(Received for publication, February 14, 1997, and in revised form, April 23, 1997)
From the Critical Care Medicine Department, National Institutes of Health, Bethesda, Maryland 20892
Using a two hybrid system screen of a human cDNA library, we have found that p11, a unique member of the S100 family of calcium-binding proteins, interacts with the carboxyl region of the 85-kDa cytosolic phospholipase A2 (cPLA2). p11 synthesized in a cell-free system interacts with cPLA2 in vitro. The p11-cPLA2 complex is detectable from a human bronchial epithelial cell line (BEAS 2B). Furthermore, p11 inhibits cPLA2 activity in vitro. Selective inhibition of p11 expression in the BEAS 2B cells by antisense RNA results in an increased PLA2 activity as well as an increased release of prelabeled arachidonic acid. This study demonstrates a novel mechanism for the regulation of cPLA2 by an S100 protein.
Phospholipase A2s (PLA2,1 EC 3.1.1.4) are a group of enzymes that catalyze the hydrolysis of the sn-2-ester bond of phospholipids, resulting in the production of free fatty acid and lysophospholipids. These lipid products may serve as intracellular second messengers or can be further metabolized to potent inflammatory mediators. The release of arachidonic acid (AA) from membranes by PLA2 and its subsequent conversion into leukotrienes, prostaglandins, and other eicosanoids play a key role in the process leading to inflammation. Several mammalian PLA2s, including the 14-kDa secreted form PLA2 and the 85-kDa cytosolic PLA2 (cPLA2), have been proposed to mediate AA release and eicosanoid generation in various cell types (1-3).
The mammalian low molecular mass 14-kDa PLA2s mainly include two types, type I and type II. Although the type I PLA2 is mainly found in pancreatic secretion, it has also been found in other mammalian tissues, and the physiological role of this extra-pancreatic type I PLA2 remains unclear. The type II PLA2 has been suggested to be involved in the pathogenesis of inflammatory reactions (1-2). It is found associated with several cells and tissues or is present extracellularly when released in response to proinflammatory mediators such as interleukin-1, interleukin-6, or tumor necrosis factor. The soluble type II 14-kDa PLA2 found in inflammatory fluids, tissue exudates, or serum has been suggested to propagate inflammation through membrane digestion and liberation of proinflammatory lipid mediators. The cell-associated type II PLA2 has been assumed to release AA for lipid mediator generation (1-3). However, because the 14-kDa PLA2 lacks selectivity for the sn-2 fatty acids of phospholipids and requires much higher Ca2+ (mM) than normal cellular Ca2+ level (nM-µM) for activation, it was not certain whether this type of PLA2 contributes to the release of AA and production of eicosanoids. Studies using antisense inhibition and selective inhibitors of different forms of PLA2s have demonstrated that the type II PLA2 participates in AA release and prostaglandin generation in the P388D1 macrophages (4-6). These findings have established an important role of the 14-kDa PLA2 for AA metabolism.
The high molecular mass 85-kDa cPLA2 exhibits a preference for arachidonic acid in the sn-2 position of the substrate phospholipids (1-3, 7-10). The enzyme activity is regulated by phosphorylation (11-14), G-proteins (15), and physiologically relevant concentrations of Ca2+ (7-9, 16). These characteristics make the 85-kDa cPLA2 a strong candidate for participation in the liberation of AA from membrane phospholipids for subsequent metabolism to eicosanoids. The cPLA2 is expressed in a wide variety of cells (1-3). Although the cPLA2 gene lacks a TATA box in its promoter (17-19), its expression is induced by various cytokines in several cell types (20-26). The cPLA2 has two functionally distinct regions, the NH2-terminal region (amino acids 1-178) containing the homologous Ca2+-dependent lipid binding motif (amino acid 36-81) and the COOH-terminal region (amino acids 179-749) containing the catalytic domain (8, 27). The NH2-terminal region Ca2+-dependent lipid binding motif mediates the Ca2+-dependent translocation of cPLA2 from cytosol to membrane substrate. The COOH-terminal region mediates the Ca2+-independent enzyme catalysis, which likely involves Ser-228 at the active site (28). Several phosphorylation sites (Ser-437, Ser-454, Ser-505, and Ser-727) have also been identified in the COOH-terminal region (29). Ser-505, located at the consensus site for mitogen-activated protein kinase, is phosphorylated by mitogen-activated protein kinase (12-13). Other protein kinases such as protein kinase C and cyclic AMP-dependent kinase (protein kinase A) also cause phosphorylation of cPLA2 in vitro (12-13, 30). Thus, both the calcium-dependent translocation mediated by the NH2-terminal Ca2+-dependent lipid binding domain and phosphorylation of the COOH-terminal region play an important regulatory role in the activation of cPLA2 and regulation of AA release.
In an effort to further investigate the regulation of the 85-kDa cPLA2, this study was designed to identify proteins interacting with the NH2-terminal or COOH-terminal region of cPLA2 protein by screening a human cDNA library using the yeast two hybrid system. An 11-kDa protein (p11), the unique member of S100 family of calcium-binding proteins, was found to interact with the COOH-terminal region of cPLA2. Further studies demonstrated that p11 associated with cPLA2 and inhibited the cPLA2 activity in vitro. The p11-cPLA2 complex was precipitated from a human bronchial epithelial cell line (BEAS 2B cell) expressing both p11 and cPLA2. Reduction of p11 expression in the BEAS 2B cells by antisense RNA reversed the inhibition of p11 on cPLA2, which resulted in an increased PLA2 activity as well as an increased AA release. This study revealed a previously unrecognized mechanism of cPLA2 regulation by the S100 protein family.
Plasmid pGBT9 (CLONTECH, Palo Alto,
CA), containing the sequences for the GAL4 DNA-binding domain (amino
acids 1-147) as well as sequences for tryptophan (TRP1),
was used to express the fusion proteins. Two cPLA2 cDNA
fragments (540 bp corresponding to the NH2 terminus and
1707 bp corresponding to the COOH terminus) were amplified by
polymerase chain reaction using human cPLA2 cDNA plasmid as the template (8). The obtained cPLA2 cDNA
fragments were cloned into pGBT9 vector in the correct orientation and
with the correct reading frame to express the GAL4-cPLA2
fusion proteins. To obtain the cPLA2 cDNA fragments,
two primer pairs were synthesized according to the cPLA2
cDNA sequence (8). The primer pair for the 540-bp cPLA2
cDNA fragment was composed of the following sequences: 5
primer-GTCGACTGATGTCATTTATAGATCCTTACCAGCAC (starting from
the translation start site; corresponding to the +1 to +27 of the coding region; containing the underlined SalI site and TG
dinucleotides for in frame cloning); 3
primer-GTCGACTCCTTCACTATTCTTTGGACCCAAGAGTTT (corresponding
to the +540 to +511 of the coding region; containing the underlined
SalI site). The primer pair for the 1707-bp
cPLA2 cDNA fragment was composed of the following
sequences: 5
primer-GTCGACTGTTGCATTCTGCACGTGATGTGCCTGTG (corresponding to +541 to +567 of the coding region; containing the
underlined SalI site and TG dinucleotides for in frame
cloning); 3
primer-GTCGACTGCTTTGGGTTTACTTAGAAACTCCTTGTT
(corresponding to the +2247 to +2218 of the coding region and
containing the underlined SalI site). The cDNA was
amplified via polymerase chain reaction utilizing Thermus
aquaticus DNA polymerase. The amplified fragments were initially
cloned into the pCRTM3 vector using the TA cloning kit
(Invitrogen). The inserts were then released from the
pCRTM3 vector by restriction enzyme digestion with
SalI and were cloned into the pGBT9 vector. The identity and
orientation of the constructs were confirmed by DNA sequencing. The
plasmid containing the 540-bp cPLA2 cDNA fragment
corresponding to the NH2-terminal cPLA2 protein was designated as plasmid pGBT9-cPLA2N. The plasmid containing the
1707-bp cPLA2 cDNA fragment corresponding to the
COOH-terminal cPLA2 protein was designated as plasmid
pGBT9-cPLA2C. These plasmids were used to transform host HF7c cells.
The transformants containing the pGBT9-cPLA2N or pGBT9-cPLA2C were
selected for tryptophan prototrophy and examined for
-galactosidase
activity using a filter assay. The
-galactosidase activity was
undetectable in colonies containing only pGBT9-cPLA2N or pGBT9-cPLA2C.
The HF7c cells containing pGBT9-cPLA2N or pGBT9-cPLA2C were then lysed for Western blotting analysis using the rabbit anti-cPLA2
antibody (Genetics Institute, Boston, MA). The identity of the
expressed fusion proteins was confirmed by the observation that both
the NH2-terminal cPLA2 fusion protein
(approximately 40 kDa) and COOH-terminal cPLA2 fusion
protein (approximately 90 kDa) reacted with the polyclonal anti-cPLA2 antibody.
A
human leukocyte MATCHMAKER cDNA library constructed in pGAD10
plasmid (CLONTECH) was screened to identify
proteins interacting with cPLA2. The Saccharomyces
cerevisiae HF7c reporter strain was used for transformation with
the library and plasmid pGBT9-cPLA2N or plasmid pGBT9-cPLA2C by the
lithium/acetate method (31). The HF7c yeast cells were sequentially
transformed with pGBT9-cPLA2N or pGBT9-cPLA2C followed by
transformation with DNA from the MATCHMAKER human leukocyte cDNA
library. Double transformants were plated in the SD synthetic medium
lacking histidine, tryptophan, and leucine at a density of
approximately 10,000 colonies per 150-mm plate (20 plates for each of
pGBT9-cPLA2N and pGBT9-cPLA2C). Colonies appeared after incubation at
30 °C for 3 days. The colonies were transferred to VWR grade 413 paper filters and assayed for -galactosidase activity using
5-bromo-4-chloro-3-indolyl-
-D-galactoside (X-Gal) as
substrate. Filters were placed in liquid nitrogen for 30 s, thawed
at room temperature, and then incubated in buffer containing X-Gal. If
transformants expressed
-galactosidase activity, the X-Gal was
cleaved and colony replicates turned blue. Positive colonies were
purified by restreaking on selection medium, and well isolated colonies
were reassayed for
-galactosidase activity.
Plasmids encoding the candidate interacting GAL4 activation domain/cPLA2 binding protein were extracted from the positive HF7c clones using a previously described method (32). The extracted plasmid DNA was transformed to an Escherichia coli strain (HB101) carrying a leuB mutation to select the activation domain hybrid plasmid that contains yeast LEU2 gene. The transformed E. coli cells were plated on M9 agar medium containing a 1 × mixture of amino acids (lacking leucine), 50 µg/ml ampicillin, 40 µg/ml proline, and 1 mM thiamine HCl. Individual colonies were isolated, and the plasmid DNA was extracted using a standard plasmid mini-prep procedure (33).
Sequence Analysis of pGAD10-cDNA ClonesInitial
sequence of the cDNA inserts was obtained by dideoxy sequencing
using a 17-mer oligonucleotide primer (5-TACCACTACAATGGATG-3
), which
corresponds to the GAL4 activation domain sequences and reads toward
the junction of the GAL4 activation domain and the cloned candidate
interacting protein. Specific oligonucleotide primers corresponding to
the insert sequence were subsequently generated to obtain the
full-length sequence of the inserts.
To construct plasmid
for p11 protein expression in bacteria, human p11 cDNA was
amplified by reverse transcription and polymerase chain reaction using
human lung poly(A)+ RNA. The primer pair, synthesized
according to the human p11 cDNA sequence (34), was composed of the
following sequences: 5
primer-GGATCCCCATCTCAAATGGAACACGCCATG (corresponding to
115-138 of the cDNA sequence; adjacent to the translation start
site; containing the underlined BamHI site) and 3
primer-CTCGAGTATCAGGGAGGAGCGAACTGCTCA (corresponding to
413-436 of the cDNA sequence; located in 3
-untranslated region;
containing the underlined XhoI site). The amplified
fragments were initially cloned into the pCRTM3 vector
using the TA cloning kit. The inserts were then released from the
pCRTM3 vector by restriction enzyme digestion
(BamHI and XhoI) and subsequently cloned into the
prokaryotic expression vector pET30a (Novagen), giving rise to
p11-pET30a. The identity and orientation of the construct was confirmed
by DNA sequencing. The pET30a-p11 vector expressed p11 protein as a
fusion protein containing a stretch of 6 consecutive histidine residues
(His-tag) at the NH2 terminus. The His-tag, removable by
digestion with protease (e.g. enterokinase), was used for
efficient purification of the recombinant p11 proteins under gentle,
native conditions for maintaining activity of soluble proteins (35).
The plasmid was transformed into E. coli strain BL21 (DE3),
a lon mutant strain containing the T7 polymerase under the
control lacUV5 promoter. Addition of
isopropyl-
-D-thiogalactopyranoside induces the
expression of p11. The expressed p11 was purified using
histidine-binding resin (Novagen) according to the manufacturer's instruction. The His-tag-p11 was digested with enterokinase, and the
released p11 without His-tag was collected and further purified by gel
filtration using Bio-Gel P-30 (Fig. 1). The purified p11 was used for subsequent studies of the regulatory effects of p11 on
cPLA2 enzyme activity.
In Vitro PLA2 Activity Assay Using Purified cPLA2 and p11
As the in vitro assayed
cPLA2 activity is dependent on the composition of the
liposome (36), we used the natural membrane vesicles prepared from the
BEAS 2B cells as cPLA2 substrate. To obtain the natural
membrane vesicles, the BEAS 2B cells grown in 175-cm2
flasks were labeled with
[5,6,8,9,11,12,14,15-3H]arachidonic acid (214 Ci/mmol) in
LHC-8 media. As incorporation of AA to membrane phospholipid classes
may be dependent on the length of labeling period, the cells were
labeled twice with [3H]AA (18-h and 10-min labeling
periods) to ensure even distribution of radiolabeled AA in various
classes of phospholipids (5-6, 37). The cells were first labeled with
[3H]AA for 18 h to allow [3H]AA to
equilibrate among phospholipids. Under these conditions, phosphatidylethanolamine accounts for the majority of esterified [3H]AA in phospholipids, the remainder being esterified
in phosphatidylcholine, phosphatidylinositol, and phosphatidylserine.
After the 18-h labeling period, the cells were labeled again with
[3H]AA for 10 min. At this short labeling term, the
distribution of AA esterified in phospholipids differs from that seen
at long incubation times in that phosphatidylcholine, not
phosphatidylethanolamine, is the phospholipid that incorporates most of
the radiolabeled AA. The dual-labeled cells were harvested and
resuspended in 0.5 ml of homogenization buffer (50 mM
HEPES, pH 7.4, 1 mM EGTA, 50 µg/ml leupeptin, 1 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, 10 µM phosphoramidon, 10 µg/ml soybean
trypsin inhibitor, 100 µg/ml aprotinin). Cells were disrupted by
repeated freeze/thaw cycles. The lysate was centrifuged at 1000 × g for 5 min to remove nuclei, unbroken cells, and debris.
The low speed supernatant was then centrifuged at 100,000 × g for 1 h to produce a crude cytosolic fraction
(supernatant) and a membrane fraction (pellet). The crude cytosol was
discarded, and the pellet was washed once, resuspended in 50 mM HEPES buffer (pH 7.4, without EGTA and protease inhibitors), and stored in aliquots at 80 °C for future use as cPLA2 substrate. Membrane phospholipid concentration was
estimated after separation of phospholipids by thin layer
chromatography on a silica gel G plate (Alltech, Dearfield, IL)
developed with chloroform/methanol/acetic acid/water, 50:30:8:4, using
standards of phosphatidylcholine, phosphatidylserine,
phosphatidylethanolamine, and phosphatidylinositol (Sigma). The
phospholipids were visualized by development with
2,7-dichlorofluorescein (Aldrich).
In vitro cPLA2 activity was measured using 3H-labeled BEAS 2B cell membrane (7.5 × 103 dpm/6.5 µg protein/assay) and purified human cPLA2 in a final volume of 100 µl (50 mM HEPES buffer, pH 7.4). The cPLA2 was kindly provided by the Genetics Institute and is purified from U937 cells. For accurate control of the Ca2+ concentration, a CaCl2/EGTA buffering system was used (38). As the physiological cellular Ca2+ level is normally at micromolar level (0.1 µM in normal resting cells and 0.3-1.0 µM in activated cells) and the cPLA2 activity is measurable under micromolar concentrations of Ca2+, a physiologically relevant Ca2+ concentration (0.5 µM free Ca2+) was used for all the assays (1 mM EGTA and 0.96 mM CaCl2). The reaction was started by the addition of substrate to the reaction mixture. The assays were then incubated at 37 °C for 1 h and terminated by the addition of 300 µl of 2:1 chloroform/methanol containing 1% acetic acid and 1 mg/ml free AA. Release of free fatty acid was analyzed using silica gel H thin layer chromatography plates as described previously (23-24). The effect of p11 was expressed as a percent of the arachidonate liberated by cPLA2 alone in each experiment (500-580 dpm).
Binding of Purified p11 and cPLA2 in VitroThe
35S-labeled p11 (carrying the His-tag at the
NH2-terminal) was synthesized from the pET30a-p11 vector by
the coupled in vitro transcription and translation using the
Single Tube ProteinTM System 2 (Novagen). For each synthesis reaction,
0.5 µg of pET30a-p11 plasmid template was transcribed in 10 µl at
30 °C for 15 min, followed by the addition of 40 µl of translation
mixture with [35S]methionine and continued incubation for
60 min. Products from individual synthesis reactions were pooled and
stored in aliquots at 20 °C for future use.
To allow formation of the p11-cPLA2 complex, 5 µl of the synthesized [35S]p11 (3.4 × 106 dpm) was incubated with 100 ng of purified human cPLA2 in 1 ml of HBSS (with calcium and magnesium) at 4 °C for 30 min. The cPLA2 was provided by Genetics Institute, Boston, MA, and is purified from U937 cells. 10 µl of rabbit anti-human cPLA2 antibody (Genetics Institute) and 25 µl of Protein G Plus/Protein A-agarose (Pierce) were then added, and the mixture was incubated at 4 °C on a rotating device for 4 h, followed by centrifugation in a microcentrifuge at 2500 rpm for 5 min at 4 °C. The supernatant was aspirated, and the pellet was washed four times with 1.0 ml of PBSTDS (1 × phosphate-buffered saline, 1% Triton X-100, 0.5% sodium deoxycholate, and 0.1% sodium dodecyl sulfate) by repeated centrifugation. Following four washings with PBSTDS, 20 µl of protein loading buffer was added to the pellet, and the sample was boiled for 10 min before electrophoresis on 18% polyacrylamide gels (Novex, San Diego, CA) using Tris glycine/SDS buffer. The gel was then dried under vacuum and exposed to the x-ray film.
Immunoprecipitation of Native p11 Protein from BEAS 2B Cell Lysate Using Anti-human cPLA2 AntibodyFor isolation
of the crude cytosolic proteins, the BEAS 2B cells grown on
175-cm2 culture flasks were digested with 0.1% collagenase
for 10 min. The cells were then lysed in 0.5 ml of homogenization
buffer containing 1 mM EGTA and protease inhibitors, and
the crude cytosolic protein was isolated as described above. The crude
cytosolic samples isolated from individual flasks were pooled and
stored in aliquots at 80 °C for future use. For
immunoprecipitation, the isolated crude cytosolic fraction (100 µl,
200 µg of protein) was added to an Eppendorf tube containing 1 ml of
HBSS (with calcium and magnesium) and 10 µl of rabbit anti-human
cPLA2 antibody. The samples were preincubated at 4 °C
for 30 min. 25 µl of Protein G Plus/Protein A-agarose was then added
to each sample, and the mixture was incubated at 4 °C on rotating
devices for 4 h, followed by centrifugation in a microcentrifuge
at 2500 rpm for 5 min at 4 °C. The supernatant was aspirated, and
the pellet was washed four times with 1.0 ml of PBSTDS by repeating
centrifugation. Following four washings with PBSTDS, 20 µl of protein
loading buffer was added to the pellet, and the sample was boiled for
10 min before electrophoresis on 16% polyacrylamide gels (Novex) using
Tris glycine/SDS buffer. The separated proteins were
electrophoretically transferred onto a nitrocellulose membrane, which
was blocked with 5% nonfat milk overnight, and then probed with a
1:1000 dilution of mouse anti-p11 monoclonal antibody. The blots were
then probed with a 1:1000 dilution of horseradish peroxidase-labeled
goat anti-mouse IgG and detected using the enhanced chemiluminescence
Western Blotting Detection System.
Crude cytosolic extracts were prepared as described above using BEAS 2B cells. 100 µl (200 µg, protein) of the cytosol was immunoprecipitated with 2.5 mg of mouse anti-p11 monoclonal antibody (Transduction Science) and Protein G/Protein A-agarose. Following washing of the Protein G/Protein A-agarose pellet with PBSTDS, 100 µl of loading buffer was added to the pellet, and the sample was boiled for 5 min before electrophoresis of 20 µl on 6% polyacrylamide gels using Tris glycine/SDS buffer. The separated proteins were electrophoretically transferred onto a nitrocellulose membrane that was blocked with 5% nonfat milk overnight and then probed with a 1:1000 dilution of rabbit anti-cPLA2 antibody (Genetics Institute). The blots were then probed with a 1:1000 dilution of horseradish peroxidase-labeled goat anti-rabbit IgG and detected using the Pierce Super SignalTM detection system.
Stable Transfection of Antisense p11 Plasmid in BEAS 2B CellsTo construct the antisense p11 expression vector, full-length human p11 cDNA was released from one of the three isolated plasmids encoding the fusion protein GAL4 activation domain/p11 protein and then cloned into the pCI-neo mammalian expression plasmid in the antisense orientation (pCI-neo-p11AS vector). The pCI-neo vector carries the human cytomegalovirus immediate-early enhancer/promoter region to promote strong, constitutive expression of the cloned p11 inserts in mammalian cells. The vector also contains the SV40 enhancer and early promoter region to control the expression of the neomycin phosphotransferase gene and the SV40 origin of replication to allow efficient replication in cell lines that express SV40 large T antigen (e.g. BEAS 2B cells). The BEAS-2B cells grown in 175-cm2 flasks were exposed to 250 µl of LipofectAMINE Reagent with 20 µg of p11AS-pCI-neo plasmid. Control cells were transfected with pCI-neo expression plasmid alone. Cells were exposed to the mixture of LipofectAMINE and plasmid for 2 h. Following removal of the transfection reagent, fresh medium was added and the incubation was continued for 24 h. The cells were then plated in medium containing 800 µg/ml geneticin (G418 sulfate). Colonies of resistant cells appeared after approximately 12 days, and the cells were subcultured at 16 days. Subsequent cultures of the selected cells were routinely grown in the presence of selective pressure.
To determine the protein levels of p11, cPLA2, and annexin II, crude cytosolic protein was isolated from cells transfected with pCI-neo vector alone or transfected with p11AS-pCI-neo vector. The cells were grown on 175-cm2 culture flasks. Each flask contained approximately 3 × 107 cells when the culture reached confluence. The cells were lysed in the homogenization buffer containing 1 mM EGTA and protease inhibitors, and the crude cytosolic protein was isolated as described above. The obtained cytosolic protein was subjected to SDS-PAGE and Western blotting analysis using a 1:1000 dilution of mouse anti-p11 (Transduction Laboratories), rabbit anti-cPLA2, or mouse anti-annexin II (Transduction Laboratories) antibodies. The blots were then probed with a 1:1000 dilution of horseradish peroxidase-labeled goat anti-mouse IgG (for detection of p11 and annexin II) or Protein A (for detection of cPLA2) and finally detected using the enhanced chemiluminescence Western Blotting Detection System.
Immunoprecipitation of Native p11 Protein and Annexin II Heavy Chain from Transfected BEAS 2B Cells Using Anti-human cPLA2 AntibodyCrude cytosolic extracts were prepared, and immunoprecipitation with anti-cPLA2 antibody was performed as described above using BEAS 2B cells stably transfected with either the antisense p11 plasmid or the pCI neo plasmid (vector). Following washing of the Protein G/Protein A-agarose pellet with PBSTDS, 100 µl of loading buffer was added to the pellet, and the sample was boiled for 5 min before electrophoresis of 20 µl on 18% polyacrylamide gels using Tris glycine/SDS buffer. The separated proteins were electrophoretically transferred onto a nitrocellulose membrane that was blocked with 5% nonfat milk overnight and then probed with a 1:1000 dilution of either mouse anti-p11 monoclonal antibody or mouse anti-annexin II monoclonal antibody. The blots were then probed with a 1:1000 dilution of of horseradish peroxidase-labeled goat anti-mouse IgG and detected using the Pierce Super SignalTM detection system.
PLA2 Activity and Arachidonic Acid Release from Cells Transfected with Antisense p11 PlasmidCellular PLA2 activity was assayed as described previously (23-24). Equal numbers of the control cells transfected with pCI-neo vector alone or cells transfected with the plasmid pCI-neo-p11AS were grown on 175-cm2 culture flasks coated with type I rat tail collagen. Each flask contained approximately 3 × 107 cells when culture reached confluence. The number of cells in each culture dish was monitored by measuring cellular DNA level using bis-benzimidazole (Hoechst 33258) (39).
For [3H]AA release, equal numbers of the control cells
transfected with pCI-neo vector alone and the cells transfected with the antisense p11 plasmid pCI-neo-p11AS were grown in
175-cm2 culture flasks and, when the cells were nearly
confluent, labeled for 18 h with 1 µCi/ml
[5,6,8,9,11,12,14,15-3H]arachidonic acid (214 Ci/mmol) in
LHC-8 media. For studies of basal release the cells were then washed
with culture media and fresh media replaced and collected after 4 hours
of culture. The samples were extracted by Sep-Pak C18
cartridges and chromatographed by reverse phase high pressure liquid
chromatography (HPLC) as described previously (23-24). The arachidonic
acid fraction of HPLC elution was collected and measured for
radioactivity. For studies of [3H]AA release after
calcium ionophore stimulation, equal numbers of the control cells
transfected with pCI-neo vector alone and the cells transfected with
the antisense p11 plasmid pCI-neo-p11AS were grown on 60-mm culture
dishes coated with type I rat tail collagen. Each dish contained
approximately 4 × 106 cells when culture reached
confluence. Experiments were performed when the cells were nearly
confluent. The cells were labeled for 18 h with 1 µCi/ml
[5,6,8,9,11,12,14,15-3H]arachidonic acid (214 Ci/mmol) in
LHC-8 media. Following repeated washing with HBSS, 5 ml of HBSS with
calcium ionophore A23187 (106 M) was added to
each dish, and the cells were incubated at 37 °C for 15 min. The
supernatants collected from four dishes were pooled as one sample for
further analysis. The samples were extracted by Sep-Pak C18
cartridges and chromatographed by reverse phase high pressure liquid
chromatography (HPLC) as described above. The arachidonic acid fraction
of HPLC elution was collected and measured for radioactivity.
In an effort to identify proteins that interact with either the NH2-terminal regulatory Ca2+-binding domain or the COOH-terminal catalytic domain of the 85-kDa cPLA2 protein, different regions of the cPLA2 cDNA were cloned into pGBT9 vector to express the fusion proteins containing the common GAL4 DNA-binding domain and the NH2-terminal or COOH-terminal fragment of the cPLA2 protein. Either the plasmid pGBT9-cPLA2N (containing the 540-bp cPLA2 cDNA fragment corresponding to the NH2-terminal cPLA2 protein) or pGBT9-cPLA2C (containing the 1707-bp cPLA2 cDNA fragment corresponding to the COOH-terminal cPLA2 protein) was used to transform the host HF7c cells. The HF7c strain containing pGBT9-cPLA2N or pGBT9-cPLA2C was used for subsequent transformation with DNA prepared from a human leukocyte MATCHMAKER cDNA library constructed in the pGAD10 plasmid. By this approach, four clones were identified to interact with the COOH-terminal region of cPLA2 protein. However, no clone was found to interact with the NH2-terminal region of the cPLA2.
To verify the true positives, the isolated four plasmids were
retransformed back into the original HF7c yeast host strain in the
presence or absence of plasmid pGBT9 containing GAL4 DNA-binding domain
alone, plasmid pGBT9-cPLA2N, or plasmid pGBT9-cPLA2C. All four positive
clones exhibited positive results in the -galactosidase activity
assay only when cotransformed with the plasmid pGBT9-cPLA2C. These
plasmids were also cotransformed into another host S. cerevisiae strain SFY526. Similar results in the
-galactosidase
activity assay were obtained in SFY526. In SFY526, the lacZ
reporter gene is under the control of a promoter different from that
used to control lacZ in HF7c. While these two promoters
share only GAL4-responsive elements in common, the rest of the two
promoter sequences differ significantly. Any positive two hybrid
interaction observed in both HF7c and SFY526 reporter strains likely
requires binding of the GAL4 DNA-binding domain specifically to the
GAL4 responsive elements (40). Therefore, we concluded that the
isolated four clones represent the true positive.
DNA sequences were obtained for inserts of the four clones identified in the genetic screen. Three of the clones were identical and contained the full-length cDNA sequence of human p11 (34), a member of the S100 protein family.
p11 Inhibits cPLA2 Enzyme ActivityThe findings
from the two hybrid screen indicated that p11 binds to
cPLA2 protein. To study the possible regulatory effect of
p11 on cPLA2 enzyme activity, experiments were designed to examine the direct effect of p11 on cPLA2 enzyme activity
in vitro. The natural membrane substrate was incubated with
100 ng of purified human cPLA2 with various concentrations
of recombinant human p11. Cytochrome c was used as a control
protein for the in vitro PLA2 activity assays as
described previously (41). As shown in Fig. 2, while
cytochrome c exhibited no effect, p11 caused a
dose-dependent inhibition of the cPLA2
activity. These results demonstrate that p11 interacts with the
cPLA2 enzyme and inhibits the cPLA2 activity. To test whether the inhibitory effect of p11 is related to substrate concentrations, the inhibition of cPLA2 by p11 was
evaluated under various amounts of substrate (2.5-20 µM
phospholipid). As shown in Fig. 3A, p11
exhibited similar inhibitory effects under different substrate
concentrations. These findings suggest that the inhibition of
cPLA2 by p11 is unlikely through an interaction between p11 and the phospholipid substrate.
As p11 is often associated with the 36-kDa annexin II to form annexin II heterotetramer in cells, we also examined whether the inhibitory effect of p11 on cPLA2 activity may be altered by the presence of annexin II. The effect of annexin II alone or combined annexin II and p11 on cPLA2 activity was studied under the same in vitro conditions. As shown in Fig. 3B, annexin II alone exhibited some inhibitory effect on cPLA2 activity. This effect was not unexpected because annexin II might sequester substrate from its access to various PLA2s under the in vitro assay conditions. While annexin II alone exhibited certain inhibitory effect on cPLA2 activity, this effect was less remarkable than the inhibition by p11. The combination of annexin II and p11 appeared to exhibit additive inhibition. These results indicate that annexin II does not significantly alter the inhibitory effect of p11 on cPLA2 activity. p11 itself is sufficient to bind and inhibit cPLA2, and the formation of p11-annexin II complex is not required for the inhibition.
p11 Binds to cPLA2 in VitroTo further
demonstrate the interaction between p11 and cPLA2, we used
the in vitro translated human p11 and purified human cPLA2 to examine their binding ability in vitro.
The bacterial expression plasmid (pET30a-p11) containing the His-tag
and full-length human p11 cDNA sequences was used as the template
to synthesize 35S-labeled p11 proteins by the coupled
in vitro transcription and translation procedures. The
synthesized protein included the His-tag-p11 fusion protein as well as
other smaller size proteins that were likely to be the translational
products from internal translation start sites (Fig. 4).
The in vitro translated 35S-labeled p11 was
incubated with purified cPLA2 protein to allow the
formation of p11-cPLA2 complex. As shown in Fig. 4, p11
protein was not precipitated by anti-cPLA2 in the absence
of cPLA2 protein, but it was precipitated if preincubated
with cPLA2. This result demonstrates the formation of
p11-cPLA2 protein complex in vitro.
Identification of p11-cPLA2 Complex from a Human Bronchial Epithelial Cell Line (BEAS 2B Cell)
The above results
demonstrated that p11 binds to cPLA2 in vivo
(based on two hybrid screen) and in vitro (based on the
in vitro binding of synthesized p11 and purified
cPLA2). To further investigate the interaction between p11
and cPLA2 in human cells, additional experiments were
designed to detect the existence of p11-cPLA2 complex in
cells expressing both p11 and cPLA2. Our previous studies
revealed the existence of cPLA2 in the BEAS 2B human
bronchial epithelial cell line (23-24). Using immunoblotting analysis,
the expression of p11 was also found in the BEAS 2B cells (see below,
Fig. 6). Therefore, we reasoned that the p11-cPLA2 complex
might be detected from the BEAS 2B cells under appropriate conditions.
As shown in Fig. 5, p11 was precipitated from the BEAS
2B cell lysate by anti-human cPLA2 antibody. This result demonstrates the existence of p11-cPLA2 complex in the BEAS
2B cells. Attempts to precipitate the complex with a monoclonal
antibody to p11 were not successful (data not shown).
Antisense Inhibition of p11 Increases PLA2 Activity and Arachidonic Acid Release from the BEAS 2B Cells
We have shown
that p11 binds to cPLA2 and inhibits the cPLA2
activity assayed in vitro. To study further the effect of
p11 on PLA2 activity and AA release in human cells, we used
the antisense RNA approach to selectively inhibit the expression of p11
and then to examine the PLA2 activity and AA release in the
BEAS 2B cells. The expression of p11 protein was successfully reduced by the antisense inhibition (Fig. 6). The antisense
inhibition of p11 expression did not alter cellular production of
annexin II. In addition, the interaction of p11 and of annexin II heavy chain with cPLA2 from vector control cells and from cells
transfected with the p11 antisense vector was studied.
Immunoprecipitation of cytosol preparations from vector control cells
and from cells transfected with the antisense vector was performed
using the polyclonal anti-cPLA2 antibody. Western blotting
was then performed, and the blots were probed with antibody to p11 or
to annexin II heavy chain. While the antisense cells contained reduced
p11 but comparable amounts of annexin II heavy chain,
immunoprecipitation of cytosol with anti-cPLA2 antibody
demonstrated less p11 and less annexin II bound to cPLA2 in
the antisense cells than in vector control cells (Fig.
7). We then determined the PLA2 activity and AA release
from these permanently transfected cells. The in vitro
assayed PLA2 activity in cells transfected with
pCI-neo-p11AS vector (91.7 ± 5.8 pmol/mg protein) was
significantly higher than in cells transfected with pCI-neo vector
alone (46.8 ± 3.4 pmol/mg protein) (p < 0.01, n = 3). The release of prelabeled AA from cells
transfected with pCI-neo-p11AS was studied under two conditions. First,
unstimulated or basal release of AA from both cell lines was studied
over a period of 4 h after labeling with [3H]AA for
18 h. Without stimulation, cells permanently transfected with
vector alone released 68 ± 8% of the arachidonate released from cells
transfected with the p11-AS vector (n = 6, p < 0.01). Second, because cPLA2 may be activated by
increases in intracellular calcium concentrations, AA release after
treatment with the calcium ionophore A23187 was studied. After
treatment with A23187, the release of prelabeled AA from cells
transfected with pCL-neo-p11AS was significantly higher than from cells
transfected with pCI-neo vector alone (Fig. 8). These
results indicate that p11 inhibits PLA2 activity and thus
decreases AA release in normal BEAS 2B cells. Decreased expression of
p11 in the antisense cells reduces the inhibitory effect of
cPLA2 by p11, which results in an increase in
PLA2 activity and AA release.
This study reveals a novel mechanism for the regulation of 85-kDa cPLA2. We demonstrate that p11, a unique member of the S100 family of calcium-binding proteins, interacts with the carboxyl region of cPLA2 and inhibits the cPLA2 activity. The interaction between p11 and cPLA2 was recognized by screening a human cDNA library using a yeast two hybrid system. Further experiments demonstrated that p11 associated with cPLA2 in vitro. Using natural membrane vesicles as substrate, p11 was found to inhibit the activity of purified cPLA2 in vitro. The inhibition of cPLA2 activity by p11 was independent of substrate concentrations. These results demonstrated that p11 directly interacts with cPLA2 and inhibits the cPLA2 enzyme activity. Additional experiments were performed to study the role of p11 in intact human cells. The p11-cPLA2 complex was identified from the BEAS 2B human bronchial epithelial cell line expressing both cPLA2 and p11. Selective inhibition of p11 expression in the BEAS 2B cells by antisense RNA caused a reduction of cellular p11 and resulted in an increased PLA2 activity as well as an increased release of prelabeled AA. Taken together, the above in vitro and in vivo observations demonstrate a novel mechanism for the regulation of cPLA2 by the S100 protein family.
p11, a Unique Member of the S100 Family of Calcium-binding Proteins, Interacts with the COOH-terminal Region of cPLA2The S100 proteins are a group of low molecular
mass (approximately 10-12 kDa) acidic Ca2+-binding
proteins, so named after the solubility of the first isolated protein
in 100% saturated ammonium sulfate (42-44). The most striking
conserved feature of these proteins is the presence of EF-hands (a term
representing the two -helices, E and F, and the intervening
Ca2+ binding loop), which function as the
Ca2+-binding sites. While other Ca2+-binding
proteins such as the parvalbumin protein family contain three EF-hands
and the calmodulin protein family and troponin C protein family contain
four EF-hands per monomer molecule, the S100 proteins have two EF-hands
(NH2-terminal and COOH-terminal EF-hands). The evident
amino acid sequence homology, conservation of calcium-binding sites,
and similarities of gene organization suggest that this group of
proteins evolved from a common ancestral EF-hand gene (42-44). The
biological functions of the S100 proteins are less clear. No specific
enzymatic property has been ascribed to any of the proteins to date.
The binding to calcium induces a conformational change in the S100
proteins, and this may then affect the secondary effector proteins.
This mode of protein-protein interaction and modulation of the activity
of the secondary effector protein is similar to that seen with
calmodulin, another family of calcium-binding proteins containing the
EF-hands (42-44). As the distribution of particular S100 proteins is
dependent on specific cell types, the S100 proteins may be involved in
transducing the signal of an increase in intracellular calcium in a
cell type-specific fashion (42-44).
p11 was first isolated as a potential substrate for the epidermal growth factor-receptor tyrosine kinase (45). Although p11 shares considerable sequence homology with the S100 family, it does not bind calcium because of amino acid deletions and substitutions in the two EF-hand motifs (46). This is an exception to the general Ca2+-binding properties displayed by other S100 proteins. Instead, p11 interacts with another protein, the 36-kDa annexin II (47-48). The p11 and annexin II proteins are linked to each other in a tight heterotetrameric complex that is present in many eukaryotic tissues and cells. p11 itself forms a tight dimer that bridges two annexin II molecules thus mediating the formation of a heterotetrameric annexin II2p112 complex. While annexin II exists as a monomer or binds to p11 forming the heterotetrameric complex, p11 has often been found as part of a complex with annexin II in many p11-positive tissues where annexin II is present in equimolar amounts or in excess (47-48). However, this phenomenon may vary with different cell types. The expression of p11 and annexin II is not always coordinated, and the ratio of p11 to annexin II varies with different cell types (49). In the F9 teratocarcinoma cells, there exists a significant amount of p11 but almost no annexin II both at the RNA and protein levels, with the unbound p11 distributed in the cytoplasm (50). Differentiation of these cells results in a large increase in annexin II and the formation of the annexin II2p112 heterotetramer (50). The formed heterotetrameric complex becomes associated with plasma membrane and submembranous cytoskeleton through the Ca2+-dependent interaction of annexin II with the negatively charged phospholipids and components of the cytoskeleton. The interaction of annexin II with actin and membrane phospholipids is dependent upon Ca2+ and p11. p11 binding alters the Ca2+ and/or phospholipid-binding properties of annexin II, with the tetrameric complex having a more than 100-fold increased affinity toward Ca2+/phospholipid compared with monomeric annexin II (51).
It remains unknown whether p11 is also able to associate with any other protein. Our results demonstrate that the 85-kDa cPLA2, a unique calcium-binding enzyme, is a novel protein that interacts with p11. Results from the two hybrid screen demonstrated that p11 associated with cPLA2 in vivo. While the exact amino acid residues involved in the interaction is unknown, only the COOH-terminal region of the cPLA2 is found to interact with the p11. The association of p11 and cPLA2 was also found under in vitro conditions. By using purified human cPLA2 protein and the in vitro translated human p11 protein, we found that p11 and cPLA2 interact with each other in vitro. Furthermore, the formation of p11-cPLA2 complex was also identified from cells expressing both p11 and cPLA2 by immunoprecipitation. Taken together, these results establish a direct interaction between p11 and the COOH region of cPLA2. In addition to annexin II, cPLA2 represents the only other protein that binds to p11.
p11 Inhibits cPLA2 Activity and Decreases the Release of Prelabeled AA from BEAS 2B CellsTo examine how the binding of p11 to cPLA2 may alter the catalytic activity of cPLA2, we studied the effects of recombinant p11 on the activity of purified human cPLA2 under in vitro assay conditions. Two aspects of the assays deserve comment. The first aspect is the use of membrane vesicles containing a mixture of phospholipids as cPLA2 substrate to mimic the in vivo interaction between cPLA2, p11, and membrane. This method was developed by Clark and colleagues (8) who used the U937 cell membrane vesicles as cPLA2 substrate. In this study, the membrane vesicles labeled with [3H]AA were prepared from the BEAS 2B cells. Another aspect of the assay is the relatively low Ca2+ concentration used in the assay. As the cPLA2 is activated by micromolar concentrations of Ca2+ and the physiological cellular Ca2+ level is normally at micromolar level as well (0.1 µM in normal resting cells and 0.3-1.0 µM in activated cells), we used the physiologically relevant Ca2+ concentration (0.5 µM free Ca2+ in the Ca2+-EGTA buffer) for all the assays. Under these assay conditions, we found that p11 inhibited cPLA2 activity in a dose-dependent fashion (Fig. 2), and this inhibitory effect is not dependent on substrate concentration (Fig. 3A). These results demonstrate that p11 binds to cPLA2 and inhibits the cPLA2 enzyme activity in vitro. Although the mechanism for the inhibition of cPLA2 by p11 is not yet known, it is unlikely that this effect is due to the interaction between p11 and phospholipid substrate because 1) p11 does not bind to phospholipids (46-47) and 2) the p11-mediated cPLA2 inhibition is not dependent on substrate concentration. Instead, we demonstrate a direct interaction between p11 and cPLA2, and it is likely that the binding of p11 to cPLA2 results in the inhibition of cPLA2 activity. However, it remains to be determined how the binding of p11 to cPLA2 inhibits the cPLA2 activity. The possibilities may include blockage of the catalytic process of cPLA2, conformational change of cPLA2, or alteration of calcium-dependent association of cPLA2 with membrane substrate.
To study further the effect of p11 on PLA2 activity and AA release in human cells, we used the antisense RNA approach to selectively inhibit the expression of p11 and then to examine the PLA2 activity and AA release in the BEAS 2B cells. The release of prelabeled AA from cells with reduced p11 expression was significantly higher than from the control cells (Fig. 7). The crude cytosolic fraction isolated from the p11-reduced cells also showed an increased PLA2 activity. These results demonstrate that p11 inhibits PLA2 activity and decreases AA release in the BEAS 2B cells. Reduced p11 expression reverses the inhibitory effect of p11 on cPLA2 and thus increases the PLA2 activity and increases the release of [3H]AA from the [3H]AA-prelabeled cells. These results suggest that p11 decreases the release of AA for eicosanoid production in intact cells by inhibiting cellular cytosolic PLA2 activity.
p11-induced cPLA2 Inhibition Does Not Depend on Annexin IIIt is well known that p11 binds to annexin II, a member of the
annexin protein family, which includes at least 13 relatively abundant
and structurally related proteins (47-48, 52-57). The annexins are
widely distributed in eukaryotic cells. The roles of the annexins are
not well understood, but they have been suggested to function in a
broad range of physiological processes including inhibition of
PLA2 (54). The biochemical hallmark of members of the
annexin protein family is their Ca2+-dependent
binding to negatively charged phospholipids, most likely reflecting
their association with cellular membranes in vivo. This
common property is mediated by the conserved
Ca2+/phospholipid-binding sites that comprise four to eight
repeats of 70-80 amino acid segments with a highly conserved 17-amino acid consensus sequence in each repeat. Typically, the annexins are
monomeric proteins with the exception of annexin II which can exist as
monomer or heterotetramer. The annexin II heterotetramer is composed of
two subunits of 36-kDa annexin II and two subunits of p11 (47-48). The
annexin II consists of two functional domains. The amino-terminal
domain contains the first 30 amino acids and includes the binding site
for p11 and both the serine and tyrosine phosphorylation sites. The
first NH2-terminal 14 amino acids of annexin II molecule
bind to p11 forming an amphipathic -helix upon complex formation.
The serine and tyrosine residues can be phosphorylated by protein
kinase C and the tyrosine kinase encoded by the src
oncogene, respectively (45). The remaining carboxyl domain of annexin
II contains four repeats of the conserved amino acid segments that bind
to Ca2+, phospholipids, and F-actin. Although p11 does not
directly bind to calcium and phospholipids, annexin II has all
Ca2+- and lipid-binding properties typical of members of
the annexin family.
For the treatment of inflammatory diseases, considerable research efforts have been made to identify selective inhibitors of PLA2. The annexins, formerly called lipocortins or calpactins, were found to inhibit cellular PLA2 activity in various cell types (54, 58). In vitro studies using the 14-kDa PLA2 demonstrated that the PLA2 activity was inhibited by annexins, and this led to the hypothesis that the inhibition of 14-kDa PLA2 by annexins was the mechanism of the anti-inflammatory action for glucocorticoids (54, 58-61). However, subsequent studies failed to demonstrate any direct interaction between the 14-kDa PLA2 and annexins. Instead, it was found that the in vitro inhibition of 14-kDa PLA2 by annexins could be overcome by increasing the substrate concentrations (41, 62-64). The extent of inhibition was more closely related to the inhibitor:substrate rather than the inhibitor:enzyme ratio, and the kinetics of inhibition were not compatible with a simple competitive or noncompetitive pattern. Based on these observations, the inhibitory action of annexins toward 14-kDa PLA2 was ascribed to substrate sequestration rather than to direct interaction with the 14-kDa cPLA2.
We have demonstrated that p11 interacts with the COOH-terminal region of cPLA2 and inhibits the cPLA2 enzyme activity. As the inhibition of cPLA2 by p11 is independent of substrate concentration, this effect is likely due to a direct interaction between p11 and cPLA2. Because p11 is often associated with the 36-kDa annexin II to form annexin II heterotetramer in cells, we also examined whether the inhibitory effect of p11 on cPLA2 activity may be altered by the presence of annexin II. The effect of annexin II alone or combined annexin II and p11 on cPLA2 activity was studied under the in vitro conditions (Fig. 3B). We found that annexin II alone exhibited some inhibitory effect on cPLA2 activity. This effect was as expected because annexin II might sequester substrate from its access to various PLA2s under the in vitro assay conditions. While annexin II alone exhibited certain inhibitory effect on cPLA2 activity, this effect was less remarkable than the inhibition by p11. The combination of annexin II and p11 appeared to exhibit additive inhibition. These results indicated that annexin II does not significantly alter the inhibitory effect of p11 on cPLA2 activity. p11 itself is sufficient to bind and inhibit cPLA2.
We thank Drs. J. D. Clark and J. L. Knopf at the Genetics Institute, Boston, MA, for providing the purified human cPLA2, anti-cPLA2 antibody, and human cPLA2 plasmid and Drs. J. E. Lechner and C. C. Harris at NCI, National Institutes of Health, Bethesda, MD, for providing the BEAS 2B cells.