) of column fractions were assayed for GEF activity as described above. Under the conditions used for gel filtration, the alpha -subunit did not comigrate with the other four subunits and did not express detectable GEF activity. No GEF activity was detected in column fractions following fractionation of extracts of cells expressing eIF-2Bepsilon alone. Therefore, for these extracts, every other fraction was pooled, and the combined fractions (1 ml each) were concentrated to ~0.1 ml using Centricon 30 centrifugal concentrators (Amicon, Inc.) prior to assay of GEF activity. eIF-2B purified from rat liver (open circle ) was also chromatographed and analyzed for GEF activity. The migration of molecular mass standards (in kilodaltons) is noted at the top of the figure.
[View Larger Version of this Image (31K GIF file)]

Effect of eIF-2alpha Phosphorylation on eIF-2B Activity

In vivo, the GEF activity of eIF-2B is competitively inhibited by phosphorylation of the alpha -subunit of eIF-2 by the protein kinases PKR, HCR, and GCN2. Genetic analyses and transfection studies with mammalian eIF-2 kinases in yeast indicate that the inhibitory effects of eIF-2alpha phosphorylation are mediated in part by the alpha -subunit of eIF-2B (37, 38). However, the studies conducted in yeast provide an indirect measurement of eIF-2B activity since the parameter used is the induction of the histidine biosynthetic pathway in response to eIF-2alpha phosphorylation. The expression of recombinant eIF-2B subunits in Sf9 cells provided an opportunity, for the first time, to directly measure the effect of eIF-2alpha phosphorylation on eIF-2B activity. We therefore examined the effects of eIF-2alpha phosphorylation on the activity of four-subunit (beta , gamma , delta , and epsilon ) and five-subunit rat eIF-2B protein complexes produced in Sf9 insect cells. GEF assays were conducted on the cell extracts following preincubation of the extract with either unphosphorylated eIF-2 or eIF-2 phosphorylated by HCR in vitro. As shown in Fig. 6, extracts of cells expressing either four eIF-2B subunits (beta , gamma , delta , and epsilon ) or the five-subunit complex exhibited similar rates of GDP exchange, and in both cases, the rate of exchange was significantly greater than in extracts of cells infected with wild-type virus. The rate of GDP exchange measured both in extracts of cells expressing all five subunits and in extracts of cells infected with wild-type virus was greatly reduced when extracts were preincubated with phosphorylated eIF-2 compared with preincubation with unphosphorylated eIF-2 (Fig. 6). In contrast, the activity of the eIF-2B complex lacking the alpha -subunit was not inhibited by phosphorylated eIF-2. The amount of eIF-2alpha in the phosphorylated form at the end of the assay was the same for each condition (data not shown), demonstrating that the lack of inhibition observed with cell extracts expressing the four-subunit eIF-2B complex (beta , gamma , delta , and epsilon ) was not the result of dephosphorylation of eIF-2alpha during the assay. The lack of inhibition of the four-subunit eIF-2B complex by phosphorylated eIF-2 demonstrates, for the first time, a regulatory role for the eIF-2B alpha -subunit using a biochemical assay.


DISCUSSION

Sf9 cells are useful for the production of biologically active recombinant proteins because the proteins they express appear to undergo proper folding and correct post-translational modifications (39). Therefore, we were interested in testing the utility of Sf9 cells to study eIF-2B as the insect cell system has been used successfully to produce active recombinant GEFs for the Ras superfamily of proteins (40, 41). An important feature of Sf9 cells is that they can be simultaneously infected with multiple viruses, and this allows the coexpression of multiple gene products. In the study described here, we focused on expressing all five subunits of rat eIF-2B in insect cells via baculovirus infection.

The multisubunit structure of eIF-2B is more complex than most GEFs, which consist of predominantly single subunit proteins. The complex structure of eIF-2B suggests that it may contain several distinct structural and functional domains. We found that expression of the eIF-2B epsilon -subunit alone resulted in detectable GDP exchange compared with extracts of cells infected with wild-type virus. In addition, GEF activity was found to cofractionate with the epsilon -subunit during gel filtration chromatography. In contrast, we have not been able to detect any GEF activity following purification of eIF-2Bepsilon using an anti-FLAG immunoaffinity resin (data not shown). It is possible that the conditions used during the purification resulted in inactivation of the protein. Nonetheless, we feel that the assumption that eIF-2Bepsilon alone manifests GEF activity must be viewed with caution until further proof validating the activity of the individual subunit is obtained.

By coinfecting insect cells with a combination of three recombinant baculoviruses, we were able to simultaneously express all five subunits of rat eIF-2B in Sf9 insect cells and to show that the expressed complex exhibited GEF activity. The presence of the eIF-2B epsilon -subunit was required for GEF activity, supporting a catalytic role for this subunit. This result is consistent with the previous observation (42) that the guanine nucleotide exchange activity of purified eIF-2B could be blocked by monoclonal antibodies recognizing the epsilon -subunit. The formation of an intact eIF-2B complex in coinfected insect cells was confirmed by coimmunoprecipitation and gel filtration assays.

In this study, we also used the insect cell expression system to explore the role of eIF-2Balpha in the regulation of eIF-2B activity. Studies in yeast suggest that a four-subunit eIF-2B complex lacking eIF-2Balpha is functional since deletion of GCN3 (eIF-2Balpha ) has no effect on cell growth under normal conditions (27). In contrast, Craddock and Proud (29) recently reported that purified rabbit eIF-2B lacking eIF-2Balpha was not functional and suggested that mammalian eIF-2B may therefore differ from yeast eIF-2B in the requirement for the alpha -subunit. By overexpressing combinations of rat eIF-2B subunits in Sf9 cells, we show that the alpha -subunit of eIF-2B is not required for GEF activity. Furthermore, using gel filtration chromatography, we were able to confirm that the GEF activity in Sf9 cell lysates containing the four-subunit eIF-2B complex (lacking eIF-2Balpha ) was due to the formation of an active four-subunit complex. It is not clear why the alpha -subunit was missing from certain purified rabbit eIF-2B preparations reported by Craddock and Proud (29). One interpretation is that the absence of the alpha -subunit in the rabbit eIF-2B preparations might indicate modifications in the other eIF-2B subunits that were not observed, but that might be inhibiting the activity of the purified four-subunit eIF-2B complex.

The alpha -subunit of eIF-2 is phosphorylated at Ser51 by three different protein kinases (PKR, HCR, and GCN2), and this phosphorylation event is an important mechanism used by cells to control translation initiation in response to a variety of stresses (3). In addition, recent evidence suggests that at least one of the eIF-2alpha kinases, PKR, may have a potential role as a tumor suppressor in normal cells and function in the restraint of cell growth (43). In yeast, deletion of the eIF-2B alpha -subunit (GCN3) suppresses the lethal effects from a constitutively active GCN2 kinase or the growth inhibitory effects resulting from high level exogenous expression of mammalian eIF-2alpha kinases (HCR and PKR) (37, 44, 45). The genetic and expression studies in yeast provide indirect evidence that the GEF activity of eIF-2B lacking the alpha -subunit is insensitive to phosphorylation of eIF-2alpha . Phosphorylated eIF-2 has an increased affinity for eIF-2B and acts as a competitive inhibitor since the complex formed does not undergo GDP/GTP exchange. The inhibition of eIF-2B activity impairs translation initiation by reducing the availability of ternary complexes (eIF-2·GTP·Met-tRNAi). In this report, we provide direct evidence that the effect of eIF-2 phosphorylation is indeed mediated by the alpha -subunit of eIF-2B because the GEF activity of a four-subunit eIF-2B complex lacking eIF-2Balpha was insensitive to eIF-2alpha phosphorylation.

Finally, the baculovirus-infected insect cell expression system may prove useful for determining the mechanism by which eIF-2B catalyzes the GEF reaction on eIF-2. At least two possible mechanisms have been proposed for the eIF-2B-catalyzed guanine nucleotide exchange reaction (46). One possible model is the substituted enzyme (ping-pong) mechanism, which involves the sequential release of GDP and the binding of GTP to eIF-2 catalyzed by eIF-2B. An alternative mechanism is based upon kinetic analysis of the GDP exchange reaction and involves complex formation between all four components of the exchange reaction (eIF-2, GDP, eIF-2B, and GTP). The four-component mechanism implies that eIF-2B contains a binding site for GTP. Dholakia et al. (47) reported the binding of a photoreactive 8-azido analog of GTP to the beta -subunit of eIF-2B and therefore suggested that eIF-2B is a GTP-binding protein. The GTP binding data are difficult to reconcile with the amino acid sequences of yeast or mammalian eIF-2B subunits since none of the subunits harbor a consensus sequence for GTP binding. Future studies using different combinations of normal and mutant versions of individual eIF-2B subunits will be required to further delineate their functional roles in the guanine nucleotide exchange reaction.


FOOTNOTES

*   This work was supported by Grants DK 13499 and DK 15658 from the National Institutes of Health.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Dagger    To whom correspondence should be addressed: Dept. of Cellular and Molecular Physiology, Pennsylvania State University College of Medicine, P. O. Box 850, Hershey, PA 17033. Tel: 717-531-8567; Fax: 717-531-7667; E-mail: jjefferson{at}cmp.hmc.psu.edu.
1   The abbreviations used are: eIFs, eukaryotic initiation factors; GEF, guanine nucleotide exchange factor; PAGE, polyacrylamide gel electrophoresis; MOPS, 4-morpholinepropanesulfonic acid; Ac NPV, A. californica nuclear polyhedrosis virus; HCR, heme-controlled repressor.

ACKNOWLEDGEMENT

We are grateful to Lynne Hugendubler for technical assistance.


REFERENCES

  1. Kimball, S. R., Vary, T. C., and Jefferson, L. S. (1994) Annu. Rev. Physiol. 56, 321-348 [CrossRef][Medline] [Order article via Infotrieve]
  2. Redpath, N. T., and Proud, C. G. (1994) Biochim. Biophys. Acta 1220, 147-162 [Medline] [Order article via Infotrieve]
  3. Pain, V. M. (1996) Eur. J. Biochem. 236, 747-771 [Abstract]
  4. Clemens, M. J. (1996) in Translational Control (Hershey, J. W. B., Mathews, M. B., and Sonenberg, N., eds), pp. 139-172, Cold Spring Harbor Laboratory Press, Plainview, NY
  5. Rhoads, R. E. (1993) J. Biol. Chem. 268, 3017-3020 [Free Full Text]
  6. Hershey, J. W. B. (1989) J. Biol. Chem. 264, 20823-20826 [Free Full Text]
  7. Dever, T. E., Yang, W., Astrom, S., Bystrom, A. S., and Hinnebusch, A. G. (1995) Mol. Cell. Biol. 15, 6351-6363 [Abstract]
  8. Welsh, G. I., and Proud, C. G. (1992) Biochem. J. 284, 19-23 [Medline] [Order article via Infotrieve]
  9. Gilligan, M., Welsh, G. I., Flynn, A., Bujalska, I., Diggle, T. A., Denton, R. M., Proud, C. G., and Docherty, K. (1996) J. Biol. Chem. 271, 2121-2125 [Abstract/Free Full Text]
  10. Kimball, S. R., and Jefferson, L. S. (1988) Biochem. Biophys. Res. Commun. 156, 706-711 [Medline] [Order article via Infotrieve]
  11. Jeffrey, I. W., Kelly, F. J., Duncan, R., Hershey, J. W. B., and Pain, V. M. (1990) Biochimie (Paris) 72, 751-757 [Medline] [Order article via Infotrieve]
  12. Rowlands, A. G., Montine, K. S., Henshaw, E. C., and Panniers, R. (1988) Eur. J. Biochem. 175, 93-99 [Abstract]
  13. Singh, L. P., Denslow, N. D., and Wahba, A. J. (1996) Biochem. J. 35, 3206-3212 [CrossRef]
  14. Dholakia, J. N., and Wahba, A. J. (1988) Proc. Natl. Acad. Sci. U. S. A. 85, 51-54 [Abstract]
  15. Welsh, G. I., and Proud, C. G. (1993) Biochem. J. 294, 625-629 [Medline] [Order article via Infotrieve]
  16. Singh, L. P., Aroor, A. R., and Wahba, A. J. (1994) Biochemistry 33, 9152-9157 [Medline] [Order article via Infotrieve]
  17. Dholakia, J. N., Mueser, T. C., Woodley, C. L., Parkhurst, L. J., and Wahba, A. J. (1986) Proc. Natl. Acad. Sci. U. S. A. 83, 6746-6750 [Abstract]
  18. Singh, L. P., Aroor, A. R., and Wahba, A. J. (1995) Biochem. Biophys. Res. Commun. 212, 1007-1014 [CrossRef][Medline] [Order article via Infotrieve]
  19. Karinch, A. M., Kimball, S. R., Vary, T. C., and Jefferson, L. S. (1993) Am. J. Physiol. 264, E101-E108 [Abstract/Free Full Text]
  20. Kimball, S. R., and Jefferson, L. S. (1995) Biochem. Biophys. Res. Commun. 212, 1074-1081 [CrossRef][Medline] [Order article via Infotrieve]
  21. Overbeck, A. F., Brtva, T. R., Cox, A. D., Graham, S. M., Huff, S. Y., Khosravi-Far, R., Quilliam, L. A., Solski, P. A., and Der, C. J. (1995) Mol. Reprod. Dev. 42, 468-476 [Medline] [Order article via Infotrieve]
  22. Quilliam, L. A., Khosravi-Far, R., Huff, S. Y., and Der, C. J. (1995) Bioessays 17, 395-404 [Medline] [Order article via Infotrieve]
  23. Price, N. T., Mellor, H., Craddock, B. L., Flowers, K. M., Kimball, S. R., Wilmer, T., Jefferson, L. S., and Proud, C. G. (1996) Biochem. J. 318, 637-643 [Medline] [Order article via Infotrieve]
  24. Price, N. T., Kimball, S. R., Jefferson, L. S., and Proud, C. G. (1996) Biochem. J. 318, 631-636 [Medline] [Order article via Infotrieve]
  25. Flowers, K. M., Mellor, H., Kimball, S. R., and Jefferson, L. S. (1995) Biochim. Biophys. Acta 1264, 163-167 [Medline] [Order article via Infotrieve]
  26. Flowers, K. M., Mellor, H., Matts, R. L., Kimball, S. R., and Jefferson, L. S. (1996) Biochim. Biophys. Acta 1307, 318-324 [Medline] [Order article via Infotrieve]
  27. Hinnebusch, A. G. (1994) Trends Biochem. Sci. 19, 409-414 [CrossRef][Medline] [Order article via Infotrieve]
  28. Hannig, E. M., and Hinnebusch, A. G. (1988) Mol. Cell. Biol. 8, 4808-4820 [Medline] [Order article via Infotrieve]
  29. Craddock, B. L., and Proud, C. G. (1996) Biochim. Biophys. Acta 220, 843-847
  30. Kimball, S. R., Karinch, A. M., Feldhoff, R. C., Mellor, H., and Jefferson, L. S. (1994) Biochim. Biophys. Acta 1201, 473-481 [Medline] [Order article via Infotrieve]
  31. Craddock, B. L., Price, N. T., and Proud, C. G. (1995) Biochem. J. 309, 1009-1014 [Medline] [Order article via Infotrieve]
  32. Richardson, C. D. (1995) Baculovirus Expression Protocols, Vol. 39, Humana Press, Totowa, NJ
  33. Kimball, S. R., and Jefferson, L. S. (1990) J. Biol. Chem. 265, 16794-16798 [Abstract/Free Full Text]
  34. Kimball, S. R., Everson, W. V., Myers, L. M., and Jefferson, L. S. (1987) J. Biol. Chem. 262, 2220-2227 [Abstract/Free Full Text]
  35. Kimball, S. R., Antonetti, D. A., Brawley, R. M., and Jefferson, L. S. (1991) J. Biol. Chem. 266, 1969-1976 [Abstract/Free Full Text]
  36. Panniers, R., Rowlands, A. G., and Henshaw, E. C. (1988) J. Biol. Chem. 263, 5519-5525 [Abstract/Free Full Text]
  37. Dever, T. E., Chen, J.-J., Barber, G. N., Cigan, A. M., Feng, L., Donahue, T. F., London, I. M., Katze, M. G., and Hinnebusch, A. G. (1993) Proc. Natl. Acad. Sci. U. S. A. 90, 4616-4620 [Abstract]
  38. Dever, T. E., Feng, L., Wek, R. C., Cigan, A. M., Donahue, T. F., and Hinnebusch, A. G. (1992) Cell 68, 585-596 [Medline] [Order article via Infotrieve]
  39. Fraser, M. J. (1992) Curr. Top. Microbiol. Immunol. 158, 131-172 [Medline] [Order article via Infotrieve]
  40. Frech, M., Cussac, D., Chardin, P., and Bar-Sagi, D. (1995) Methods Enzymol. 256, 125-129 [Medline] [Order article via Infotrieve]
  41. Zheng, Y., Hart, M. J., and Cerione, R. A. (1995) Methods Enzymol. 256, 77-84 [Medline] [Order article via Infotrieve]
  42. Oldfield, S., Jones, B. L., Tanton, D., and Proud, C. G. (1994) Eur. J. Biochem. 221, 399-410 [Abstract]
  43. Barber, G. N., Wambach, M., Thompson, S., Jagus, R., and Katze, M. G. (1995) Mol. Cell. Biol. 15, 3138-3146 [Abstract]
  44. Ramirez, M., Wek, R. C., de Aldana, C. R. V., Jackson, B. M., Freeman, B., and Hinnebusch, A. G. (1992) Mol. Cell. Biol. 12, 5801-5815 [Abstract]
  45. Chong, K. L., Feng, L., Schappert, K., Meurs, E., Donahue, T. F., Friesen, J. D., Hovanessian, A. G., and Williams, B. R. G. (1992) EMBO J. 11, 1553-1562 [Abstract]
  46. Price, N., and Proud, C. (1994) Biochimie (Paris) 76, 748-760 [CrossRef][Medline] [Order article via Infotrieve]
  47. Dholakia, J. N., Francis, B. R., Haley, B. E., and Wahba, A. J. (1989) J. Biol. Chem. 264, 20638-20642 [Abstract/Free Full Text]

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.