J Biol Chem, Vol. 274, Issue 53, 37620-37628, December 31, 1999

The Annexin Protein Lipocortin 1 Regulates the MAPK/ERK Pathway^*

Louise C. Alldridge^§, Hayley J. Harris^§, Robin Plevin^¶, Robert Hannon, and Clare E. Bryant^§**

From the  Department of Clinical Veterinary Medicine, University of Cambridge, Madingley Road, Cambridge, CB3 OES, United Kingdom, the ^¶ Department of Physiology and Pharmacology, University of Strathclyde, Glasgow, Scotland, and the  Department of Biochemical Pharmacology, St. Bartholomew's and the Royal London School of Medicine and Dentistry, Queen Mary and Westfield College, University of London, Charterhouse Square, London EC1M 6BQ, United Kingdom

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Lipocortin 1 (annexin 1) is a calcium- and phospholipid-binding protein that modulates anti-inflammatory responses including those induced by lipopolysaccharide. To investigate the precise role of lipocortin 1 in regulating the lipopolysaccharide-induced signal transduction pathways, we generated stable RAW 264.7 macrophage cell lines expressing decreased and increased lipocortin 1 protein. Several RAW 264.7 clones with increased lipocortin 1 protein levels showed constitutive activation of the mitogen-activated protein kinase extracellular signal-regulated kinase, which was down-regulated following lipopolysaccharide treatment. Conversely, clones with decreased lipocortin 1 protein expression showed prolonged extracellular signal-regulated kinase activity, following lipopolysaccharide activation. Lipocortin 1 specifically regulates the components of the extracellular signal-regulated kinase pathway, since changes in lipocortin 1 protein expression had no affect on the related mitogen-activated protein kinases p38 and c-Jun N-terminal kinase. Lipocortin 1 modulated upstream components of the extracellular signal-regulated kinase pathway and associated with the adaptor protein growth factor binding protein. The downstream consequences of altered extracellular signal-regulated kinase activity were independent of the proinflammatory transcription factor nuclear factor kappa B. These data indicate that lipocortin 1 specifically regulates proximal signaling components of the extracellular signal-regulated kinase signal transduction pathway, resulting in the modulation of biochemical functions in RAW macrophages.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Lipocortin 1 (annexin 1) is a member of the annexin family of calcium and phospholipid-binding proteins of which 20 are known at present. Annexins are ubiquitous and expressed in diverse organisms (1-3). They are structurally homologous proteins with a highly conserved 70-amino acid repeat sequence. This comprises the core domain, which confers both the calcium and phospholipid binding properties of annexins. The diverse biological activities are dictated by the N-terminal domain, which is variable in both length and sequence (4). Annexins are cytosolic or associated with the membrane or the cytoskeleton in a calcium-dependent manner. Their precise functions are unclear; however, they have been implicated in a broad range of cellular functions, including signal transduction, DNA replication, cell transformation, ion channel formation, and apopotosis (5).

Lipocortin 1 is one of the more extensively studied annexins. It is a steroid-regulated protein and thus implicated in some of the beneficial actions of glucocorticoids including inhibition of cell proliferation, anti-inflammatory effects, the regulation of cell differentiation (6), and, more recently, membrane trafficking (7, 8). Lipocortin 1 may reduce inflammation by a number of mechanisms including reduced paw edema (9), decreased polymorphonuclear cell migration (10, 11), antipyretic effects (12), and an antiendotoxic action (13). The molecular mechanisms by which lipocortin 1 modulates these cellular responses are not known. Decreased inducible nitric-oxide synthase expression (13, 14) is believed to be important in the antiendotoxic action of lipocortin 1. Another effect of lipocortin 1 is the inhibition of cytosolic phospholipase 2 activity and hence arachidonic acid and lysophospholipid generation (6), which may explain some of its anti-inflammatory and antiproliferative functions.

Lipocortin 1 is a substrate for protein kinase C and protein-tyrosine kinases; this, coupled with multiple phosphorylation sites and calcium and phospholipid binding properties, may be indicative of a role in signal transduction as a means to effecting its pleiotropic physiological roles. The N-terminal domain of lipocortin 1 possesses a region with sequence similarity to a SH2¹ recognition domain. Thus lipocortin 1 may function in signaling complexes with proteins containing SH2 domains, although there is little evidence to support this theory (15).

In this study, the role of lipocortin 1 in LPS-mediated signal transduction in RAW 264.7 macrophages was assessed to discover how lipocortin 1 modulates cellular responses to LPS and protects against endotoxic shock. The major cell surface receptor for LPS on macrophages is the 55-kDa glycosylphosphatidyl inositol membrane-anchored glycoprotein CD14 (16). Recent work also suggests a role for the Toll-like receptor in the initial LPS signal transduction events (17). Subsequent events are not well defined, but strong evidence indicates a role for protein-tyrosine kinases in LPS signaling (18). Following LPS-activation, numerous proteins become phosphorylated, including members of the mitogen-activated protein kinase family: p38, extracellular signal-regulated kinases 1 and 2 (ERK1 and -2), and c-Jun N-terminal kinase (JNK) (19-22). Recent studies have suggested that each of these kinases play a pivotal role in expressing a number of proinflammatory molecules in macrophages (23, 24). This is the first study of the regulation of the mitogen-activated protein kinase family by the annexin lipocortin 1.

The putative downstream consequences of modulating these signaling cascades by altering lipocortin 1 protein levels are also important. LPS activates the nuclear translocation of the transcription factor NF-B in macrophages via the phosphorylation and ubiquitination of IB proteins. Many proinflammatory genes have NF-B recognition sequences in their promoters, including inducible nitric-oxide synthase (25), interleukin-1 (26), interleukin-6 (27), and tumor necrosis factor  (28), which together contribute to the pathophysiology of endotoxic shock. Some of these responses have been shown to be modulated by the action of lipocortin 1. Dexamethasome-induced lipocortin 1 has been shown to inhibit LPS-induced expression of inducible nitric-oxide synthase and subsequent nitric oxide production (13, 14). In addition, rats treated with a lipocortin 1 N-terminal peptide are protected against LPS-mediated endotoxic shock (13).

In the present study, we describe the effects of altered lipocortin 1 expression on LPS-induced activation of ERK1/2, JNK, and p38 MAPK. The p38 and JNK responses to LPS are unchanged by alteration of lipocortin 1 protein levels. Increased expression of lipocortin 1 protein results in constitutive activation of ERK, which is inhibited following LPS activation. Conversely, reduced lipocortin 1 protein expression leads to prolonged ERK activity following LPS-activation. Similar patterns of activation and inhibition are seen with the ERK-regulating kinase MEK, suggesting that lipocortin 1 acts on upstream components of this pathway. Modulation of early tyrosine phosphorylation events and the association of lipocortin 1 with Grb 2 also suggest a putative role for lipocortin 1 in proximal signaling events in the ERK cascade. The downstream consequences of the lipocortin 1-mediated changes in ERK1/2 phosphorylation do not appear to involve changes in the nuclear translocation of the NF-B.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Generation of Lipocortin 1 Expression Plasmids-- Total RNA was isolated from murine skin samples (Micro-scale Total RNA Separator Kit; CLONTECH). Full-length lipocortin 1 cDNA was amplified from murine RNA by reverse transcriptase-polymerase chain reaction using the following primers: forward, 5'-TAC TTC TCT AAA AAT GGC AAT; reverse, 3'-CAG AAT ATT TTA ACA AAA TAA at an annealing temperature of 57 °C.

The lipocortin 1 cDNA was subcloned into pCR 2.1 (Invitrogen) and sequenced. After verification of the sequence the lipocortin 1 cDNA was cloned into the pRc/RSV eukaryotic expression vector (Invitrogen) in either the sense or antisense orientation. The plasmids were cultured overnight, and DNA was purified using Wizard Maxipreps (Promega), cesium-banded, and dialyzed to remove endogenous LPS prior to transfection.

Cell Culture and Transfection-- LPS-responsive RAW 264.7 cells were cloned and cultured in Dulbecco's modified Eagle's medium containing 10% fetal calf serum supplemented with 2 mM glutamine, 200 units/ml penicillin, 100 µg/ml streptomycin. Cells were placed in serum-free medium and then transfected for 24 h using Lipofectamine (Life Technologies) and 5 µg of plasmid DNA. Cells were grown in complete medium for a further 24 h prior to splitting and subsequent dilution cloning in medium containing 500 µg/ml G418. Clones were selected and expanded prior to characterization of lipocortin 1 expression using Western blot analysis and ELISA assays. Expression of lipocortin 1 in the transfected cells was serum-dependent; therefore, all studies were performed using the same batch of serum.

Characterization of Cellular Lipocortin 1 Levels-- Since the levels of lipocortin 1 expression are critical to this study, the protein expression was measured by Western blot and ELISA. In addition, cell surface levels of lipocortin were assessed in order to ensure that changes in expression were reflected in this pool, which is thought to be important in the biological function of this protein (10, 13, 14). Cell surface lipocortin 1 was removed using an 1 mM EDTA wash containing protease inhibitors PMSF (1 mM), pepstatin A (0.5 µg/ml), and leupeptin (0.5 µg/ml). Total cell lysates were prepared in buffer containing 10 mM EDTA and 1% Triton X-100 with the same combination of protease inhibitors. Western blot analysis for lipocortin 1 was performed using samples with equivalent protein concentration, as described previously (13), using sheep anti-human lipocortin 1 antibody (a gift from J. Croxtall) at a concentration of 1:50,000. Protein bands were visualized by enhanced chemiluminescence (Amersham Pharmacia Biotech). Densitometric analysis of the Western blots was performed, and data are expressed as -fold changes in optical density. To further determine the levels of lipocortin 1 expression, a crude ELISA was developed. ELISA cell surface washes were prepared as for Western blot analysis and coated for 18 h at 4 °C onto 96-well plates (Falcon). After blocking with phosphate-buffered saline containing 0.1% Tween 20 and 5% milk powder for 1 h at 37 °C, lipocortin 1 levels were detected using the sheep anti-human lipocortin 1 antibody at a concentration of 1:2000. Following incubation with anti-sheep horseradish peroxidase-conjugated secondary antibody (1:2000), lipocortin 1 was detected using o-phenylenediamine dihydrochloride, and the color reaction was read at an OD of 492 nm. Optical density readings were normalized to viable cell number as assessed by cell counting using trypan blue exclusion.

Assay of Cellular Proliferation-- Cells were seeded at 3 × 10⁴ cells/35-mm plate and grown for 4 days in normal growth medium or medium supplemented with G418 for the transfected cell lines. At 4 days, cells were harvested and counted microscopically using a hemocytometer. Multiplication rate (r) and population doubling times (PDT) were calculated using the following equations, where n_H is the number of cells harvested at 96 h (t₂) and n_I is the number of cells seeded at time zero (t₁).

(Eq. 1)

(Eq. 2)

All experiments were counted in duplicate, and data represent three separate experiments.

ERK/MAPK Activity Assay-- Cells were lysed on ice in buffer containing 10 mM Tris/HCl, pH 7.5, 150 mM NaCl, 1% (v/v) Nonidet P-40, 1 mM EGTA, 1 mM EDTA, 1 mM Na₃V0₄, 1 mM PMSF, 50 mM NaF, 40 mM Na₄P₂O₇, and 0.5 µg/ml leupeptin and aprotinin. After clearing by centrifugation, ERK1/2 was immunoprecipitated for 1 h at 4 °C, using a polyclonal antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) immobilized on protein G-Sepharose. Immunoprecipitates were washed three times in lysis buffer and once in PBS. Kinase assays were conducted by incubating immunocomplexes in 30 µl of kinase buffer containing 30 mM Tris/HCl, pH 7.4, 20 mM MgCl₂, 2 mM MnCl₂, 10 µM ATP, and 7.5 µg of myelin basic protein (MBP) at room temperature for 30 min. The reaction was stopped by the addition of 30 µl of 2× Laemmli sample buffer and boiled for 5 min. 5-µl samples were resolved by 12% (w/v) SDS-PAGE and transferred to polyvinylidene difluoride membrane (Millipore Corp.) prior to probing with an antibody to phosphorylated MBP (Upstate Biotechnology). The phosphorylated substrate was visualized using enhanced chemiluminescence (Amersham Pharmacia Biotech). Blots from four separate experiments were quantified using Kodak 1D software. For the neutralizing antibody studies, cells were incubated with antisera to lipocortin 1 or sheep serum (1:60 dilution) for 4 h prior to cell lysis (13).

JNK and p38 Activity Assays-- Protein kinase activity of p38 was measured in affinity precipitates of p38 bound to recombinant GST-MAPKAP kinase-2 (29) immobilized on glutathione GSH-Sepharose beads (Amersham Pharmacia Biotech). Stimulated cells were lysed on ice in buffer containing 20 mM Tris/HCl, pH 7.4, 150 mM NaCl, 1% Triton X-100, 10% glycerol, 2 mM EDTA, 20 mM NaF, 2.5 mM -glycerophosphate, 0.2 mM PMSF, and 0.5 µg/ml leupeptin and aprotinin. Whole cell lysates were clarified by centrifugation and added to 1 µg of GST-MAPKAP kinase-2/GSH-Sepharose and circulated for 2 h at 4 °C. The bead complexes were recovered by centrifugation and washed three times in lysis buffer and once in kinase buffer (25 mM Hepes, pH 7.5, 20 mM MgCl, 5 mM -glycerophosphate, 0.1 mM Na₃V0₄, 2 mM dithiothreitol). The bead complexes were resuspended in kinase buffer containing 5 µCi of [-³²P]ATP and incubated for 30 min at 30 °C.

The reaction was stopped by the addition of 10 µl of 4× Laemmli sample buffer and boiled for 5 min. Samples were resolved by 11% (w/v) SDS-PAGE, the gels were dried, and the incorporation of [³²P]phosphate into GST-MAPKAP kinase-2 was detected using autoradiography. Autoradiographs from four separate experiments were quantified using Kodak 1D software.

The activity of JNK was quantified as for p38 activity except that cells were lysed in buffer containing 20 mM Hepes, pH 7.7, 50 mM NaCl, 1% (v/v) Triton X-100, 0.1 mM EDTA, 0.2 mM PMSF, 0.5 µg/ml aprotinin, and leupeptin and circulated with recombinant glutathione S-transferase-tagged truncated N terminus of cJun (GST-c-Jun-(5-89)) immobilized on GSH-Sepharose (30).

Electrophoretic Mobility Shift Assay-- Nuclear pellets were prepared from LPS-activated cells. Cells (3 × 10⁶) were washed twice with ice-cold PBS and then pelleted by centrifugation at 13,000 rpm for 1 min. The cell pellet was resuspended in 400 µl of buffer 1 (10 mM Hepes, pH 7.9; 10 mM KCl; 0.1 mM EDTA; 0.1 mM EDTA; 0.1 mM dithiothreitol; 0.5 mM PMSF; and 10 µg/ml leupeptin, pepstatin, and aprotinin) and put on ice to swell for 15 min. After the addition of 25 µl of 10% (w/v) Nonidet P-40, the samples were vortexed for 10 s and then centrifuged at 13,000 rpm for 20 s. The cell pellets were resuspended in 50 µl of buffer 2 (20 mM Hepes, pH 7.9, 25% (w/v) glycerol, 0.4 M NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM dithiothreitol, 0.5 mM PMSF, and 10 µg/ml leupeptin, pepstatin, and aprotinin) and, after light vortexing, were extracted on ice for 15 min. The extract was sonicated on ice for 30 s and then centrifuged at maximum speed for 15 min at 4 °C. The supernatant containing the nuclear fraction was stored at 70 °C.

The oligonucleotide sequences (Promega) were as follows. Murine intronic k-chain B site was 5'-AGT TGA GGG GAC TTT CCC AGC C (NF-B consensus), and AP1 consensus was 5' TTC CGG CTG ACT CAT CAA GCG. Oligonucleotides were labeled with [-³²P]ATP using T4 polynucleotide kinase (Promega). EMSAs were performed in 10 µl of reaction mixture containing 5 µg of nuclear extract, 5% glycerol, 1 mM MgCl₂, 0.5 mM EDTA, 0.5 mM dithiothreitol, 50 mM NaCl, 10 mM Tris-HCl, pH 7.5, 0.05 mg/ml poly(dI-dC)·poly(dI-dC), and 0.2 ng of DNA probe. The reactions were incubated for 20 min at room temperature. After the addition of 1 µl of gel-loading buffer (250 mM Tris-HCl, pH 7.5, 0.2% bromphenol blue, 40% glycerol), the reaction products were analyzed on 4% acrylamide gel, and the radioactive bands were visualized by autoradiography. To check for binding specificity to NF-B, two competition reactions were set up with either 1.75 pmol of cold NF-B oligonucleotide or 1.75 pmol of cold AP1 oligonucleotide. Autoradiographs from four separate experiments were quantified using Kodak 1D software.

MEK Activity Assay-- MEK activity was measured in vitro using histidine-tagged substrates following immunoprecipitation of the kinase. Cells were lysed in buffer containing 20 mM Tris/HCl, pH 7.4, 137 mM NaCl, 1% Triton X-100, 2 mM EDTA plus the same complement of protease and phosphatase inhibitors used for the ERK assay. Following centrifugation to clear the insoluble material, lysates were incubated with rabbit anti-MEK precoupled to protein G-Sepharose for 90 min at 4 °C. Immunoprecipitates were recovered by centrifugation, washed four times in lysis buffer, and then incubated for 15 min at 30 °C in 30 µl of buffer containing 25 mM Hepes, pH 7.4, 25 mM -glycerophosphate, 15 mM MgCl₂, 1 mM dithiothreitol, 50 µM ATP, 5 µCi of [-³²P]ATP, 1 µg of wild type ERK, and 200 µM epidermal growth factor receptor peptide substrate. The reaction was terminated on ice, and a proportion of the sample was assayed for peptide phosphorylation by ion exchange chromatography, while the remaining sample was resolved by 10% SDS-PAGE and exposed to autoradiographic film.

Tyrosine Phosphorylation-- Protein tyrosine phosphorylation induced by LPS (1 µg/ml) was detected by immunoblotting using anti-phosphotyrosine antibody 4G10 (TCS). Cells were lysed in buffer containing 50 mM NaCl; 1 mM EDTA; 10 mM Tris-HCl, pH 7.4; 1% Nonidet P-40; 1 mM EGTA; 1 mM PMSF; 1 µg/ml aprotinin, pepstatin, and leupeptin; 1 mM Na₄P₂O₇; 25 mM NaF. Proteins were resolved by 5-15% gradient SDS-PAGE and transferred to polyvinylidene difluoride (Millipore). Membranes were blocked in 3% (w/v) nonfat dried milk in PBS followed by incubation with 4G10 (1 µg/ml) for 4 h at room temperature. After three washes in PBS containing 0.05% Tween (PBS-T), the membranes were incubated with goat anti-mouse horseradish peroxidase conjugate diluted at 1:3000 for 1 h at room temperature. Phosphotyrosine-containing proteins were visualized using enhanced chemiluminescence (Amersham Pharmacia Biotech).

Analysis of Grb 2-associating Proteins-- Grb 2-associating proteins were isolated from untreated and LPS-activated cells using GST-Grb 2 fusion protein coupled to GSH-Sepharose. Cells were lysed in the buffer used for phosphotyrosine assays and cleared by centrifugation. Lysates were circulated with GST-Grb 2/GSH-Sepharose at 4 °C for 2 h. The bead complexes were recovered by centrifugation, washed four times in lysis buffer, and boiled in 2× Laemmli sample buffer for 8 min. Proteins were resolved and detected as for the phosphotyrosine assay. The same blots were reprobed with sheep antiserum to lipocortin 1 as described previously.

To further confirm the association between lipocortin 1 and Grb 2, immunoprecipitation of lipocortin 1 was conducted. Depletion of Grb 2 in lysates subjected to serial lipocortin 1 immunoprecipitation was conducted due to the close proximity of Grb 2 to the immunoglobulin light chain on blots of the immunoprecipitated complex. Lysates were prepared as above and circulated with the sheep antisera to lipocortin 1 coupled to protein G-Sepharose for 4 h at 4 °C. Four rounds of immunodepletion were conducted. The whole cell lysates and depleted lysates were denatured by boiling in 2× Laemmli sample buffer for 5 min and resolved by 12% SDS-PAGE. Following transfer to polyvinylidene difluoride, Grb 2 was detected using rabbit antisera (Santa Cruz Biotechnology), and lipocortin 1 was detected as descrbed previously.

Data Analysis-- Data were analyzed using Instat software. One-way analysis of variance was used for all statistical analysis with the exception of the r and PDT data, for which unpaired Student's t tests were conducted. Significance was taken at p < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Generation and Characterization of Stably Transfected Cell Lines-- After transfection, 10 antisense, 32 sense, and 7 control cell lines were selected by G418 resistance. Western blot analysis of lipocortin 1 expression in both the biologically relevant cell surface washes and the total cell lysates were used to identify the under- and overexpressing cell lines for expansion and analysis. Three cell lines transfected with antisense cDNA and three transfected with sense cDNA that showed marked under- or overexpression of lipocortin 1 were selected. All control cell lines produced equal levels of lipocortin 1 in amounts equivalent to untransfected RAW cells. Representative blots of both cell surface and total cellular lipocortin in over- and underexpressing cell lines with the respective densitometric analysis from 13 Western blots are shown in Fig. 1, A and B, respectively. As expected, the differences in lipocortin 1 expression between the cell lines were reflected and, interestingly, more striking, in the biologically relevant cell surface pool lipocortin 1, thus indicating that the altered cellular expression levels are conferred onto the cell surface, a pool that may be crucial to the study of the function of lipocortin 1. A crude ELISA for lipocortin 1 was developed to further verify the protein expression. The overexpressing cells showed an increase in lipocortin 1 expression of 181%, and the antisense cell line showed a slight decrease of 20% (compared with control cells).

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Characterization of lipocortin 1 expression. Western blot analysis of cell surface (A) and total cellular (B) lipocortin 1 was performed on equal quantities of protein using sheep anti-human lipocortin 1 antibody at a concentration of 1:50,000. The upper panels show representative blots, and the lower panels show data from densitometric analysis of 13 separate blots. CV, cells transfected with control empty vector showing control levels of lipocortin 1 expression; R1A, cells transfected with antisense lipocortin 1 showing reduced levels of lipocortin 1 expression; KF1A, cells transfected with lipocortin 1 cDNA showing increased lipocortin 1 expression. Lipocortin 1 was detected as a 37-kDa band, sometimes accompanied by a clipped form of the protein in cell surface samples.

Subsequent experiments were conducted on several clones demonstrating the maximum change in lipocortin 1 expression to ensure that the results obtained were not artifactual.

Overexpression of Lipocortin 1 Inhibits Cellular Proliferation in RAW Macrophages-- Lipocortin 1 is antiproliferative in a number of cell lines (6). To determine whether the changes generated in lipocortin 1 expression were biologically relevant, proliferation of the cell lines was assayed by growth curve analysis. Multiplication rate and population doubling times were calculated (see Table I). Cells with increased levels of lipocortin 1 had a significantly reduced multiplication rate of 0.66 generations/24 h, compared with 0.98 generations/24 h for cells transfected with the control plasmid. Increasing the levels of lipocortin 1 also significantly prolonged the population doubling time from 24.7 h in the control cells to 36.4 h. Cells with reduced levels of lipocortin 1 had a slightly higher multiplication rate (1.11 generations/24 h) and a reduced population doubling time (22.3 h), but these were not found to be statistically significant.

Alteration of Lipocortin 1 Protein Expression Changes the Constitutive and LPS-induced ERK/MAPK Activity-- LPS (1 µg/ml) affected ERK/MAPK activity in all cell lines (Fig. 2A). In the control cell lines (CV and RAW), ERK/MAPK becomes enzymatically active at 5 min after stimulation with LPS and returns to control levels after 30 min. The small amount of phosphorylated substrate in control lanes represents the presence of phosphorylated MBP in the preparation purchased from Sigma. Cells expressing reduced levels of lipocortin 1 protein (R1A) displayed a prolonged activation of ERK/MAPK with enzyme activity still present 30 min after exposure to LPS. The cells expressing increased levels of lipocortin 1 protein displayed constitutively activated ERK/MAPK, which was subsequently inhibited after stimulation with LPS. ERK/MAPK activity was inhibited by 50% at 5 min and by 90% at 30 min after exposure to LPS. Constitutive activation of ERK is only seen in cells that express excess lipocortin 1 protein (Fig. 2B). ERK protein shift assays were also conducted. Gel-retarded, phosphorylated ERK protein followed the same activation kinetics seen in the in vitro kinase assays (data not shown). Preincubation with a neutralizing antibody to lipocortin 1 resulted in abrogation of constitutive ERK activity and restoration of LPS-mediated ERK activation in cells overexpressing lipocortin 1 (Fig. 2C).

View larger version (23K): [in this window] [in a new window]   Fig. 2.   ERK activity in cells expressing different amounts of lipocortin 1 protein. A, ERK activity was determined by in vitro kinase assay using myelin basic protein as exogenous substrate. The top panel shows Western blots of phosphorylated myelin basic protein (P-MBP). Residual phosphorylation of the substrate is seen at time 0 due to the presence of phosphorylated myelin basic protein in the preparation from Sigma. The lower panel shows the same blots reprobed with anti-ERK antibody to show equivalent immunoprecipitation of ERK in LPS-treated and untreated cells. The graphs show -fold change in activity of ERK in the different cell lines, quantified by densitometric analysis of phosphorylated myelin basic protein. B, graph to show the activity of ERK in the untreated cell lines. Phosphorylated myelin basic protein was quantified by densitometric analysis. Only KF1A cells show ERK activity in the absence of LPS. n = 6; RAW, parental cell line; CV, cells transfected with control vector expressing control levels of lipocortin 1; R1A, cells underexpressing lipocortin 1; KF1A, cells overexpressing lipocortin 1. C, Western blots of phosphorylated myelin basic protein following in vitro ERK kinase assays in independently isolated clones. C6 and F3 are clones transfected to underexpress lipocortin 1 and show the same pattern of ERK activty as the original R1A clone (A). 16 and 26 are clones transfected to overexpress lipocortin 1 and show the same pattern of ERK activty as the original KF1A clone (A). D, Western blots of phosphorylated myelin basic protein following in vitro ERK kinase assays in lipocortin 1 overexpressing KF1A cells pretreated with either normal sheep serum (NSS) or lipocortin 1-neutralizing antiserum (LC-1).

To ensure that our observations were not the result of random clonal selection, we repeated these experiments in other clones either under- or overexpressing lipocortin 1. Constitutive activation of ERK and subsequent inhibition by LPS was also observed in two alternative lipocortin 1 overexpressing clones. Two separate lipocortin 1 underexpressing clones showed prolonged ERK activation following LPS activation (Fig. 2D).

The Effects of Lipocortin 1 Are Specific to the ERK/MAPK Pathway: Effect of Lipocortin 1 on LPS-induced p38 MAPK, JNK MAPK, and NF-B Activation-- LPS (1 µg/ml) induced enzymatic activation of p38 MAPK in all cell lines (Fig. 3A). The kinetics of activation by LPS was also similar in the cell lines. KF1A, CV, and RAW cells showed peak activity at 15 min post-LPS stimulation. The R1A cells displayed slightly slower kinetics with peak activity at 30 min after LPS stimulation.

View larger version (33K): [in this window] [in a new window]   Fig. 3.   p38 and JNK activation in cells expressing different amounts of lipocortin 1 protein. A, p38 activity was determined by in vitro kinase assay of affinity immobilized p38 to GST-MAPKAP kinase-2. The upper panel shows autoradiographs of phosphorylated MAPKAP-kinase 2 substrate. The lower panel shows -fold increases in p38 activity following LPS treatment, quantified by densitometric analysis of phosphorylated MAPKAP-kinase 2. B, JNK activity was determined by in vitro kinase assay of affinity immobilized JNK to GST-c-Jun-(5-89). The upper panel shows autoradiographs of phosphorylated c-Jun substrate. The lower panel shows -fold increases in JNK activity following LPS treatment, quantified by densitometric analysis of phosphorylated c-Jun. n = 4; RAW, parental cell line; CV, cells transfected with control vector expressing control levels of lipocortin 1; R1A, cells underexpressing lipocortin 1; KF1A, cells overexpressing lipocortin 1.

LPS (1 µg/ml) induced enzymatic activation of JNK in all cell lines. The kinetics of activation by LPS was also similar in the cell lines (Fig. 3B). All four cell lines showed peak activation at 15-30 min post-LPS stimulation. JNK activity was still present at 90 min after exposure to LPS in all cell lines.

Electrophoretic mobility shift assays were conducted in order to assess changes in lipocortin 1 expression results in downstream regulation of NF-B. Specific binding of NF-B was confirmed using unlabeled oligonucleotide containing the consensus binding sequence for NF-B (Sp) and unlabeled oligonucleotide containing the consensus binding sequence for AP1 (Nsp; Fig. 4A). LPS (1 µg/ml) caused NF-B translocation to the nucleus by 30 min, which was still detectable after 6 h, in all cell lines. The was no significant difference in the kinetics of NF-B activation in any of the cell lines (Fig. 4B). Western blots conducted to assess the phosphorylation and degradation of IB also showed no significant differences between the cell lines (data not shown).

View larger version (34K): [in this window] [in a new window]   Fig. 4.   NF-B activation in cells expressing different amounts of lipocortin 1 protein. NF-B activity was determined using electrophoretic mobility shift assay. A, autoradiograph shows binding specificity of NF-B using two competition reactions set up with either 1.75 pmol of cold NF-B oligonucleotide or 1.75 pmol of cold AP1 oligonucleotide in LPS-treated CV cells (, untreated cells; +, cells treated with LPS 1 µg/ml; Sp, cells treated with LPS 1 µg/ml with excess unlabeled oligonucleotide to NF-B; NSp, cells treated with LPS 1 µg/ml with excess unlabeled oligonucleotide to AP1). B, upper panel shows autoradiographs of LPS-induced NF-B translocation in cell lines expressing different levels of lipocortin 1. The lower panel of graphs shows the relative optical density of gel-retarded NF-B, quantified by densitometric analysis. n = 4; CV, cells transfected with control vector expressing control levels of lipocortin 1; KF1A, cells overexpressing lipocortin 1; R1A, cells underexpressing lipocortin 1.

Lipcortin 1 Affects the ERK/MAPK Pathway at a Site Upstream of MEK in the Signal Transduction Pathway-- Potentially, lipocortin 1 could modulate ERK activity by regulating key phosphatases involved in the dephosphorylation and inactivation of ERK or by affecting protein kinases upstream of ERK in the signal transduction cascade. To determine by which of these modes lipocortin 1 regulates ERK, we assayed MEK activity in the different clones. Cells overexpressing lipocortin 1 showed constitutive activation of MEK, which was inhibited by LPS, thus reflecting the profile of ERK activity in these cells. Similarly, cells with reduced levels of lipocortin 1 protein showed prolonged MEK activity (Fig. 5A).

View larger version (23K): [in this window] [in a new window]   Fig. 5.   MEK activity in cells expressing different amounts of lipocortin 1 protein. MEK activity was determined by in vitro kinase assay using histidine-tagged substrates following immunoprecipitation of the kinase. Autoradiographs of phosphorylated substrate are shown in A. In addition, phosphorylation of the peptide was assayed by ion exchange chromatography. Peptide phosphorylation (dpm) is displayed graphically in B. n = 4; CV, cells transfected with control vector expressing control levels of lipocortin 1; R1A, cells underexpressing lipocortin 1; KF1A, cells overexpressing lipocortin 1.

Analysis of peptide phosphorylation by ion exchange chromatography was also conducted. In overexpressing cells, the peptide phosphorylation was detected in the untreated cells; however, following LPS treatment, minimal phosphorylation of the substrate was observed. In the underexpressing cells, phosphorylated peptide was detected following a 5-min exposure to LPS, and the peptide remained phosphorylated after 10 min (Fig. 5B). Thus, the phosphorylation of the peptide followed the same pattern as the in vitro MEK assay and, more significantly, the ERK assay.

Alteration of Lipocortin 1 Protein Expression Modifies the Pattern of LPS-induced Protein Tyrosine Phosphorylation-- Different levels of lipocortin 1 protein affect qualitative and quantitative differences in protein tyrosine phophorylation. Over 13 separate experiments were conducted. The most consistent clone-specific changes in tyrosine phosphorylation occurred in higher molecular mass proteins, in particular two proteins with approximate molecular masses of 150 and 80 kDa, indicated with arrows on the representative blot shown in Fig. 6. The former is tyrosine-phosphorylated only in cells overexpressing lipocortin 1 following LPS-stimulation for 1 min, returning to basal phosphorylation at 5 min. In contrast, the 80-kDa protein is dephosphorylated 5 min post-LPS stimulation. Again, this reflects a protein tyrosine phosphorylation pattern unique to the cells with excess lipocortin 1 expression.

View larger version (43K): [in this window] [in a new window]   Fig. 6.   LPS-induced protein tyrosine phosphorylation in cell lines expressing different amounts of lipocortin 1 protein. Protein tyrosine phosphorylation in cell lines expressing different levels of lipocortin 1 was assessed by Western blotting using the anti-phosphotyrosine antibody 4G10. Different patterns of protein tyrosine phosphorylation were seen in higher molecular weight proteins. Two proteins with the approximate molecular masses of 150 and 80 kDa, indicated with arrows on the representative blot, show consistently altered patterns of tyrosine phosphorylation and dephosphorylation following LPS treatment of KF1A cells. n = 13; CV, cells with control levels of lipocortin 1; R1A, cells underexpressing lipocortin 1; KF1A, cells overexpressing lipocortin 1.

Lipocortin 1 Forms a Signaling Complex That Contains Grb 2-- Data from the MEK and phosphotyrosine assays along with the membrane localization of lipocortin 1 suggested that the site at which lipocortin 1 modifies the ERK cascade is proximal. In addition, since lipocortin 1 has a putative SH2 bindng domain (15), it seemed relevant to assess the tyrosine-phosphorylated proteins that bind to Grb 2 and whether lipocortin 1 itself formed associations with Grb 2 in the different clones (Fig. 7). Quantitative and qualitative differences in tyrosine-phosphorylated proteins associating with the GST-Grb 2 fusion protein were only observed in the cells that expressed increased lipocortin 1 levels. Phosphotyrosine proteins of 130 and 58 kDa were abundantly associated in the overexpressing cells. No significant changes in the profile of Grb 2-associated tyrosine phosphorylated proteins were observed following LPS activation.

View larger version (49K): [in this window] [in a new window]   Fig. 7.   The association of phosphotyrosine-containing proteins and lipocortin 1 with Grb 2. Cell lysates from LPS-activated cells were incubated with GST-Grb 2 fusion protein coupled to GSH-Sepharose beads. The resulting complexes were separated by PAGE, transferred and probed with the anti-phosphotyrosine antibody 4G10 (upper panel). The same blot was then probed with anti-lipocotin antisera (lower panel). n = 4; CV, cells transfected with control vector expressing control levels of lipocortin 1; KF1A, cells overexpressing lipocortin 1; R1A, cells underexpressing lipocortin 1.

Reprobing of these blots with antiserum to lipocortin 1 revealed that it forms an association with the GST-Grb 2 fusion protein. A similar band was not detected on the phosphotyrosine blot, indicating that this pool of lipocortin 1 is not tyrosine-phosphorylated. Again, no significant changes in the amount of lipocortin 1 associated with the GST-Grb 2 fusion protein were seen following LPS activation except for a slight decrease after 5 min in the overexpressing cells.

To confirm that lipocortin 1 and Grb 2 form a complex, lipocortin 1 immunoprecipitates were denatured, resolved, and immunoblotted for Grb 2. A band at the correct size of 25 kDa was obscured by cross-reaction of the secondary antibody with the immunoglobulin light chain. Therefore, lysates were depleted of lipocortin 1 by successive immunoprecipitations and blotted for Grb 2 to detect simultaneous depletion (Fig. 8). Immunodepletion of lipocortin 1 resulted in the simultaneous depletion of Grb 2, confirming their association.

View larger version (36K): [in this window] [in a new window]   Fig. 8.   Immunodepletion of lipocortin 1 results in significant depletion of Grb 2. Western blot is shown of whole cell lysates (WCL) and cell lysates depleted of lipocortin 1 by four successive rounds of immunoprecipitation (DL). The top half was probed with anti-lipocortin antiserum (upper panel), and the bottom was probed with anti-Grb 2 antiserum (lower panel). n = 4; CV, cells transfected with control vector expressing control levels of lipocortin 1; KF1A, cells overexpressing lipocortin 1; R1A, cells underexpressing lipocortin 1.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In this study, we have demonstrated that lipocortin 1 specifically modulates the ERK signaling cascade at an upstream site probably by associating with key signaling components including the adaptor protein Grb 2. Increasing the expression of lipocortin 1 leads to constitutive activation of ERK1/2 kinase in RAW macrophages. LPS activation of these cells results in inhibition of this activity. Cells with decreased lipocortin 1 protein show prolonged activation of ERK1/2 kinase in response to LPS. Identical patterns of activation are seen with the ERK1/2-activating kinase MEK, suggesting that lipocortin 1 acts upstream in the signaling cascade. The finding that lipocortin 1 modulates early tyrosine phosphorylation events and associates with Grb 2 also suggests a proximal function in this pathway. These alterations in activation appear to be specific to the ERK/MAPK pathway, since both p38 and JNK activity are relatively unaltered by changes in lipocortin 1 protein expression. Activation of NF-B, a downstream regulator of several LPS responses potentially modulated by the action of ERK (31-35), was also unaffected by lipocortin 1, in untreated and LPS-stimulated cells. Thus, the downstream consequences of lipocortin 1-mediated changes in ERK activity are independent of NF-B activation, suggesting that there is no direct interaction between the ERK pathway and the activation of NF-B in RAW macrophages. The changes observed in the ERK pathway result from alteration of lipocortin 1 expression and are not due to clonal selection of RAW macrophages, since assays were repeated using independent clones, and identical results were obtained.

We utilized a cellular model that constitutively over- or underexpressed lipocortin 1, confirmed by both Western blot analysis and ELISA. The validity of this model, in the study of lipocortin 1 function, was confirmed by two important observations. First, clones with altered levels of lipocortin 1 expression are able to process the protein in the usual manner, since the changes are also observed on the cell surface. Antibody studies show that this surface-associated pool is functionally significant (10, 13, 14), and interestingly, in our cell lines, changes in lipocortin 1 expression are more evident in the cell surface-associated protein fraction. Second, the exogenous lipocortin 1 is biologically functional, as shown by the significant reduction in cellular proliferation of cells transfected with sense cDNA. The cells under-expressing lipocortin 1 showed a slight increase in proliferation; however, this was not statistically significant. The reduced amount of lipocortin 1 protein expressed by these cells is probably sufficient to permit proliferation at a rate close to that of control cells.

The constitutive activation of ERK in cells overexpressing lipocortin 1 is a surprising result. To our knowledge, constitutive activation of ERK has only been reported in oncogenically transformed cells (36) and in discrete microtubule-associated pools in PC12 cells (37). Lipocortin 1 may act by constitutive activation of kinases upstream of ERK or by inhibition of key phosphatases that deactivate the enzyme. The constitutive activity of MEK, the differences observed with Grb 2-associated tyrosine-phosphorylated proteins, and the association of lipocortin 1 with Grb 2 support the former mode of ERK regulation. The constitutive activation of ERK may reflect the importance of lipocortin 1 in modulating cellular proliferation (38).

Following LPS activation, the presence of excess lipocortin 1 protein results in specific inhibition of the ERK pathway, while decreased lipocortin 1 protein expression leads to prolonged LPS activation of ERK. This inverse relationship between the amount of lipocortin 1 expressed and the activity of ERK, following LPS activation, may be indicative of the anti-inflammatory effects of this protein. The data from untreated and LPS-activated cells indicate that lipocortin 1 has a dual role in modulating the ERK cascade depending on LPS activation. This may reflect the activation and/or phosphorylation status of the lipocortin 1 protein itself, which is tyrosine phosphorylation after 1 min of exposure to LPS in RAW macrophages.²

Interestingly, the constitutive activation of ERK in cells overexpressing lipocortin 1 is abrogated by antibody adsorption of the protein. Similarly, antibody neutralization of lipocortin 1 restores LPS induction of ERK activity. These data confirm the functional importance of cell surface-associated lipocortin 1 (also demonstrated for the related protein annexin 2 (39)) and supports our hypothesis that this annexin I is regulating the ERK pathway in RAW macrophages. Since our results suggest that lipocortin 1 acts intracellularly, as part of a signaling complex, it is possible that the transportation of external lipocortin 1 into the cell is a critical event. In addition, inhibition of external lipocortin 1 alludes to a paracrine and/or an autocrine function in the regulation of ERK. However, since our cell lines are macrophage-derived, the internalization of the antibody complex leading to effects on intracellular lipocortin 1 could complicate interpretation of this part of the study.

Analysis of upstream signaling, in the LPS-activated clones, has confirmed a role for lipocortin in the proximal signaling events of the ERK cascade. The ERK kinase MEK was found to follow an identical pattern of activation and inhibition. Therefore, signaling components upstream of ERK in the cascade are regulated by lipocortin 1 and not phosphatases that can directly regulate ERK activity. Alteration of tyrosine phosphorylation is usually the earliest biochemical change detected in many signaling systems including LPS activation. The differences observed in the pattern of tyrosine phosphorylation, therefore, suggest that lipocortin 1 functions early in LPS-induced signaling events. This would agree with the observed localization of lipocortin 1 along the plasma membrane in macrophages (7). It is of interest that an 80-kDa tyrosine phosphoprotein is dephosphorylated during LPS activation, while a 150-kDa protein becomes tyrosine-phosphorylated following LPS activation. This may indicate a role for a protein-tyrosine phosphatase as well as protein-tyrosine kinases in lipocortin 1 modulation of LPS-induced signal transduction in macrophages.

Since lipocortin 1 contains a motif with a degree of sequence homology to a SH2-binding domain (15), the function of Grb 2 was assessed in the clones. Grb 2, a small adaptor protein, is usually found complexed with the Ras exchange factor SOS, via its SH3 domains. This complex mediates Ras activation when recruited to the plasma membrane by SH2-mediated binding of phosphotyrosine motifs on several receptor tyrosine kinases or other adaptor proteins associated with these receptors (40-44). This results in Ras activation of Raf and a cascade of phosphorylation events leading to ERK activation. The profile of tyrosine-phosphorylated proteins associating with Grb 2 was analyzed as a preliminary screen. Although the cells overexpressing lipocortin 1 show a different complement of phosphotyrosine proteins bound to Grb 2, the profile does not alter significantly following LPS activation. This may suggest, again, a dual role for lipocortin 1 in the regulation of ERK in untreated and LPS-activated cells. Lipocortin 1 may not modulate the binding of proteins with phosphotyrosine motifs to the SH2 domain or alternatively may function downstream of Grb 2 following LPS activation. Since lipocortin 1 was found to bind the Grb 2 fusion protein and this association was confirmed by simultaneous depletion of Grb 2 with lipocortin 1 immunoprecipitations, it is probable that it functions at this point in the cascade. Further work will determine the precise nature of this association and the composition of the protein complex in untreated and LPS-treated cells.

It is of particular interest that changes in the expression of lipocortin 1 only affect the activity of the ERK signaling cascade and not the related JNK and p38 MAPK pathways. There exists a degree of overlap in the regulation of MAPK by their MAPK kinases and their MAPK kinase kinases. In addition, all of the MAPKs possess overlapping substrate specificities, each phosphorylating the minimum sequence X(S/T)P, where  is a proline or aliphatic. Mechanisms must therefore exist to achieve signaling specificity to ensure the correct biological response. It is thought that protein-protein interactions between the components of a pathway increase this specificity. Recently, two mammalian scaffold proteins, MP1 (45) and JIP-1 (46), have been identified that help route the ERK and the JNK pathways, respectively. Identification of components complexed with lipocortin 1 and Grb 2 may show that lipocortin 1 plays a similar role, specifically enhancing the ERK signal transduction pathway and excluding others.

The multifactoral downstream effects of the action of lipocortin 1, including antiproliferative and antiendotoxic effects, may be translocated by the ERK signaling pathway. The ERK cascade serves to couple a wide range of stimuli to the regulation of a variety of cellular functions. Activation of ERK1 and -2 results in phosphorylation of cytoplasmic and nuclear substrates, including epidermal growth factor receptor, ribosomal S6 kinase II (Rsk), Elk-1, c-Jun, and c-Myc (47, 48). Although ERK activation has been traditionally associated with increased cellular proliferation, recent work suggests that sustained ERK activation, as observed in clones overexpressing lipocortin 1, can lead to expression of proteins such as p21^cip1/waf1, culminating in cell cycle arrest (49-51). Our growth curve data indicate that the cells overexpressing lipocortin 1 protein have a reduced proliferation rate that is likely to be related to the constitutive activation of ERK1 and -2. Hence, lipocortin 1 may reduce proliferation by inducing the expression of p21^cip1/waf1 via constitutive activation of ERK. Preliminary results have shown that the lipocortin 1 overexpressing cells do express p21^cip1/waf1.

The transcription factor NF-B plays a central role in the regulation of LPS-inducible genes implicated in the inflammatory response, including inducible nitric-oxide synthase and various cytokines (52). Interestingly, lipocortin 1 down-regulates LPS-induced inducible nitric-oxide synthase expression (13, 14). In this study, we found that lipocortin 1 did not affect basal or LPS-induced activation of NF-B. Thus, lipocotin 1 does not appear to influence inflammatory gene expression by modulation of the NF-B pathway. In addition, several recent publications have suggested a link between ERK activity and NF-B activation in macrophages (34) and other cell types (33, 35). Here, since lipocortin 1-mediated inhibition of LPS-induced ERK had no effect on the nuclear translocation of NF-B in our cells, we hypothesize that the ERK pathway does not directly influence LPS-induced NF-B activation in RAW cells. This observation suggests that lipocortin 1 may be important in directing the signal preferences of ERK to specific substrates with distinct biological roles. Our data contrast with the studies that suggest a potential link between ERK and NF-B activity. However, these observations were made using either stimuli other than LPS, potentially nonselective inhibitors, or transient transfection of constitutively active ERK (which may drive the expression and secretion of cytokines that can subsequently activate NF-B). In RAW macrophages, the potential influence of the ERK pathway on LPS-induced NF-B activation needs further investigation.

Since lipocortin 1 is an important antiendotoxic and anti-inflammatory protein (6, 13), our data have important implications in explaining how lipocortin 1 functions pathophysiologically. We have shown that lipocortin 1 specifically regulates the activity of the ERK cascade but has no effect on the LPS-induced JNK, p38 MAPK, or NF-B activation in RAW macrophages. We hypothesize that modulation of the ERK pathway by lipocortin 1 occurs at a proximal site involving association with key upstream signaling components such as Grb 2, possibly to form signaling complexes. Determining how lipocortin 1 regulates the ERK pathway as well as potential downstream targets should explain the mechanism of the cellular effects of this protein and suggest novel targets for anti-inflammatory drugs.

	ACKNOWLEDGEMENT

We thank Suzi Keating for preparation of the lipocortin 1 cDNA.

	FOOTNOTES

^* 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.

^§ Supported by a Wellcome Research Career Development Fellowship award (to C. E. B.).

^** To whom correspondence should be addressed. Fax: 44-1223-337610; E-mail: ceb27@cus.cam.ac.uk.

² L. Alldridge, unpublished observation.

	ABBREVIATIONS

The abbreviations used are: SH2, Src homology 2; ERK, extracellular signal-regulated kinase; Grb 2, growth factor-binding protein 2; GST, glutathione S-transferase; IB, inhibitory B; JNK, c-Jun N-terminal kinase; MAPK, mitogen-activated protein kinase; MEK, MAPK/ERK-activating kinase; MBP, myelin basic protein; MAPKAP, MAPK-activated protein; NF-B, nuclear factor B; ELISA, enzyme-linked immunosorbent assay; PMSF, phenylmethylsulfonyl fluoride; PBS, phosphate-buffered saline; PAGE, polyacrylamide gel electrophoresis; LPS, lipopolysaccharide.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES