(Received for publication, February 18, 1997, and in revised form, March 31, 1997)
From the Institut de Pharmacologie Moléculaire
et Cellulaire, CNRS UPR 411, Sophia Antipolis, 660 route des Lucioles,
06560 Valbonne, France,
U406 INSERM Génétique
Médicale, et Développement, Faculté de
Médecine, 27 Boulevard Jean Moulin, 13385 Marseille cedex 05, France, and ¶ URA 1283 CNRS, Centre Hospitalier Universitaire
Saint Antoine, 27 rue Chaligny, 75012 Paris, France
Secretory phospholipases A2 (sPLA2s) represent a rapidly expanding family of structurally related enzymes found in mammals as well as in insect and snake venoms. In this report, a cDNA coding for a novel sPLA2 has been isolated from human fetal lung, and its gene has been mapped to chromosome 16p13.1-p12. The mature sPLA2 protein has a molecular mass of 13.6 kDa, is acidic (pI 5.3), and made up of 123 amino acids. Key structural features of the sPLA2 include: (i) a long prepropeptide ending with an arginine doublet, (ii) 16 cysteines located at positions that are characteristic of both group I and group II sPLA2s, (iii) a C-terminal extension typical of group II sPLA2s, (iv) and the absence of elapid and pancreatic loops that are characteristic of group I sPLA2s. Based on these structural properties, this sPLA2 appears as a first member of a new group of sPLA2s, called group X. A 1.5-kilobase transcript coding for the human group X (hGX) sPLA2 was found in spleen, thymus, and peripheral blood leukocytes, while a less abundant 0.8-kilobase transcript was detected in the pancreas, lung, and colon. When the hGX sPLA2 cDNA was expressed in COS cells, sPLA2 activity preferentially accumulated in the culture medium, indicating that hGX sPLA2 is an actively secreted enzyme. It is maximally active at physiological pH and with 10 mM Ca2+. hGX sPLA2 prefers phosphatidylethanolamine and phosphatidylcholine liposomes to those of phosphatidylserine.
Phospholipases A2 (PLA2s; phosphatidylcholine 2-acylhydrolase, EC 3.1.1.4)1 represent a growing family of enzymes that catalyze the hydrolysis of glycerophospholipids at the sn-2 position, producing free fatty acids and lysophospholipids (1-3). PLA2s generate rate-limiting precursors in the biosynthesis of various types of biologically active lipids, including prostaglandins, hydroxy fatty acids, leukotrienes, thromboxanes, and platelet-activating factor. Over the last two decades, two major classes of mammalian PLA2s (intracellular and secretory) have been characterized and cloned. Secretory PLA2s were first characterized and then classified into different groups according to their molecular structure and the localization of their disulfide bridges (4, 5). Recently, an updated classification of PLA2s has been proposed that include both intracellular and secretory types of PLA2s (2). A number of intracellular PLA2s have been characterized and now comprise the well known Ca2+-sensitive arachidonoyl-specific 85-kDa cPLA2 (6, 7), multiple Ca2+-independent PLA2s (1, 2, 8) as well as several other types of cytosolic enzymes less extensively characterized (9-13).
Secretory low molecular mass PLA2s (sPLA2s, 13-18 kDa) form a class of structurally related enzymes whose members are also rapidly increasing (2). These enzymes are characterized by the presence of several disulfide bridges, an absolute catalytic requirement for millimolar concentration of Ca2+, and a broad specificity for phospholipids with different polar head groups and fatty acyl chains (14). sPLA2s have been purified from a variety of sources including not only mammalian pancreas, spleen, lung, platelets, and extracellular fluids, but also insect and snake venoms (1, 15-17). Pancreatic sPLA2, nonpancreatic inflammatory sPLA2, bee venom sPLA2, and a novel sPLA2 highly expressed in heart, are prototypes of group I, II, III, and V, respectively (2). The pancreatic group I sPLA2, as its name indicates, was originally purified from pancreatic juice and then identified and cloned in other tissues, such as lung, spleen, kidney, and ovary (3, 18-20). Besides its primary function in digestion of dietary lipids, the pancreatic group I sPLA2 has been proposed to play a role in cell proliferation (21), smooth muscle contraction (22, 23), as well as acute lung injury (24). The inflammatory group II sPLA2 (initially called nonpancreatic sPLA2) has been purified and cloned from various sources, including platelets and extracellular fluids (16, 25, 26). It is highly expressed in the plasma and synovial fluids of patients with various inflammatory diseases, such as rheumatoid arthritis, acute pancreatitis, Crohn's disease, and in endotoxic shock (27-31), as well as in various gastrointestinal cancers (32, 33). It is considered as a potent mediator of the inflammatory process (16, 27, 29, 30, 34) and has been recently proposed as a tumor suppressor gene of intestinal tumorigenesis (35). The human group V sPLA2 has been cloned from brain and is found strongly expressed in heart (36, 37). More recently, this sPLA2 has been detected in P388D1 murine macrophages where it may act as a novel sPLA2 effector involved in lipid mediators production (38). A fourth sPLA2 has been cloned in rat and mouse (39), but appears to be a nonfunctional pseudogene in humans (40). This sPLA2 belongs to group IIC and is prevalently expressed in testis. Other low molecular mass PLA2s have been characterized from various tissues including spermatozoa, brain, and lung, suggesting a larger diversity of PLA2s (41-44). Moreover, a sPLA2-related gene has been cloned from human teratocarcinoma cells and found to code for a protein of 689 amino acids containing two domains of high homology with sPLA2s (45).
A wealth of sPLA2s have also been described in snake and insect venoms. Besides a role in prey digestion, several venom sPLA2s are potent toxins that display neurotoxicity and myotoxicity among a variety of other toxic effects (17, 46, 47). Studies with several iodinated venom sPLA2s, including OS1 and OS2 purified from Taipan snake venom, have shown the existence of various high affinity sPLA2 receptors (48). N-type sPLA2 receptors were first identified in rat brain membranes and are made up of several protein subunits of 36-51 kDa and 85 kDa (49). These receptors display a high affinity for neurotoxic sPLA2s, but not for nontoxic sPLA2s, suggesting that N-type receptors may contribute to sPLA2 neurotoxic effects (49-51). M-type sPLA2 receptors were first identified in skeletal muscle cells (52). They consist of a single 180-kDa subunit and bind a number of toxic and nontoxic venom sPLA2s. The M-type sPLA2 receptor has now been cloned in various species (53-56), and its molecular properties have been analyzed in detail (57-61). Although the physiological roles of M- and N-type receptors remain to be discovered, M-type receptors from various species have been shown to associate with high affinity pancreatic group I sPLA2s as well as inflammatory group II sPLA2s with Kd values of 1-10 nM, suggesting that mammalian endogenous sPLA2s might be the natural ligands of these receptors (21, 48, 54, 61).
We now report the cloning, chromosomal mapping, and recombinant expression of a novel human sPLA2. Based on its structural properties, this sPLA2 appears as a first member of a new group of mammalian sPLA2s, called group X, according to the PLA2 nomenclature defined recently (2).
Protein sequences of different sPLA2s were
used to search for homologues in gene data bases stored at the National
Center for Biotechnology (NCBI) by using the tBLASTn sequence alignment program (62). Translation of an EST sequence (I.M.A.G.E. Consortium clone identification 309343, 3, GenBankTM accession no.
[GenBank]) (63) in one frame presented a significant sequence similarity
(p = 9.9 e
21) with several
sPLA2s. This 445-bp sequence was originally obtained from a
human fetal lung cDNA library. A 327-bp DNA fragment corresponding to the EST sequence was then amplified by reverse transcriptase-PCR using human fetal lung cDNA. For that purpose, poly(A)+
mRNA (CLONTECH Laboratories, Inc.) was
reverse-transcribed with the Moloney murine leukemia virus reverse
transcriptase (Life Technologies, Inc.) according to the
manufacturer's protocol and used for PCR with a sense primer
(5
-GGAATTCGCCTATATGAAATATGGT-3
) and an antisense primer
(5
-GGAATTCAAGGTAGTCAGTCACACTTG-3
) containing EcoRI
restriction sites. PCR reactions were performed using a low error rate
DNA polymerase (PWO polymerase, Eurogentec Corp., Belgium) in a total
volume of 50 µl containing 10 mM Tris-HCl, pH 8.85, 25 mM KCl, 2.0 mM MgCl2, 200 µM dNTPs, 200 µM of each primer, 20 ng of
cDNA template, and 1 unit of PWO DNA polymerase. PCR conditions
were: 94 °C, 30 s; 50 °C, 30 s, and 72 °C, 30 s, for 33 cycles. The 327-bp DNA fragment was then subcloned into pBlueScript (Stratagene Cloning Systems) and sequenced to verify its
identity with the EST sequence by using an automatic sequencer (Applied
Biosystems model 373A) with the dideoxy nucleotide method.
In situ hybridization was performed on chromosome
preparations obtained from phytohemagglutinin-stimulated human
lymphocytes cultured for 72 h. 5-Bromodeoxyuridine (60 µg/ml)
was added for the final 7 h of culture to ensure a
posthybridization chromosomal banding of good quality. The 5 Alu
repeat element was removed from the full-length hGX sPLA2
cDNA clone by restriction with ApoI, and the resulting
DNA fragment (nucleotides
370 to 580, see Fig. 1) inserted into
pBlueScript was tritium-labeled by nick-translation to a specific
activity of 108 dpm/µg. The radiolabeled probe was
hybridized to metaphase spreads at a final concentration of 200 ng/ml
of hybridization solution as described previously (64). After coating
with nuclear track emulsion (Kodak NTB2), slides were
exposed for 20 days at 4 °C and then developed. To avoid any
slipping of silver grains during the banding procedure, chromosome
spreads were first stained with buffered Giemsa solution and
metaphase-photographed. R-banding was then performed by the
fluorochrome-photolysis-Giemsa method, and metaphases were
rephotographed before analysis.
Northern Blot Analysis
Three human multiple tissue Northern
blots (CLONTECH Laboratories, Inc., catalog nos.
7759-1, 7760-1, and 7770-1) were first probed with the randomly primed
32P-labeled PCR fragment corresponding to a partial
sequence of hGX sPLA2 cDNA (nucleotide 179-507 in Fig.
1) in 50% formamide, 5 × SSPE (0.9 M NaCl, 50 mM sodium phosphate, pH 7.4, 5 mM EDTA), 5 × Denhardt's solution, 0.1% SDS, 20 mM sodium phosphate,
pH 6.5, and 250 µg/ml denatured salmon sperm DNA at 42 °C for
18 h. Blots were washed to a final stringency of 0.1 × SSC
(30 mM NaCl, 3 mM trisodium citrate, pH 7.0)
with 0.1% SDS at 55 °C and then exposed to Kodak X-Omat AR films
with two intensifying screens. Northern blots were then stripped and
hybridized in the same conditions as above with the entire coding
sequences of the human group V sPLA2 (36), the human group
II sPLA2 (25), the human group I sPLA2 (19),
and finally with the manufacturer-supplied -actin probe.
The full-length 1020-bp cDNA clone coding for hGX
sPLA2 was subcloned directly into the expression vector
pcDNA I (Invitrogen Corp.) by using convenient restriction enzyme
sites. hGX sPLA2 clones Met-42 and Met-32 were prepared by
a PCR-assisted strategy using a low error rate PWO DNA polymerase
(Eurogentec Corp., Belgium) and respective sense oligonucleotides
primers (Met-42, 5-GGAATTCGCGGCCGCCATGGGGCCGCTACCTGTGT-3
; Met-32, 5
-GGAATTCGCGGCCGCCATGCTGCTCCTGCTACTGC-3
) in combination with a common antisense oligonucleotide
(5
-GTCTAGAGTCAGTCACACTTGGGCGAGT-3
). Primers contained restriction
sites to facilitate subcloning of the PCR fragments and consensus Kozak
sequences were added to the sense primers to enhance protein expression
(65). The PCR products were subcloned into pcDNA I and sequenced to
verify the integrity of DNA constructions. The various plasmids were
then transfected into COS or HEK 293 cells by a modification of the DEAE-dextran/chloroquine method (66) or by the Ca/PO4
procedure (67), respectively. Cells at 50% confluence were transfected with 5 µg of plasmid DNA/75-cm2 Petri dish. Cell
supernatants were harvested at different times post transfection and
assayed for hGX sPLA2 activity. On day 3, washed cells were
harvested and disrupted by sonication in 300 µl of 20 mM
Tris-HCl, pH 7.4, 2 mM EDTA, and 1 mM
phenylmethylsulfonyl fluoride to measure cell-associated
sPLA2 activity. To test whether hGX sPLA2 is
secreted as a proenzyme, serum-free cell supernatants containing hGX
sPLA2 activity were treated with 3 µg/ml trypsin (Sigma,
T-1005 type XI from bovine brain) in 20 mM Tris-HCl, pH 8.0, 10 mM CaCl2 for 3 h at 37 °C and
then assayed for sPLA2 activity with
[3H]oleate-labeled Escherichia coli
membranes.
Preparation of autoclaved E. coli membranes and sPLA2 assays were performed essentially as described previously (57). Unless otherwise specified, sPLA2 assays were performed at 25 °C in a total volume of 100 µl consisting of 140 mM NaCl, 50 mM Tris-HCl, pH 7.4, 10 mM CaCl2, 0.1% bovine serum albumin, and 100,000 dpm of [3H]oleate-labeled E. coli membranes. Incubation times and sample volumes were adjusted to ensure hydrolysis rates within the linear range of enzymatic assays. Typically, 3-10 µl of solution containing hGX sPLA2 were incubated for 30-60 min at 20 °C to measure sPLA2 activity. Reaction mixtures were stopped by adding 300 µl of 0.1 M EDTA, pH 8.0, and 1% fatty acid-free bovine serum albumin. After centrifugation at 10,000 × g for 3 min, 300 µl of supernatant containing hydrolyzed phospholipids were counted. Control incubations in the absence of added sPLA2 were carried out in parallel and used to calculate specific hydrolysis.
sPLA2 Activity Assay on Mixed Phospholipid LiposomesLiposomes were prepared by sonication using different
phospholipid mixtures (phosphatidylethanolamine, phosphatidylcholine, or phosphatidylserine, with
D--dipalmitoylphosphatidylcholine, 75:25 mol%).
D-
-Dipalmitoylphosphatidylcholine is a nonhydrolyzable phospholipid that is used for formation of liposomes with the various
hydrolyzable phospholipids. L-
-Phosphatidylcholine (from egg yolk), L-
-phosphatidylethanolamine (from egg yolk),
L-
-phosphatidylserine (from bovine brain), and
D-
-dipalmitoylphosphatidylcholine were purchased from
Sigma. Phospholipids were dissolved in chloroform, dried under a
nitrogen stream, and resuspended at 0.33 mM in 100 mM Tris-HCl buffer, pH 7.4. The lipid suspension was then
sonicated twice for 2 min with a 20-kHz MSE tip probe at 100 W. Incubations were carried out for 15 min at 37 °C in a total volume
of 500 µl containing 200 nmol of hydrolyzable phospholipids
resuspended in 100 mM Tris-HCl, pH 7.4, 10 mM
CaCl2, and 0.1% fatty acid-free BSA. hGX sPLA2
activity was measured by adding 20 µl of a 72-h COS cell supernatant,
and incubation was carried out for 15 min at 37 °C. Released fatty
acids were extracted by a modification of Dole's procedure (68),
methylated with diazomethane, and quantified as described below by
GC-MS measurements. Control incubations in the absence of added
sPLA2 were carried out in parallel and used to calculate
specific hydrolysis.
Human erythrocytes were isolated from fresh citrated blood by the procedure of Steck et al. (69). Briefly, blood was centrifuged for 10 min at 4 °C at 100 × g. Red blood cells were then collected, washed several times with 5 volumes of an ice-cold phosphate saline buffer (150 mM NaCl, 5 mM sodium phosphate, pH 8.0), and hemolyzed in 40 volumes of 5 mM sodium phosphate, pH 8.0. Erythrocyte ghosts were collected by centrifugation for 20 min at 22,000 × g and washed several times with 5 mM phosphate buffer until "white ghosts" were obtained. sPLA2 assays were performed in a total volume of 500 µl containing erythrocyte membranes (representing 200 nmol of hydrolyzable phospholipids) resuspended in 100 mM Tris-HCl, pH 7.4, 10 mM CaCl2, and 0.1% fatty acid-free bovine serum albumin. Incubations were carried out for 15 min at 37 °C with translational shaking. 20 µl of a 72-h COS cell supernatant were used to measure hGX sPLA2 activity. The released fatty acids were extracted by a modification of the Dole's procedure (68), methylated with diazomethane, and quantified by GC-MS measurement as described below. Control incubations in the absence of added sPLA2 were carried out in parallel and used to calculate specific hydrolysis.
GC-MS MeasurementsThe released fatty acids extracted by Dole's procedure were methylated with diazomethane and separated by gas chromatography on a capillary column of Supelcowax-10 bonded phase (0.32 mmx30 m, Supelco) with a Hewlett Packard 5890 Series II gas chromatograph. Fatty acids were detected by mass spectrometry (Nermag 10-10C, France) in the chemical ionization mode with ammonia (0.1 bar) as the reagent gas. The positive quasi-molecular ions were monitored and time-integrated. Quantification referred to an internal standard of heptadecanoic methyl ester with response factors was calculated with the various fatty acid methyl ester calibrators.
Protein sequences of various
sPLA2s were used to search for related sequences in gene
data bases by using the tBLASTn sequence alignment program (62). As a
result, we identified a human EST of 445 bp that was originally
obtained from a fetal lung cDNA library (Fig. 1).
The deduced amino acid sequence of the EST displayed a high homology to
other sPLA2s, including several amino acids invariably
conserved in catalytically active sPLA2s. We thus
postulated that this EST was a partial copy of a mRNA coding for a
novel low molecular mass sPLA2. This sPLA2 was
assigned as hGX sPLA2 in accordance with the recent
numbering system proposed by Dennis (2). Two primers were then designed
from the EST sequence and used in PCR amplification to screen for the
presence of hGX sPLA2 in human testis, human placenta, and
human fetal lung cDNAs. Only human fetal lung cDNAs produced
the expected 327-bp PCR fragment predicted from the EST sequence. This
fragment was cloned and sequenced, confirming its identity with the EST
sequence. This PCR fragment was then used as a probe to hybridize to
several human multiple Northern blots to analyze the transcription
pattern of hGX sPLA2. A 1.5-kb transcript coding for this
sPLA2 was found in spleen, thymus, and peripheral blood
leukocytes, while no such expression was observed in several other
human tissues, including human fetal lung (see Fig. 5). These data
indicate that this novel sPLA2 is expressed at a fairly low
level in a limited number of human tissues. This prompted us to request
the clone from which the EST sequence was derived at the "I.M.A.G.E.
consortium" (LLNL) cDNA clones data base (63). This clone was
found to contain a cDNA insert of 1020 bp bearing an open reading
frame of 166 amino acid residues, thus coding for the entire protein
sequence of the novel sPLA2 (Fig. 1). The 5-noncoding
region was found to contain an Alu repetitive element, while the short
3
-noncoding region contained a canonical polyadenylation site located
18 bases upstream from a poly(A) tail (Fig. 1).
After removal of the 5 Alu repeat, the cDNA clone was used as a
probe to map the hGX sPLA2 gene by in situ
hybridization on human metaphase chromosomes. In the 100 metaphase
cells examinated after in situ hybridization, there were 191 silver grains associated with chromosomes and 59 of these (30.9%) were
located on chromosome 16. The distribution of grains on this chromosome
was not random, since 74.6% (44/59) of them mapped to the p13.1-p12
region of chromosome 16 short arm (Fig. 2). These
results indicate that the hGX sPLA2 gene maps to the
16p13.1-p12 region of the human genome. The chromosomal localization of
this novel sPLA2 is thus distinct from those of other known
sPLA2s which have been mapped on chromosomes 1 and 12 (Table I) (40, 70, 71).
|
The hGX sPLA2 cDNA clone
predicts a mature sPLA2 protein of 123 amino acids
(calculated molecular mass, 13.6 kDa) and the presence of a signal
peptide with a maximal length of 42 amino acids (Fig. 1). The
calculated isoelectric point of the mature protein is 5.29, thus
placing hGX sPLA2 as the most acidic human sPLA2 (Table I). Sequence pattern analysis indicates that
hGX sPLA2 contains a potential glycosylation site at
position 71, which is also found in the sequence of the mouse group IIC
sPLA2 (Table I and Fig. 3). Two potential
initiator methionines located at positions 42 and
32 relative to
the mature protein are found in the hGX sPLA2 sequence
(Fig. 1). Interestingly, two potential initiator methionines were also
observed in the mouse group IIC sPLA2 (39). Although the
second methionine of the hGX sPLA2 sequence does not appear
in a more favorable context for initiation of translation as compared
with the first methionine (65), the use of this second methionine as a
translation initiation site will reduce the length of the prepropeptide
to 32 amino acids, i.e. to a size more consistent with those
of other sPLA2 signal peptide sequences (19, 25, 26, 36,
37, 39). The presence of an arginine doublet and of other polar
residues preceding the mature sPLA2 protein strongly
suggests that the signal sequence of hGX sPLA2 is a
prepropeptide. It was, however, difficult to assign the position of the
signal peptidase cleavage site that separates prepeptide and propeptide
sequences (72). In addition, no evident homology was observed between
the putative hGX sPLA2 propeptide sequence and that of
other sPLA2s containing propeptides (19, 39). Although the
maturation processing of hGX sPLA2 remains to be
elucidated, cleavage at dibasic amino acid motifs, including an
arginine doublet, is known to be efficiently catalyzed by
subtilisin-like protein convertases in the Golgi apparatus (73), thus
suggesting that hGX sPLA2 might be secreted as an active
enzyme after removal of its propeptide sequence into the cell
interior.
An alignment of the amino acid sequence of various mammalian
sPLA2s with the sequence of hGX sPLA2 is
presented in Fig. 3, and the biological properties of these
sPLA2s are compared in Table I. Fig. 3 shows that hGX
sPLA2 displays only 27-35% identity with other mammalian
sPLA2s, indicating no preferential homology with most of
the other sPLA2s. However, comparison of the complete hGX
protein sequence with data bases indicates a slightly higher level of
identity with group II sPLA2s including those of snake venom origin. For example, hGX sPLA2 displays 35% identity
with human group II sPLA2, while only 29% identity is
observed with human group I sPLA2. As a hallmark of all
other sPLA2s, the hGX sPLA2 protein contains a
large number of cysteines that are most probably assembled into
disulfide bridges. Of the 16 cysteines found in the structure of hGX
sPLA2, two are located at positions 11 and 77, which are
characteristic of group I sPLA2s, while two others are
found at positions 50 and 137, which are typical of group II
sPLA2s (Table I). The other cysteines are located at positions observed in both group I and II sPLA2s, but not
at positions that are characteristic of group IIC sPLA2s
(Fig. 3 and Table I). This cysteine pattern places the newly cloned
sPLA2 in a novel group, assigned as group X, according to
the PLA2 nomenclature recently refined by Dennis (2). In
fact, the cysteine pattern of hGX sPLA2 resembles that of
PLA2L3 (Table I), except that this latter
sPLA2 contains an extra cysteine at position 18 (Fig. 3).
It is also interesting to note that hGX sPLA2 has an amino acid C-terminal extension, which is characteristic of group II sPLA2s, while it contains neither the elapid loop
characteristic of group IA nor the pancreatic loop, a particular
feature of group IB sPLA2s, or the 4 amino acid insertion
found in group IIC sPLA2s (Table I and Fig. 3) (5). Fig. 3
also shows that hGX sPLA2 contains all of the amino acids
that are absolutely conserved in active sPLA2 enzymes,
indicating that the newly cloned sPLA2 could be
catalytically active.
A phylogenetic tree was derived from an alignment of all known protein
sequences of mammalian sPLA2s (Fig. 4). A
first division was observed in this tree between
PLA2L3/PLA2L5
and other mammalian sPLA2s, that places these two sPLA2 domains on
a branch distinct from that of other mammalian sPLA2s. This
is particularly interesting if one considers that these two branches
correspond to a separation between catalytically active and inactive
sPLA2s. Indeed, it is likely that the PLA2L3
and PLA2L5
domains (45), if expressed separately as
sPLA2 proteins, would correspond to catalytically inactive
enzymes, since these structures do not have several of the residues
found in catalytically active sPLA2s (Fig. 3). According to
this tree, a second duplication event would have occurred to generate
the present group I sPLA2s and another branch from which the other mammalian sPLA2s originate. Interestingly, hGX
sPLA2 appears to result from an ancestral duplication event
occurring after the separation of group I sPLA2s and other
sPLA2s, but before the divergence between group II and
group V sPLA2s (Fig. 4). Therefore, it is likely that
sPLA2s from groups II, V, and X have emerged from a common
ancestor, while sPLA2s from group I are more distantly related and arose before the development of these more recent groups.
This dendrogram is also in accordance with the chromosomal localization
of sPLA2s (Table I). Indeed, the three closely related human sPLA2s from group IIA, IIC, and V, which would have
emerged from recent gene duplication events, map to chromosome 1, while the more distant human group I and human group X sPLA2s map
to chromosome 12 and 16, respectively. Finally, it appears from this tree that group IIC sPLA2s are more distant from group IIA
and group V sPLA2s that would have emerged from a common
ancestor (Fig. 4).
Transcription Pattern of hGX sPLA2 and Comparison with That of Other Human sPLA2s
The tissue expression pattern of hGX sPLA2 was analyzed by probing several human multiple tissue Northern blots at high stringency (Fig. 5). Transcripts of 1.5 kb coding for hGX sPLA2 were detected in adult spleen and at a lower level in adult thymus and peripheral blood leukocytes. This result could indicate a specific expression of hGX sPLA2 in cells related to the immune system and inflammation. It is then possible that hGX sPLA2, as with group II and group V sPLA2s (3, 16, 27, 38, 74), may participate in the release of lipid mediators of inflammation. Whether the expression of hGX sPLA2, as with human group II sPLA2 (3, 16, 27, 74), is up-regulated in certain pathological conditions remains to be determined. Although the cDNA coding for hGX sPLA2 was originally cloned from a human fetal lung library, no transcript was detected in human fetal lung, indicating that hGX sPLA2 is not highly expressed in this fetal tissue (although the absence of expression may also be due to differences in sample preparations from different individuals). No expression was detected in other fetal tissues such as heart, liver, and brain (Fig. 5). Another low abundance 0.8-kb transcript was detected in the colon, lung, and pancreas, which could result from distinct sites of initiation or termination of transcription or from alternative splicing.
The pattern of expression of hGX sPLA2 was then compared with those of other human sPLA2s. In accordance with previous data (19, 55), a huge amount of human group I sPLA2 transcript was detected in adult pancreas and at lower levels in the spleen, lung, and prostate. No detectable expression was observed in the kidney and small intestine, contrasting with previous data (18, 75). Interestingly, a 0.5-kb transcript was also observed in the ovary as well as in fetal lung and fetal liver, while a transcript of higher size was detected in the testis (Fig. 5). Fig. 5 shows that human group II sPLA2 is widely distributed in several human adult tissues, including heart, liver, skeletal muscle, small intestine, and prostate. As for the human pancreatic sPLA2, human group II sPLA2 is expressed also in fetal liver (Fig. 5). These results are in accordance with the previous detection of human group II sPLA2 in a variety of mammalian tissues and cells, although the expression levels might be rather variable, depending on cell activation, which is known to highly modulate the expression of human group II sPLA2 (3, 16, 27, 29, 75). For example, no human group II sPLA2 transcripts are detected in spleen and peripheral blood leukocytes, from which human group II sPLA2 was characterized (3, 16, 27, 76). As previously observed (36), human group V sPLA2 is strongly expressed in adult heart and at lower levels in the lung and liver (Fig. 5). A low expression is also observed in the prostate, testis, ovary, small intestine, and colon, while no expression is detected in fetal tissues (Fig. 5). Taken together, it appears that the tissue distribution of these four human sPLA2s is different, suggesting diverse biological functions for each of them. Besides the expression of hGX sPLA2 in spleen, where both human group I (Fig. 5) and human group II sPLA2s (76) are also expressed, hGX sPLA2 is expressed in the thymus, a tissue in which other human sPLA2s are not detected (Fig. 5). The presence of a secretory PLA2 activity in the thymus either at adult stages or during the fetal thymic development is not well documented. Group II sPLA2 is known to be strongly expressed in rat thymus after endotoxic shock induced by lipopolysaccharides (77), and PLA2 activity has been detected in T lymphocytes and epithelial cells from rat thymus, but the molecular nature of these PLA2s has not been identified (78, 79). The biological role of PLA2 activity in thymus is poorly understood, although certain reports suggest a role in rat thymocyte apoptosis (80, 81). Interestingly, p-bromophenacyl bromide, a known inhibitor of sPLA2s (14, 30, 82, 83), prevents apoptosis of rat thymocytes and lymphoma cells (84).
Recombinant Expression of hGX sPLA2 and Characterization of sPLA2 ActivityTo test whether
the hGX sPLA2 cDNA really encodes a catalytically
active sPLA2, this cDNA was inserted into the
expression vector pcDNA I and then transfected into eukaryotic
cells. Three different constructs of hGX sPLA2 were
prepared. The first construct was generated by insertion of the
full-length cDNA clone (1020 bp) into the expression vector. Two
other constructs were prepared by a PCR-assisted strategy to insert
into the expression vector only the cDNA sequence coding for the
mature protein preceded by the prepropeptide sequence, starting either
at methionine-42 (Met-42 construct) or at methionine-32 (Met-32
construct). In addition, a consensus site for initiation of translation
was inserted in both of these latter constructs to increase the
efficiency of expression of hGX sPLA2 (65). The three
different constructs were transfected into COS cells, and the cell
supernatants were assayed for sPLA2 activity at different
times after transfection (Fig. 6). Cell supernatants
from COS cells transfected with the three different constructs
accumulated hGX sPLA2 activity, whereas cells transfected
with the parent vector did not show sPLA2 activity (Fig.
6A). A sPLA2 activity representing 6% of the
total cellular activity was recovered from the cell lysate obtained
72 h after cell transfection, indicating that hGX
sPLA2 was actively secreted by COS cells (Fig.
6A). The expression level of hGX sPLA2 activity obtained with Met-42 and Met-32 clones was about 10-fold higher relative to that observed with the full-length cDNA construct, probably due to the presence of the consensus leader sequence added in
the PCR-made constructs (Fig. 6). Furthermore, the expression level of
sPLA2 activity was not significantly modified when the translation of hGX sPLA2 started at methionine-32,
suggesting that this methionine may be functional in vivo as
an initiator site of translation. Finally, based on the presence of a
propeptide sequence ending with an arginine doublet in the structure of
hGX sPLA2, it was of interest to verify whether hGX
sPLA2 could be secreted as a zymogen. For that purpose, COS
cells were transfected and transferred into serum-free medium, and the
resulting supernatants were treated with trypsin and then assayed for
sPLA2 activity. Treatment with trypsin (which is assumed to
cleave efficiently at the propeptide arginine doublet if present) did
not significantly increase the supernatant sPLA2 activity,
suggesting that most of the sPLA2 released into the cell
incubation medium was already active and thus that the cleavage of the
propeptide sequence might have already occurred during the cell
maturation of hGX sPLA2. When human HEK 293 cells were used
instead of COS cells for the same experiments, similar results were
obtained, suggesting that hGX sPLA2 is an active enzyme
efficiently secreted by diverse eukaryotic cells (data not shown). As
shown in Fig. 6, B and C, recombinant
sPLA2 required 10 mM Ca2+ for
maximal enzymatic activity and was optimally active at physiological pH, with a rapid drop in sPLA2 activity observed below pH
6.0 and above pH 8.0.
The substrate preference of hGX sPLA2 was determined using as substrate artificial liposomes of natural lipids as well as erythrocyte ghost membranes. Table II indicates that hGX sPLA2 hydrolyzes phosphatidylethanolamine slightly more efficiently than phosphatidylcholine, while phosphatidylserine is poorly hydrolyzed. This substrate preference is very similar to that known for other sPLA2s (14, 25-27, 30). The results presented in Table II also indicate that hGX sPLA2 hydrolyzes more efficiently phospholipids containing polyunsaturated fatty acids at the sn-2 position, whatever the polar head group of phospholipids. This may suggest that hGX sPLA2 prefers lipids containing polyunsaturated fatty acids. Whether this apparent specificity is preserved for physiological substrates, such as membranes from live cells, remains to be addressed.
|
In conclusion, this report describes the molecular cloning and the characterization of a novel low molecular mass sPLA2. Besides its particular structural features, which assign this sPLA2 to a new group, the cloned enzyme is expressed in the spleen, thymus, and peripheral blood leukocytes where it may realize particular function(s). While the origin of cells that express this novel sPLA2 is currently unknown, hGX sPLA2 expression in the thymus, spleen, and leukocytes strongly suggests a role in relation to the immune system and/or inflammation. In that respect, it will be particularly interesting to analyze the expression of hGX sPLA2 during fetal thymic development and upon inflammatory challenge. The hGX sPLA2 gene has been mapped to chromosome 16, in a region that is closely associated with T cell leukemias (85). Further work is clearly needed to establish the exact physiological function(s) of this new sPLA2 and to determine whether it has any relevance in disease states.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) U95301[GenBank].
We thank M. Bordes and R. Waldmann for expert advice with computer program and Dr. A. Patel for a critical reading of the manuscript. The excellent technical assistance of N. Gomez is greatly appreciated and acknowledged, as well as the secretarial assistance of D. Doume and the expert photographical work of F. Aguila.