Contributed by Martha Vaughan, January 4, 2007 (received for review December 28, 2006)
Brefeldin A-inhibited guanine nucleotide-exchange proteins (GEPs) BIG1 and BIG2 activate ADP-ribosylation factor (ARF) GTPases, which are required for vesicular trafficking. Both molecules contain one or more sites for binding protein kinase A, i.e., A kinase-anchoring protein (AKAP) sequences. Elevation of cell cAMP caused PKA-catalyzed phosphorylation and nuclear accumulation of BIG1 but not BIG2. We then asked whether BIG1 phosphorylation altered its GEP activity. Incubation of BIG1 or BIG2 with PKA catalytic subunits and ATP resulted in retardation of their electrophoretic migration, consistent with PKA phosphorylation. Okadaic acid inhibits many protein phosphatases, including protein phosphatase 1 (PP1) and PP2A, that can reverse PKA-catalyzed phosphorylation. Incubation of HepG2 cells with okadaic acid caused concentration-dependent accumulation of presumably phosphorylated BIG1 and BIG2 with decreased mobility, which was increased by subsequent incubation in vitro with specific recombinant phosphatases, PP1γ > PP2A ≫ PP1α. For assays of GEP activity, BIG1 and BIG2 were immunoprecipitated from cells that had been depleted, respectively, of BIG2 and BIG1 by using specific siRNA. GEP activity of each was significantly decreased after incubation with recombinant PKA plus ATP and restored by incubation with PP1γ. In agreement with a role for PP1γ in regulation of BIG, endogenous PP1γ, but not PP1α or β, was immunoprecipitated with BIG1 or BIG2 from microsomal fractions. All observations are consistent with the effects of BIG1 and BIG2 phosphorylation on vesicular trafficking, via alterations in ARF activation and regulatory roles for cAMP, PKA, and PP1γ in ARF activation by BIG1 and BIG2.
ADP-ribosylation factors (ARFs) are 20-kDa guanine nucleotide-binding proteins with multiple intracellular functions, including critical roles in vesicular trafficking. Conversion of inactive ARF with GDP bound to active ARF–GTP requires a guanine nucleotide-exchange protein (GEP) to accelerate GDP release (1–3). GEPs, like ARFs, are ubiquitous in eukaryotic cells. All GEP molecules contain structurally similar, central Sec7 domains that are responsible for ARF activation and its inhibition by brefeldin A but differ otherwise in size, structure, and biochemical properties (3–5). Two ≈200-kDa brefeldin A-inhibited GEPs, BIG 1 and BIG2, first purified together in macromolecular complexes from bovine brain (6), are 74% identical in overall amino acid sequences, with 90% identity in their Sec7 domains (7). Li et al. (8) had found that BIG2 interacted in a yeast two-hybrid screen with RIα, a regulatory (R) subunit of PKA and identified three R-binding, or A kinase-anchoring protein (AKAP), sequences in BIG2 with different R-subunit specificities. Several other proteins with AKAP domains are known to be phosphorylated (9), and it was later shown that phosphorylation of the consensus PKA substrate site, Ser-883, in BIG1 was required for its nuclear accumulation after elevation of intracellular cAMP (10).
Many biological processes are regulated by reversible phosphorylation. Cells use this posttranslational modification to alter the properties of a great many proteins involved in metabolic, mechanical, and signaling activities. Protein phosphorylation is controlled dynamically by kinases that add and phosphatases that remove phosphate. Eukaryotic cells contain hundreds of protein kinases, which are much more numerous than the phosphatases, each with its own substrate specificity, intracellular localization, and regulation (11–13). The cAMP-dependent PKA, a serine/threonine kinase that exists in several isoforms, is a tetrameric protein comprising two catalytic (C; isoforms α, β, or γ) and a homodimer of type Iα, Iβ, IIα, or IIβ R subunits (11). In cells, the holoenzyme is usually tethered by interaction of the R dimer with AKAP, which acts as a scaffold for assembly of multimolecular machines in which kinase, substrate, phosphatase, and other molecules can act in concert to locate, coordinate, and limit, spatially as well as temporally, the effects of cAMP (14). AKAP domains are amphipathic helices of 14–18 aa (15, 16), three of which have been identified in BIG2; only one of those is present in BIG1. Consistent with their potential AKAP functions, endogenous BIG1 and BIG2 have been coimmunoprecipitated with antibodies against individual R or C subunits from HepG2 cytosol (8). BIG2 lacks a PKA phosphorylation site corresponding to Ser-833 in BIG1 but contains at least two other potential consensus sites.
In eukaryotic proteins, dephosphorylation of most serine/threonine phosphates is catalyzed by four enzymes, protein phosphatase 1 (PP1), PP2A, PP2B, and PP2C, which differ in metal ion requirements, substrate specificities, and sensitivities to inhibitors (13, 17). PP1, a major tissue serine/threonine phosphatase, is abundant in virtually all cell compartments, has a key regulatory role in cell cycle progression, and is also involved, for example, in muscle contraction, glycogen metabolism, gene expression, calcium transport, neurotransmission, and intracellular transport (17). The PP1 holoenzyme comprises a catalytic monomer (three mammalian genes, α, β, or γ isoforms), with regulatory subunits that can influence catalytic activity, subcellular localization, and substrate specificity (13, 17). PP1 and PP2A, which also has a wide range of biological functions, together account for >90% of all serine/threonine phosphatase activity in mammalian cells (17). The numerous regulatory or targeting proteins that interact with PP1 in a mutually exclusive manner do so via amino acid sequence Arg/Lys–Val/Ile–Xaa–Phe, which exists in many more proteins than are known to associate with PP1 (18). The same two consensus sequences are present in both BIG1 and BIG2, plus two potentially functional variants (19) in BIG2. Modulation of GEP activity by phosphorylation/dephosphorylation of PKA and other kinase sites in these molecules could clearly be important.
We report here that incubation in vitro with PP1γ increased GEP activity of immunoprecipitated BIG1 and BIG2 that had been inhibited by prior incubation with recombinant PKA C subunit plus ATP. All data are consistent with roles for these enzymes in cAMP regulation of Golgi structure, as recently described (20).
Because endogenous BIG1 and BIG2 had been coimmunoprecipitated from HepG2 cells with antibodies against PKA C or R subunits (8), we investigated whether PKA would modify endogenous BIG1 and BIG2 in vitro. Incubation of each immunoprecipitated protein with three different recombinant PKA C subunits plus ATP for 30 min resulted in significant retardation of its electrophoretic migration, which was greater with Cβ and Cγ than with Cα for both BIG1 and BIG2 (Fig. 1). ATP alone did not alter mobilities; this was consistent with PKA-catalyzed phosphorylation.
In vitro phosphorylation of endogenous BIG1 or BIG2 by PKA. Samples of proteins precipitated from postnuclear supernatants with antibodies against BIG1 (A) or BIG2 (B) were applied directly (NI) to 4% Tris-glycine gel or after incubation at 30°C for 30 min without (C) or with 10 unit of PKA α, β, or γ catalytic subunits and 50 μM ATP in PKA buffer (total volume, 50 μl). Proteins were then separated by SDS/PAGE and reacted with antibodies against BIG1 or BIG2. Distance (millimeter) between the center of each immunoreactive band and the center of 150-kDa protein marker was measured on prints from Fuji Image Gauge software (version 4.0). Data are means ± SD of values from three experiments. ∗, P < 0.01 vs. NI; ∗∗, P < 0.01.
For clues to phosphatases that might be responsible for intracellular dephosphorylation, we assessed effects of several phosphatase inhibitors on the electrophoretic mobility of endogenous BIG1 and BIG2. Migration of each was clearly retarded after incubation of cells for 3 h with 50 nM okadaic acid (OA), and the effect increased with increasing OA concentration to a maximum at 200–500 nM (Fig. 2A). Effects of calyculin A (CA) were similar to those of OA but occurred at concentrations at least 10-fold lower (Fig. 2B). These findings might reflect the participation of both PP1 and PP2A in dephosphorylation of BIG1 and BIG2. Treatment with cyclosporin A or orthovanadate did not significantly alter BIG1 or BIG2 migration, although with vanadate BIG2 was consistently slightly retarded, perhaps because of a tyrosine phosphorylation (Fig. 2 C and D).
Effect of phosphatase inhibitors on phosphorylation of BIG1 and BIG2 in cells. Samples (10 μg) of total proteins are from cells that had been incubated for 3 h at 37°C without (C) or with the indicated concentrations of OA (A), CA (B), 100 nM cyclosporin A (CsA) or 500 nM OA (C), and 1 mM sodium orthovanadate (OV) or 500 nM OA (D). Data are reported as in Fig. 1.
Because OA and CA, inhibitors of both PP1 and PP2A, apparently inhibited intracellular dephosphorylation of BIG1 and BIG2 (Fig. 2), we assessed the effects of individual phosphatases in vitro on the endogenously phosphorylated BIG1 and BIG2 that had accumulated in cells incubated for 3 h with OA (Fig. 3). Migration of both BIG1 and BIG2 was maximally increased after incubation for 30 min with 1 or 2 units of PP2A and, to a significantly greater extent, with the same amounts of PP1γ (P < 0.01). Effects of PP1α were much smaller and were not maximal at the concentrations used.
In vitro dephosphorylation of endogenously phosphorylated BIG1 and BIG2. Phosphorylated BIG1 (A) or BIG2 (B), immunoprecipitated from cells that had been incubated with 500 nM OA for 3 h, were incubated at 30°C for 30 min without (NI) or with the indicated amount of PP1α, PP1γ, or PP2A in a total volume of 50 μl before measurement of electrophoretic migration, as in Fig. 1. Data (Right) are means ± SD of values from three experiments, with representative blots from one experiment (Left).
Effects of altered phosphorylation on the distributions of BIG1 and BIG2 were assessed in cells that had been incubated with 500 nM OA to inhibit PP1 and PP2A or with 100 μM H89 (a potentially selective inhibitor of PKA) for 3 h before preparation of homogenate fractions (Fig. 4). From cells treated with OA, as seen in other experiments, electrophoretic migration of (presumably phosphorylated) BIG1 and BIG2 was significantly slower than that from control cells. OA treatment also significantly increased amounts of BIG1 and BIG2 in cytosol, whereas they were decreased in membrane fractions (Fig. 4). Effects of H89 treatment, which may have inhibited protein kinases in addition to PKA, were essentially opposite, with levels of both BIG1 and BIG2 in membrane fractions increased and those in cytosol decreased. Their migrations were similar to those of the proteins in control cells, although BIG2 seemed to have migrated faster in all fractions of cells incubated with H89, consistent with decreased phosphorylation.
Effects of OA or H89 on localization of BIG1 and BIG2 in HepG2 cell fractions. Cells were incubated for 3 h at 37°C without (C) or with 500 nM OA or 100 μM H89 before homogenization. Samples (20 μg) of proteins from whole homogenate, cytosol, membrane, and crude nuclear fractions were separated by SDS/PAGE in 4% or 4–12% gel before immunoblotting, as indicated, with antibodies against BIG1 and BIG2 or α-tubulin, Golgi 58K, and histone H1. Means ± SD of densitometric values from three experiments quantified by Fuji Image Gauge version 4.0 and expressed relative to that for C in the same experiment set equal to 100 are shown with blots from a representative experiment. ∗, P < 0.01 vs. untreated cells.
On confocal immunofluorescence microscopy of cells incubated with OA, both BIG1 and BIG2 disappeared from their apparent colocalization with GM130 at perinuclear, presumably Golgi, structures and were seen in all cell nuclei (Fig. 5). Effects of H89 on the distribution of BIG2 seemed more obvious than those on BIG1, although both remained, in part, colocalized with GM130 (Fig. 5). These data are consistent with the results of Western blotting of subcellular fractions, except that no changes were detected in amounts of the two proteins in crude nuclear fractions, for reasons that are not clear.
Effect of OA or H89 on localization of BIG1 and BIG2 in HepG2 cells by confocal immunofluorescence microscopy. Cells were incubated for 75 min at 37°C without (C) or with 500 nM OA or 100 μM H89 before staining of immunoreactive BIG1 (green) or BIG2 (green) and GM130 (red) and visualization by confocal laser-scanning microscopy. Data were similar in two other experiments. (Scale bar, 20 μm.)
Believing that the observed alterations in electrophoretic migration and subcellular distribution of BIG1 and BIG2 reflected, at least to some extent, alterations in their phosphorylation, we asked whether they were associated with changes in GEP activity of the proteins. In those experiments, to minimize coimmunoprecipitation of the other GEP and additional proteins, cells were first depleted of BIG1 or BIG2 selectively by incubation for 72 h by using specific siRNAs. The GEP activity of BIG1 or BIG2 from cells that had been incubated for 3 h with OA after 72 h of siRNA treatment was significantly less than that from untreated cells (Fig. 6). In additional experiments, we assessed GEP activity of BIG1 and BIG2 that were phosphorylated by PKA and then dephosphorylated by PP1 in vitro. GEP activities of both BIG1 and BIG2 were decreased after PKA-catalyzed phosphorylation and restored by PP1 dephosphorylation (Fig. 7).
Effect of OA treatment of cells on GEP activity of BIG1 or BIG2. Immunoprecipitated BIG1 (B1) or BIG2 (B2) from cells that had been depleted, respectively, of BIG2 or BIG1 followed by incubation for 3 h without (N) or with 500 nM OA, was assayed for GEP activity with incubation at 37°C for 1 h. Activities were calculated as picomoles of [35S]GTPγS (after correction for binding by beads with nonimmune IgG immunoprecipitate from untreated cells that were incubated without ARF) divided by densitometric value for BIG1 or BIG2 in the immunoprecipitate assayed. Means ± SD of values from three experiments (Left) were calculated after activity of each of the OA-treated samples was expressed as a percentage of its control (N = 100). Representative blots from one experiment (Right). ∗∗, P < 0.05 vs. without OA.
Effect of BIG1 or BIG2 phosphorylation on GEP activity. Experiments were performed as in Fig. 6, but starting with 15 × 105 cells for each homogenate and 3-fold amounts of all materials, because 95% of each immunoprecipitate from 5 × 105 cells was used for GEP assay. Of the three samples (1, 2, 3) from each preparation of immunoprecipitated BIG1 (B1) or BIG2 (B2), one was incubated in 100 μl of TENDS buffer alone (N) and two with 20 units of PKA Cγ in 100 μl of PKA buffer for 30 min before washing with TENDS (samples 1 and 2) or with PP1 buffer (sample 3), followed by incubation (30 min) in 100 μl of TENDS (samples 1 and 2) or PP1 buffer and 4 units of PP1γ (sample 3), washing with TENDS buffer, and assay of GEP activity (Fig. 6). Means ± SD of values from three experiments (Left) are shown with representative blots from one experiment (Right). ∗, P < 0.01 vs. TENDS buffer alone; ∗∗, P < 0.01 vs. PKA.
PP1γ, the most effective phosphatase for both BIG1 and BIG2 dephosphorylation, was immunoprecipitated from microsomal fractions with antibodies against BIG1 or BIG2. PP1α and PP2A were not detected in the immunoprecipitate (Fig. 8).
Coimmunoprecipitation of endogenous protein phosphatases from HepG2 cells. HepG2 cell cytosol (A) and microsome (B) fractions were incubated with BIG1 (B1), BIG2 (B2), or normal rabbit IgG (N). Precipitated proteins (IP), 10 μg of cytosol (C), or microsomal fraction (M) were separated by SDS/PAGE before immunoblotting (IB) with indicated antibodies.
Li et al. (8) reported that endogenous BIG1 and BIG2 were coimmunoprecipitated by antibodies against PKA subunits RIα, RIIα, or C and that endogenous RIα was immunoprecipitated by antibodies against BIG1 or BIG2, although direct interaction of the proteins was not demonstrated. Phosphorylation of consensus PKA site, Ser-883 in BIG1, was required for nuclear accumulation of BIG1 after the elevation of cAMP in HepG2 cells by incubation with forskolin or 8-bromo-cAMP (10). This serine sequence is not present in BIG2, but both proteins contain other potential PKA substrate sites. We show here that incubation of immunoprecipitated endogenous BIG1 or BIG2 with ATP and recombinant PKA C subunit resulted in slowed electrophoretic migration, consistent with its phosphorylation by PKA in vitro. Many kinds of posttranslational protein modifications can, of course, alter electrophoretic mobility, and we have not yet quantified the stoichiometry of BIG1 and BIG2 phosphorylation, either in cells or in vitro. We believe, however, that changes in migration of BIG1 and BIG2 after incubation of the immunoprecipitated proteins individually with PKA (plus ATP) or PP, in all likelihood, reflect changes in phosphorylation.
The magnitude of difference in migration, which increased with time of incubation of cells with OA, for example, seems consistent with accumulation of additional phosphate as phosphatase inhibition continued and perhaps even increased or extended to additional locations or less-sensitive enzymes. BIG1 and BIG2 contain, respectively, 145 and 140 serines, 112 and 99 threonines, and 42 and 39 tyrosines (7), representing numerous potential sites of phosphorylation by diverse kinases. Many of those probably contributed to alterations of mobility of proteins from cells incubated with compounds like OA that inhibit more than one phosphatase. In PP1, one C subunit associates with different regulatory subunits that modulate substrate specificity as well as catalytic activity (13, 17). The heterogeneity of phosphorylated BIG1 and BIG2 in cells incubated for 3 hours with OA is reflected in the different effects of subsequent in vitro incubation with PP1α, PP1γ, or PP2A on their electrophoretic migration. PP1γ caused the largest increases in mobility of both BIG1 and BIG2, which were maximal with 1 unit, as were those of PP2A. The effects of PP1γ on distances migrated, however, were distinctly larger than those of PP2A. The same amounts of PP1α had lesser effects that were apparently not maximal with 2 units.
Earlier experiments in the laboratory had shown that both BIG1 and BIG2 were immunoprecipitated from cytosol by antibodies against PP1 (R. Adamik, J.M., and M.V., unpublished data). Conversely, antibodies against BIG1 or two different BIG2 sequences also precipitated PP1. Identification of PP1 isoforms (α, β, γ) was not completed because demonstration of an effect of PKA-catalyzed BIG2 phosphorylation on its GEP activity was a priority when the AKAP sequences in BIG2 were characterized (8). The later observation that elevation of cell cAMP content resulted in nuclear accumulation of BIG1, which contains only one of the three AKAP sequences identified in BIG2, led to finding that after PKA-catalyzed phosphorylation of Ser-883, BIG1 concentrated in nuclei (10). Obvious questions then included whether BIG1 in the nucleus (perhaps phosphorylated) had ARF GEP activity. That question has not been answered. We show here, however, that PKA-catalyzed phosphorylation in vitro of BIG1 immunoprecipitated from cells that had been selectively depleted of BIG2 with targeted siRNA resulted in significantly decreased GEP activity assayed with recombinant ARF1. Analogous experiments with similarly prepared BIG2 had parallel results. The less-active, PKA-phosphorylated BIG1 and BIG2 seemed more abundant in cytosol and the less-phosphorylated, more-active proteins in membrane fractions.
It had seemed that phosphorylation of Ser-883 near the C terminus of the BIG1 Sec7 domain might have effects on conformation or function, perhaps involving the ARF-interaction site or catalytic Glu-793, which is critical for GEP activity. In BIG2, however, alanine occupies the site corresponding to Ser-883 in BIG1, where its replacement by Ala abolished nuclear accumulation (10). Two predicted PKA phosphorylation sites in BIG2 are outside the Sec7 domain and offer no basis for speculation on mechanisms of influencing GEP activity. Our experiments do not exclude the possibility that PKA-catalyzed phosphorylation results in a stable association of BIG1 or BIG2 with a molecule(s) that prevents its effective interaction with ARF substrate, for example, with 14-3-3 proteins that bind and alter the activity of other phosphorylated AKAPs (21, 22). The molecules associated with immunoprecipitated endogenous BIG1 and BIG2 that we used in GEP assays remain to be identified. Participation of additional, presently unknown proteins in the inhibition of BIG1 and BIG2 GEP activity by PKA phosphorylation seems, in fact, quite possible.
We have not yet identified the R subunit involved in cAMP-induced phosphorylation of BIG1 or BIG2. Replication factor C (RFC) is a complex of five proteins that catalyzes ATP-dependent binding of proliferating cell nuclear antigen to DNA. This is required for replication, repair, and recombination, but RFC subunits also participate in other molecular interactions. RIα associates with RFC40 through its N-terminal dimerization/docking domain (23), just as it does with AKAP structures (24), although apparently without C subunits. RIα competed for binding to the same region of RFC40 with another subunit, RFC37, that could displace RIα and allow RFC complex formation. Specific depletion of RIα in tumor cells decreased amounts of the RFC40–RIα complex, as well as proliferation and survival, indicating a critical role for the dynamic RFC40–RIα complex (23). The presence of free RIα (without associated C subunits) in several tissues had led to many studies of its potential role in malignant cell growth, in addition to PKA regulation (25, 26). More recently, RIα was described as a tumor suppressor gene in humans, in whom its haploinsufficiency is associated with cardiac myxomas in Carney syndrome (27).
We do not know which R subunit(s) participated in any of the AKAP or phosphatase-binding actions of BIG1 and BIG2 that we believe can modify the ARF-activating GEP function via its phosphorylation state. Although RIα is the only one that interacted with BIG2 in a two-hybrid screen (8), other R and C subunits were coimmunoprecipitated with both BIG1 and BIG2. RIα was coimmunoprecipitated with BIG1 from cytosol and from crude nuclear fractions both before and after their nuclear accumulation was increased by elevation of cell cAMP (10). Phosphorylation of the PKA site in BIG1 was required for its nuclear accumulation after elevation of cell cAMP, consistent with its entrance to the nucleus and retention there in a PKA-phosphorylated, GEP-inhibited state. This, however, remains to be proved, and evidence that PKA-catalyzed phosphorylation always results in nuclear accumulation of BIG1 (or BIG2) is lacking. It seems probable that cAMP regulates ARF activation at more than one intracellular site via scaffolding functions of BIG1 and BIG2, each in specific molecular complexes that localize, respond to, and integrate multiple signals, not all of which involve a GEP function of BIG1 or BIG2 regulation.
Preparation and purification of antibodies against BIG1 and BIG2 have been published in ref. 28. Goat polyclonal antibodies against PP1α and PP1γ C subunit were purchased from Santa Cruz Biotechnology (Santa Cruz, CA); sheep polyclonal antibody against PP1β C subunit from Abcam (Cambridge, MA); mouse mAbs against α-tubulin and Golgi 58K from Sigma (St. Louis, MO), against histone H1 and PP2A C subunit from Upstate Biotechnology (Lake Placid, NY), and against GM130 from BD Biosciences (Franklin Lakes, NJ); horseradish peroxidase-conjugated anti-rabbit IgG and anti-mouse IgG from Promega (Madison, WI); fluorescein- or Texas red-labeled anti-rabbit and anti-mouse IgG from Vector Laboratories (Burlingame, CA); okadaic acid, cyclosporin A, and recombinant human PP1γ C subunit from EMD Biosciences (San Diego, CA); H-89 and orthovanadate from Sigma; CA, recombinant human PP1α, PP1β, PP2A, and PKA C subunit α from Upstate Biotechnology; recombinant human PKA C subunits β and γ from Biomol (Plymouth Meeting, PA); proteinase inhibitor mixture from Roche (Indianapolis, IN); and reagents for siRNA from Dharmacon Research (Lafayette, CO).
HepG2 cells (American Type Culture Collection, Manassas, VA) were grown on collagen I-coated dishes at 37°C in DMEM (GIBCO, Grand Island, NY) with 10% FBS (GIBCO), penicillin (100 units/ml), and streptomycin (100 μg/ml), in an atmosphere of 5% CO2/95% air.
Cells (≈5 × 105) collected by centrifugation after scraping were homogenized (20 strokes) in a Dounce tissue grinder (Wheaton Science, Millville, NJ) in 5 ml of TENDS buffer (20 mM Tris·HCl, pH 8.0, 1 mM EDTA, 1 mM NaN3, 1 mM DTT, 250 mM sucrose) containing protease inhibitors. Homogenates were centrifuged (800 × g, 10 min), supernatants were diluted to protein concentration of 1 mg/ml, and 750-μl samples were incubated for 1 h with 30 μl of protein G–Sepharose CL-4B bead slurry (50% vol/vol). Beads were discarded, ovalbumin (1 mg/ml) and antibodies against BIG1 (7.5 μg) or BIG2 (7.5 μg) were added, and, after incubation overnight, protein G–Sepharose CL-4B beads (30 μl) were added for 2 h. Beads collected by centrifugation were washed twice with homogenization buffer (1.5 ml) containing 150 mM NaCl and twice with the buffer indicated for each experiment (i.e., TENDS, PKA, PP1, PP2A). For in vitro phosphorylation or dephosphorylation, beads were dispersed in wash buffer (160 μl), and 20-μl samples of suspended beads were used.
For fractionation, postnuclear supernatants prepared as described for immunoprecipitation were centrifuged (105,000 × g, 1 h) to separate cytosol and membrane fractions, which were homogenized in homogenizing buffer (0.5 ml). Samples (20 μg) of proteins from whole homogenate, cytosol, membrane, and crude nuclear fractions were separated by SDS/PAGE in 4% or 4–12% gel and transferred to nitrocellulose membranes for reaction with antibodies against BIG1 (5 μg/ml), BIG2 (1 μg/ml), α-tubulin (0.5 μg/ml), Golgi 58K (1 μg/ml), or histone H1 (1 μg/ml). Secondary antibodies, horseradish peroxidase-conjugated goat anti-rabbit and horse anti-mouse IgG, were detected by using SuperSignal Chemiluminescent substrate (Pierce Biotechnology, Rockford, IL). Immunoreactive bands were quantified by densitometry by using Fuji Image Gauge software (version 4.0; Fujifilm, Tokyo, Japan).
For phosphorylation, samples of proteins immunoprecipitated, as described above, with antibodies against BIG1 or BIG2 were incubated at 30°C for 30 min without or with the indicated amount of PKA catalytic subunit in PKA buffer (20 mM Mops, pH 7.2, 25 mM β-glycerol phosphate, 5m M EGTA, 15 mM MgCl2, 1 mM DTT) containing 0.1 mM ATP in a total volume of 50 μl. For dephosphorylation, samples of immunoprecipitated BIG1 or BIG2 from cells that had been incubated for 3 h with 500 nM OA were incubated at 30°C for 30 min without or with the indicated phosphatase in PP1 buffer (50 mM Hepes, pH 7.0, 0.1 mM EGTA, 100 mM NaCl, 1 mM MnCl2, 0.025% Tween-20, 2 mM DTT) or PP2A buffer (50 mM Tris·HCl, pH 7.0, 0.1 mM CaCl2, 1 mM MgCl2, 1 mM DTT) in a total volume of 50 μl.
Before immunoprecipitation of BIG1 or BIG2, cells were incubated for 72 h with siRNA targeting human BIG2 or BIG1, respectively, and DharmaFECT4 siRNA transfection reagent according to the instructions of the manufacturer (Dharmacon Research), as described (10, 29). Washed beads with immunoprecipitated proteins from ≈5 × 105 cell were dispersed in 500 μl of TENDS buffer; a 25-μl sample was taken for Western blotting to quantify BIG1 and BIG2, and GEP assay reagents were added to remaining beads. To quantify GTPγS [guanosine 5′-γ-(thio)triphosphate] binding, recombinant ARF1–GST (25 pmol) was incubated (37°C) with 4 μM 5-(γ-[35S]thio)triphosphate ([35S]GTPγS; ≈2.5 × 106 cpm) in TENDS buffer containing 5 mM MgCl2, BSA (50 μg), and phosphatidylserine (20 μg) plus indicated immunoprecipitate in a total volume of 50 μl. Samples were transferred to nitrocellulose filters, which were washed six times (2 ml each) with ice-cold buffer (25 mM Tris·HCl, pH 8.0, 2.5 mM DTT, 5 mM MgCl2, 100 mM NaCl) and dried before radioassay in scintillation fluid (5 ml). GEP activity was calculated as the ratio of bound [35S]GTPγS (pmol) to the amount of BIG1 or BIG2 protein in each assay quantified by densitometry of Western blots.
HepG2 cells grown on type I collagen-coated four-well CultureSlides (BD Biosciences) were fixed for 20 min (all procedures at room temperature) with 3% paraformaldehyde in PBS, washed with PBS, permeabilized for 4 min with 0.1% Triton X-100 in PBS, washed with PBS, and incubated for 1 h with blocking buffer (5% goat serum, 5% horse serum, and 3% BSA in PBS). Cells were washed with PBS and incubated with anti-BIG1 (5 μg/ml) or anti-BIG2 (1 μg/ml), and anti-GM130 (2.5 μg/ml) Abs for 3 h, washed with PBS, and incubated for 1 h with FITC-conjugated anti-rabbit IgG and Texas red-conjugated anti-mouse IgG (each 1.5 μg/ml). After washing with PBS, cells were mounted in Vectashield (Vector Laboratories) and inspected with a confocal fluorescence microscope (LSM 510; Carl Zeiss, Jena, Germany).
We thank Dr. Toyoko Hiroi for providing ARF1-GST and Dr. Vincent C. Manganiello for important discussions and manuscript review. This work was supported by the Intramural Research Program of the National Institutes of Health, National Heart, Lung, and Blood Institute.
Author contributions: F.K., J.M., and M.V. designed research; F.K. performed research; F.K. and M.V. analyzed data; and F.K., J.M., and M.V. wrote the paper.
The authors declare no conflict of interest.
