Human type IIA secretory phospholipase A2 (sPLA2-IIA) is induced in association with several immune-mediated inflammatory conditions. We have evaluated the effect of sPLA2-IIA on PG production in primary synovial fibroblasts from patients with rheumatoid arthritis (RA). At concentrations found in the synovial fluid of RA patients, exogenously added sPLA2-IIA dose-dependently amplified TNF-α-stimulated PGE2 production by cultured synovial fibroblasts. Enhancement of TNF-α-stimulated PGE2 production in synovial cells was accompanied by increased expression of cyclooxygenase (COX)-2 and cytosolic phospholipase A2 (cPLA2)-α. Blockade of COX-2 enzyme activity with the selective inhibitor NS-398 prevented both TNF-α-stimulated and sPLA2-IIA-amplified PGE2 production without affecting COX-2 protein induction. However, both sPLA2-IIA-amplified PGE2 production and enhanced COX-2 expression were blocked by the sPLA2 inhibitor LY311727. Colocalization studies using triple-labeling immunofluorescence microscopy showed that sPLA2-IIA and cPLA2-α are coexpressed with COX-2 in discrete populations of CD14-positive synovial macrophages and synovial tissue fibroblasts from RA patients. Based on these findings, we propose a model whereby the enhanced expression of sPLA2-IIA by RA synovial cells up-regulates TNF-α-mediated PG production via superinduction of COX-2. Therefore, sPLA2-IIA may be a critical modulator of cytokine-mediated synovial inflammation in RA.

Rheumatoid arthritis (RA)4 is a systemic autoimmune disease characterized by inflammation of the joint synovium that results in pain, joint erosion, and dysfunction. The severity of joint inflammation fluctuates, resulting in exacerbations and remissions of disease activity. The cytokines, TNF-α and IL-1β, play an important role in the pathogenesis of RA. Neutralization of these cytokines alleviates synovial inflammation in both animal models and human RA (1). These animal studies, together with IL-1β and TNF-α in vivo gene deletion experiments, have shown that IL-1β is consistently important in mediating cartilage breakdown, whereas TNF-α is a key inducer of synovial inflammation (2). Moreover, transgenic mice overexpressing TNF-α develop spontaneous arthritis (3). TNF-α and IL-1β activate the transcription factor NF-κB (4) and the p38 and c-Jun N-terminal kinase mitogen-activated protein kinase (MAPK) pathways (5) to induce a host of proinflammatory proteins. Direct inhibition of NF-κB (6), Jun/Fos (AP-1) (7) or p38 MAPK (8) reduces disease severity in animal models of arthritis, confirming the importance of these signaling pathways in the inflammatory and/or the erosive component of arthritis.

Prostaglandin E2 (PGE2) contributes to pain and swelling during inflammation through induction of hyperalgesia and increased vascular permeability (9) and modulates bone resorption through stimulation of osteoclast formation from precursor stem cells (10). PGE2 production by cultured rheumatoid synovial fibroblasts (RSFs) is induced within hours by IL-1β mediated by NF-κB- and MAPK-dependent coordinate induction of cytosolic phospholipase A2 (cPLA2)-α and cyclooxygenase (COX)-2 (11, 12). Although COX-1 is constitutively expressed by RSFs, IL-1β-stimulated PGE2 production occurs exclusively via COX-2 (13). In RA synovium both COX-1 and -2 are expressed, with COX-2 expression elevated in relation to the degree of inflammation in synovial tissue (14). Recently, COX-2-selective inhibitors that maintain the anti-inflammatory properties of nonsteroidal anti-inflammatory drugs (NSAIDs) (15, 16) have been developed; but, unlike the latter compounds, they have a favorable gastrointestinal side effect profile.

A low molecular mass (14 kDa) human type IIA secretory phospholipase A2 (sPLA2-IIA) has been identified in rheumatoid synovium (17); however, the importance of this enzyme in synovial pathology is poorly defined. sPLA2-IIA, first purified from the synovial fluid of patients with RA (18), is found at high levels in the colonic mucosa of patients with ulcerative colitis and Crohn’s disease (19), in the bronchoalveolar lavage fluid of patients with asthma following Ag challenge (20) and in the serum of septic shock patients (21). Serum sPLA2-IIA concentrations are elevated in patients with RA (22) and correlate with severity of disease (23). Enzyme expression is increased in RA synovial macrophages and fibroblasts relative to synovium from nonarthritic patients and correlates with histological markers of synovial inflammation (17). Human sPLA2-IIA is acutely inflammatory when injected into rabbit joints (24); however, transgenic mice overexpressing sPLA2-IIA do not develop arthritis (25).

Given the presence of sPLA2-IIA at concentrations up to several micrograms per milliliter in RA synovial fluid (26), and the established inflammatory activities of PGs (9), we have examined here both the relationship between sPLA2-IIA and PGE2 production in cultured RSFs and the cellular localization of sPLA2-IIA and COX-2 in rheumatoid synovial tissue. The results of these studies demonstrate that concentrations of sPLA2-IIA found in RA synovial fluid enhance TNF-α-stimulated PGE2 production in RSFs by superinducing COX-2 protein levels. Further, sPLA2-IIA and COX-2 colocalize in discrete subpopulations of rheumatoid synovial macrophages and fibroblasts. These findings indicate that sPLA2-IIA may be an important amplifier of cytokine-mediated PG production and may thereby contribute to the severity of the synovial inflammatory response in RA.

sPLA2-IIA was purified from the conditioned medium and cell pellets of a Chinese hamster ovary cell line (5A2) stably expressing human sPLA2-IIA cDNA (27) by immunoaffinity chromatography on an AKTA explorer system (Pharmacia Biotech, Uppsala, Sweden) and quantified by ELISA (22). sPLA2-IIA was a single 14-kDa band on silver-stained PAGE gels, and N-terminal amino acid sequence analysis (27) confirmed its identity. sPLA2-IIA contained <0.1 ng endotoxin/mg protein (Limulus amebocyte lysate pyrochrome, Associates of Cape Cod, Falmouth, MA) and was enzymatically active in a [3H]arachidonate-labeled Escherichia coli membrane assay (27).

Synovial tissues were obtained as described (17) using procedures approved by the St. Vincent’s Hospital Ethics Committee. RSFs were isolated by trypsin (0.5%)/EDTA (5.3 mM) digestion (15 min, 37°C in DMEM/Ham’s F12 medium), followed by collagenase (200 U/ml; Life Technologies, Gaithersburg, MD), and cells were grown to confluence twice in medium containing 10% FBS (Commonwealth Serum Laboratory, Melbourne, Australia), penicillin (100 U/ml), streptomycin (100 μg/ml), and amphotericin B (0.125 μg/ml) (Life Technologies), before storage in liquid nitrogen. Cells were phenotypically stable (CD14-negative, prolyl-5-hydroxylase-positive by immunofluorescence microscopy) to passage and were used from passage 4 to 10. The human neonatal lung fibroblast (NLF) cell line CCD34Lu (American Type Culture Collection, Manassas, VA) was grown in 2% FBS to synchronize the growth rate with that of the RSFs. For all experiments, confluent cell monolayers were grown antibiotic-free and received fresh medium containing 0.1% BSA (endotoxin-free, fatty acid-free; Boehringer Mannheim, Sydney, Australia) before stimulation.

Inhibitors LY311727 (28) and NS-398 (29) (Cayman Chemical, Ann Arbor, MI) were prepared as 10-mM stocks in DMSO. Both control and inhibitor-treated cultures contained a final solvent concentration of 0.1% (v/v) DMSO. LY311727 was inhibitory toward sPLA2-IIA (IC50 2.5 μM at an sPLA2-IIA concentration of 10 ng/ml, [3H]E. coli membrane assay (27)), whereas NS-398 was noninhibitory up to 100 μM (data not shown). Confluent fibroblast monolayers were stimulated with sPLA2-IIA, human rTNF-α (PeproTech, Rocky Hill, NJ), or IL-1β (R&D Systems, Minneapolis, MN) alone or in combination, in the presence or absence of inhibitors for 24 h, and media were stored at −80°C. PGE2 was quantified in triplicate at three dilutions by enzyme immunoassay (Cayman Chemical).

Cell lysates were prepared by resuspension in PBS containing 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM EGTA, aprotinin (50 μg/ml), leupeptin (200 μM), and PMSF (1 mM) followed by repeated passage through a 21-gauge needle, incubation on ice (30 min), centrifugation (20 min, 13,800 × g, 4°C), and storage of supernatants at −80°C. Lysates were analyzed by Western blotting following SDS-PAGE (4–20% gradient gels; Novex, San Diego, CA). Primary Abs were anti-ovine COX-1 mAb (cat. no. 160110, 1.7 μg/ml; Cayman Chemical), anti-human COX-2 mAb (cat. no. 160112, 0.5 μg/ml; Cayman Chemical), anti-human cPLA2-α mAb (cat. no. sc-454, 0.1 μg/ml; Santa Cruz Biotechnology, Santa Cruz, CA), anti-β-actin mAb (clone AC-15, 0.3 μg/ml; Sigma, St. Louis, MO), and anti-human ICAM-1 goat polyclonal Ab (cat. no. sc-1510, 0.2 μg/ml; Santa Cruz Biotechnology). Proteins were visualized using enhanced chemiluminescence (NEN, Boston, MA) and quantified by densitometry (Molecular Dynamics, Sunnyvale, CA).

Synovial tissue from five RA patients receiving, alone or in combination, NSAIDs, auranofin, azathioprine, cyclosporin, sulfasalazine, or methotrexate, but not prednisolone, was examined by immunofluorescence microscopy as described (17). Sections were sequentially incubated with anti-human COX-2 goat polyclonal Ab (cat. no. sc-1745, 5 μg/ml, 45 min; Santa Cruz Biotechnology) or negative control goat IgG (5 μg/ml; Sigma), donkey anti-goat rhodamine red-X conjugate (1/100, 30 min; Jackson ImmunoResearch, West Grove, PA), a second primary Ab (45 min), and donkey anti-mouse-indodicarbocyanine (Cy5) conjugate (1/200, 30 min; Jackson ImmunoResearch). For double-labeling studies, second primary Abs were anti-human CD68 (clone EBM11, 4.3 μg/ml; Dako, Glostrup, Denmark), anti-human prolyl 4-hydroxylase (clone 5B5, 1.5 μg/ml; Dako), anti-human Von Willebrand factor (clone F8/86, 4.4 μg/ml; Dako), or isotype-matched (IgG1k) negative control mAb 81193 (Bioquest, Sydney, Australia).

For triple-labeling studies, second primary Abs were anti-human sPLA2 (clone 9C1 (22), 2 μg/ml; Bioquest), isotype-matched (IgG1k) negative control 81193 (2 μg/ml), anti-human cPLA2-α (4 μg/ml), or isotype-matched (IgG2b) negative control (clone MOPC-195, 4 μg/ml; Immunotech, Marseille, France). The specificity of Ab 9C1 for sPLA2-IIA in immunohistochemistry has been partially but not exhaustively determined (13, 17). The Ab recognizes a conformational epitope on sPLA2-IIA, identifies sPLA2-expressing cells in rheumatoid synovium with the same specificity as two other independent mAbs (4A1 and 10B2) raised to sPLA2-IIA, and a capture ELISA using 9C1 together with Ab 4A1 does not recognize human sPLA2-V or sPLA2-1B. Sections were blocked (4% mouse serum, 30 min; Sigma) and incubated with CD14 mAb-FITC conjugate (clone RMO52, monocyte marker, 1/5; Immunotech) or isotype-matched (IgG2a) negative control mAb-FITC conjugate (clone U7.27, 1/5; Immunotech). Sections were analyzed by confocal microscopy as described (17). Images were exported into Adobe Photoshop 5.0 (Adobe Systems, San Jose, CA) and Canvas 5.02 (Deneba Software, Miami, FL) for presentation.

Statistical evaluations were performed on primary data using the Wilcoxon signed rank test or the Student’s unpaired t test with StatView software (Abacus Concepts, Berkeley, CA).

To determine the responsiveness of the cPLA2-α/COX pathway to TNF-α, RSFs were stimulated with increasing concentrations of TNF-α, then PGE2 production and COX and cPLA2-α protein levels were determined. Basal PGE2 production varied between cultures (range 46 ± 2 to 245 ± 10 pg/ml); however, in each of the five cultures, TNF-α dose-dependently stimulated PGE2 production (range 0.32 ± 0.07 to 13.4 ± 3.1 ng/ml at 10 ng/ml TNF-α) (Fig. 1,A). Increased PGE2 production correlated with accumulation of COX-2 protein (Fig. 1,B). cPLA2-α was weakly but consistently up-regulated with COX-2, whereas COX-1 was unaffected. The adhesion molecule ICAM-1 was near-maximally induced at 1 ng/ml TNF-α (Fig. 1 B).

FIGURE 1.

TNF-α-stimulated induction of PGE2 production and COX-2 protein expression in RSF. PGE2 production (A) and COX-2, COX-1, cPLA2-α, ICAM-1, or β-actin protein levels (B) following TNF-α stimulation (PGE2, 24 h; protein, 18 h) was measured in RSF cultures. Data are from triplicate determinations in RSF cultures established from five independent RA patients and are presented as fold increase in PGE2 (mean ± SE) relative to untreated controls. PGE2 concentration in unstimulated cultures (n = 5) ranged from 46 ± 2 to 245 ± 10 pg/ml. ∗∗, p < 0.01; ∗∗∗, p < 0.001 relative to unstimulated control, Wilcoxon signed rank. Western blot data are representative (n = 3).

FIGURE 1.

TNF-α-stimulated induction of PGE2 production and COX-2 protein expression in RSF. PGE2 production (A) and COX-2, COX-1, cPLA2-α, ICAM-1, or β-actin protein levels (B) following TNF-α stimulation (PGE2, 24 h; protein, 18 h) was measured in RSF cultures. Data are from triplicate determinations in RSF cultures established from five independent RA patients and are presented as fold increase in PGE2 (mean ± SE) relative to untreated controls. PGE2 concentration in unstimulated cultures (n = 5) ranged from 46 ± 2 to 245 ± 10 pg/ml. ∗∗, p < 0.01; ∗∗∗, p < 0.001 relative to unstimulated control, Wilcoxon signed rank. Western blot data are representative (n = 3).

Close modal

Treatment of RSFs with TNF-α (10 ng/ml) in the presence of the COX-2-selective inhibitor NS-398 (1 μM) (29) abrogated TNF-α-stimulated PGE2 production whereas the sPLA2-selective inhibitor LY311727 (28) reproducibly suppressed PGE2 by 30% (Fig. 2).

FIGURE 2.

Effect of sPLA2 and COX-2 inhibitors on TNF-α-stimulated PGE2 production. PGE2 production was measured in RSFs (n = 5) treated for 24 h with 0.1% DMSO (Control) or 0.1% DMSO containing TNF-α (10 ng/ml) either alone or in combination with an sPLA2-selective inhibitor LY311727 (10 μM) or a COX-2-selective inhibitor NS-398 (1 μM). Control PGE2 range was 46 ± 2 to 245 ± 10 pg/ml. ∗∗∗, p < 0.001 relative to TNF-α-stimulated cultures, Wilcoxon signed rank.

FIGURE 2.

Effect of sPLA2 and COX-2 inhibitors on TNF-α-stimulated PGE2 production. PGE2 production was measured in RSFs (n = 5) treated for 24 h with 0.1% DMSO (Control) or 0.1% DMSO containing TNF-α (10 ng/ml) either alone or in combination with an sPLA2-selective inhibitor LY311727 (10 μM) or a COX-2-selective inhibitor NS-398 (1 μM). Control PGE2 range was 46 ± 2 to 245 ± 10 pg/ml. ∗∗∗, p < 0.001 relative to TNF-α-stimulated cultures, Wilcoxon signed rank.

Close modal

RSFs were treated with increasing concentrations of sPLA2-IIA in the presence or absence of TNF-α (10 ng/ml). sPLA2-IIA alone did not stimulate PGE2 production at any concentration examined (Fig. 3,A). Coaddition of sPLA2-IIA with TNF-α resulted in a dose-dependent enhancement of TNF-α-stimulated PGE2 production with a mean 3-fold augmentation of the TNF-α response over the five RSF cultures at 10 μg/ml sPLA2-IIA (Fig. 3 A). The response was significant at both 1 and 10 μg/ml sPLA2-IIA, even though only four of the five RSF cultures were responsive.

FIGURE 3.

Effect of sPLA2-IIA on PGE2 production and COX-2 protein expression. A, PGE2 production was measured in RSFs (n = 5) treated for 24 h with increasing concentrations of sPLA2-IIA alone (○) or in combination with TNF-α (10 ng/ml) (•). Unstimulated PGE2 concentration range was 46 ± 2 to 245 ± 10 pg/ml. ∗, p < 0.05; ∗∗, p < 0.01 relative to TNF-α-stimulated control, Wilcoxon signed rank. Protein levels for COX-2, COX-1, cPLA2-α, or β-actin (BD) were measured by SDS-PAGE immunoblotting and densitometry in RSF cultures (n = 3) stimulated for 18 h with either TNF-α (10 ng/ml, B and C) or sPLA2-IIA (5 μg/ml) alone or in combination or (D) increasing concentrations of TNF-α in the presence or absence of sPLA2-IIA (5 μg/ml).

FIGURE 3.

Effect of sPLA2-IIA on PGE2 production and COX-2 protein expression. A, PGE2 production was measured in RSFs (n = 5) treated for 24 h with increasing concentrations of sPLA2-IIA alone (○) or in combination with TNF-α (10 ng/ml) (•). Unstimulated PGE2 concentration range was 46 ± 2 to 245 ± 10 pg/ml. ∗, p < 0.05; ∗∗, p < 0.01 relative to TNF-α-stimulated control, Wilcoxon signed rank. Protein levels for COX-2, COX-1, cPLA2-α, or β-actin (BD) were measured by SDS-PAGE immunoblotting and densitometry in RSF cultures (n = 3) stimulated for 18 h with either TNF-α (10 ng/ml, B and C) or sPLA2-IIA (5 μg/ml) alone or in combination or (D) increasing concentrations of TNF-α in the presence or absence of sPLA2-IIA (5 μg/ml).

Close modal

Although sPLA2-IIA alone had no effect on PGE2 production in RSFs (Fig. 3,A), sPLA2-IIA (5 μg/ml) consistently increased COX-2 protein to a level similar to that induced by TNF-α (10 ng/ml) (Fig. 3, BD). Coaddition of sPLA2-IIA with TNF-α resulted in a synergistic increase in COX-2 protein levels (Fig. 3, BD). The increase over sPLA2-IIA alone was clearly detectable at 0.1 ng/ml TNF-α (Fig. 3,D). COX-1 protein was not affected by these treatments, whereas a small increase in cPLA2-α protein levels was consistently observed with TNF-α (>0.1 ng/ml)/sPLA2-IIA treatment (Fig. 3, B and D). sPLA2-IIA did not increase TNF-α-stimulated ICAM-1 protein expression (data not shown). Although sPLA2-IIA also showed augmentation of IL-1β-stimulated PGE2 production and COX-2 induction (data not shown), the response was small relative to sPLA2-IIA augmentation of the TNF-α-stimulated response and did not occur consistently in RSF cultures.

To determine whether the responsiveness of RSFs to sPLA2-IIA was a general feature of human fibroblasts, comparable experiments were performed using a human NLF cell line. In NLF cells, PGE2 production was significantly but weakly increased by TNF-α (10 ng/ml) from 71 ± 3 pg/ml (n = 4) to 90 ± 10 pg/ml (n = 2) (p < 0.05). However, the weak induction was not due to a lack of responsiveness of the pathway to stimulation because IL-1β (0.1 ng/ml) stimulated basal PGE2 production by 12-fold (p < 0.001) (Student’s unpaired t test). Also, COX-1, COX-2, cPLA2-α, and ICAM-1 expression following TNF-α or IL-1β stimulation was similar to that observed in RSFs (data not shown). NS-398 (1 μM) abrogated TNF-α-stimulated PGE2 production, whereas LY311727 (10 μM) had no effect. sPLA2-IIA (10 μg/ml) did not augment TNF-α-stimulated PGE2 production by the NLF cell line. In NLF cultures, sPLA2-IIA, like TNF-α, also increased COX-2 protein expression, whereas coaddition of sPLA2-IIA and TNF-α had only an additive effect on COX-2 protein levels that was not associated with increased PGE2 production over TNF-α alone (data not shown).

RSFs were stimulated with TNF-α (10 ng/ml) and increasing concentrations of sPLA2-IIA in the presence or absence of the sPLA2 inhibitor LY311727 (10 μM). LY311727 reduced both PGE2 production (Fig. 4,A) and COX-2/cPLA2-α protein (Fig. 4, B and C) to levels observed in TNF-α-stimulated RSFs. The COX-2-selective inhibitor NS-398 (1 μM) reduced PGE2 production to basal levels (Fig. 4,A) without affecting TNF-α/sPLA2-IIA-stimulated COX-2 protein (Fig. 4, B and C). Levels of COX-1 were not affected by either treatment (Fig. 4 B).

FIGURE 4.

Effect of sPLA2 and COX-2 inhibitors on sPLA2-IIA-stimulated PGE2 production and COX-2 induction. PGE2 production (A) was measured in RSFs (n = 5) treated for 24 h with 0.1% DMSO and increasing concentrations of sPLA2-IIA plus TNF-α (10 ng/ml) either alone (○), or in combination with LY311727 (10 μM) (•) or NS-398 (1 μM) (□). Unstimulated PGE2 levels ranged from 46 ± 2 to 245 ± 10 pg/ml. ∗∗, p < 0.01; ∗∗∗, p < 0.001 between LY311727-treated cells and sPLA2-IIA/TNF-α-treated cells, Wilcoxon signed rank. COX-2, COX-1, cPLA2-α, or β-actin levels (B and C) were measured by SDS-PAGE immunoblotting and densitometry in RSF cultures (n = 3) stimulated for 18 h with sPLA2-IIA (5 μg/ml) plus TNF-α (10 ng/ml) and 0.1% DMSO either alone (TNF-α/sPLA2) or in combination with LY311727 (10 μM) or NS-398 (1 μM).

FIGURE 4.

Effect of sPLA2 and COX-2 inhibitors on sPLA2-IIA-stimulated PGE2 production and COX-2 induction. PGE2 production (A) was measured in RSFs (n = 5) treated for 24 h with 0.1% DMSO and increasing concentrations of sPLA2-IIA plus TNF-α (10 ng/ml) either alone (○), or in combination with LY311727 (10 μM) (•) or NS-398 (1 μM) (□). Unstimulated PGE2 levels ranged from 46 ± 2 to 245 ± 10 pg/ml. ∗∗, p < 0.01; ∗∗∗, p < 0.001 between LY311727-treated cells and sPLA2-IIA/TNF-α-treated cells, Wilcoxon signed rank. COX-2, COX-1, cPLA2-α, or β-actin levels (B and C) were measured by SDS-PAGE immunoblotting and densitometry in RSF cultures (n = 3) stimulated for 18 h with sPLA2-IIA (5 μg/ml) plus TNF-α (10 ng/ml) and 0.1% DMSO either alone (TNF-α/sPLA2) or in combination with LY311727 (10 μM) or NS-398 (1 μM).

Close modal

To determine whether the expression of PLA2 enzymes was detectable in synovial tissue cells expressing COX-2, the localization of COX-2 relative to sPLA2-IIA and cPLA2-α in RA synovial membrane sections was examined using immunofluorescence confocal microscopy. The population of cells expressing COX-2 was first defined by double immunofluorescence with Abs to macrophage-like cells (CD68), fibroblast-like cells (5B5), and endothelial cells (factor VIII) (data not shown). COX-2 staining intensity in synovial sections derived from five independent RA patients was consistently strongest in the macrophage-like cells. Positive staining was also observed in both synovial lining and subsynovial lining fibroblast-like cells, whereas endothelial cells were usually negative or weakly COX-2 positive. Overall, the COX-2 staining pattern observed was consistent with that described previously for RA (14) with the exception that strong COX-2-positive endothelial cell staining was not consistently seen.

RA synovium from three patients was then triple-labeled with Abs to COX-2, CD14, and sPLA2 (Fig. 5) or COX-2, CD14, and cPLA2-α (Fig. 6). All cells positive for COX-2 were also sPLA2-positive (Fig. 5,D), although sPLA2-positive/COX-2-negative cells were observed. The majority of CD14-positive cells were COX-2-negative, and a subpopulation of those cells were sPLA2-positive (data not shown). A small number of COX-2/CD14/sPLA2-positive cells were observed. The pattern of staining described (Fig. 5) is consistent with that observed in synovial sections derived from the two other RA patients examined (data not shown). All COX-2-positive cells were also cPLA2-α-positive, although cPLA2-α-positive/COX-2-negative cells were observed, notably the endothelial cells lining vessels (Fig. 6). As in the sPLA2 studies, the majority of CD14-positive cells were COX-2-negative, and a subpopulation of these were cPLA2-α-positive. Several COX-2/CD14/cPLA2-α-positive cells were observed, as were CD14-negative, cPLA2-α- positive, and COX-2-positive cells.

FIGURE 5.

Colocalization of sPLA2 with COX-2 in RA synovium. Synovial sections from patients with RA were stained simultaneously and imaged for COX-2 (A, red), sPLA2 (B, blue), and CD14 (C, green). The overlaid image is shown in D. ▸, sPLA2/COX-2-positive, CD14-negative cell; →, sPLA2-positive, COX-2/CD14-negative cell; and ➤, sPLA2/COX-2/CD14-positive cell. Adjacent sections were used for control staining with isotype-matched irrelevant primary Ab controls: goat IgG (E), IgG1κ murine monoclonal 81193 (F), and an IgG2a mAb-FITC conjugate (G). The composite image of the three negative controls is shown in H. The representative section presented was derived from a patient on NSAIDs but not disease modifying anti-rheumatic drugs, and the image shown is taken from an avascular region of synovial sublining layer. Bar = 10 μm.

FIGURE 5.

Colocalization of sPLA2 with COX-2 in RA synovium. Synovial sections from patients with RA were stained simultaneously and imaged for COX-2 (A, red), sPLA2 (B, blue), and CD14 (C, green). The overlaid image is shown in D. ▸, sPLA2/COX-2-positive, CD14-negative cell; →, sPLA2-positive, COX-2/CD14-negative cell; and ➤, sPLA2/COX-2/CD14-positive cell. Adjacent sections were used for control staining with isotype-matched irrelevant primary Ab controls: goat IgG (E), IgG1κ murine monoclonal 81193 (F), and an IgG2a mAb-FITC conjugate (G). The composite image of the three negative controls is shown in H. The representative section presented was derived from a patient on NSAIDs but not disease modifying anti-rheumatic drugs, and the image shown is taken from an avascular region of synovial sublining layer. Bar = 10 μm.

Close modal
FIGURE 6.

Colocalization of cPLA2-α with COX-2 in RA synovium. The representative section shown is from the patient described in Fig. 5. Synovial sections from RA patients were simultaneously stained for COX-2 (A, red), cPLA2-α (B, blue), and CD-14 (C, green). The composite image from all three channels is shown in D. ➤, COX-2/cPLA2-α/CD14-positive cell; ▸, COX-2/cPLA2-α-positive CD14-negative cell; and →, cPLA2-α-positive COX-2/CD-14-negative cell. Adjacent sections were stained with isotype-matched negative control Abs: goat IgG (E), IgG2b clone MOPC-195 (F), and IgG2a-FITC conjugate (G). The composite image from all three channels for the control is shown in H. The image is taken from the sublining layer (sll) of the synovium. v, Vessel. Bar = 50 μm.

FIGURE 6.

Colocalization of cPLA2-α with COX-2 in RA synovium. The representative section shown is from the patient described in Fig. 5. Synovial sections from RA patients were simultaneously stained for COX-2 (A, red), cPLA2-α (B, blue), and CD-14 (C, green). The composite image from all three channels is shown in D. ➤, COX-2/cPLA2-α/CD14-positive cell; ▸, COX-2/cPLA2-α-positive CD14-negative cell; and →, cPLA2-α-positive COX-2/CD-14-negative cell. Adjacent sections were stained with isotype-matched negative control Abs: goat IgG (E), IgG2b clone MOPC-195 (F), and IgG2a-FITC conjugate (G). The composite image from all three channels for the control is shown in H. The image is taken from the sublining layer (sll) of the synovium. v, Vessel. Bar = 50 μm.

Close modal

Though sPLA2-IIA is significantly elevated at sites of inflammation in immune-mediated disorders, the role of secreted PLA2 forms in PG production in primary cells relevant to human inflammatory pathology or in the clinical manifestations of inflammation in vivo is not clear. Transgenic overexpression of sPLA2-IIA causes hyperplasia of the skin but not arthritis (25). Blocking sPLA2-IIA activity with selective inhibitors in rat adjuvant-induced arthritis shows inconsistent results between studies, whereas inhibitors are consistently effective in blocking acute carrageenan-induced inflammation (30, 31). However, interpretation of intervention studies blocking sPLA2 in animal models is complicated because there are at least nine mammalian forms of sPLA2 (32, 33, 34, 35, 36) and the tissue distribution and regulation of expression of these forms is different in rodents relative to humans. We show here that, unlike TNF-α alone, sPLA2-IIA alone has no effect on PGE2 production in RSFs, although it consistently up-regulated COX-2 protein to levels comparable to those seen with TNF-α alone. These data suggest that COX-2 up-regulation is necessary but insufficient for PGE2 production. Additional signals are provided by TNF-α stimulation, which sPLA2-IIA alone cannot provide. It is possible that sPLA2-IIA, although enzymatically active, can neither supply substrate directly to COX-2 nor indirectly activate cPLA2-α. TNF-α signaling likely results in post-translational activation of cPLA2-α at the level of phosphorylation and/or mobilization to membranes. Alternatively, TNF-α may also regulate the recently cloned PGE synthase (37) immediately distal to COX-2 in the PGE2 pathway. Importantly, COX-2 has recently been shown to mediate effects on cellular proliferation independently of its enzyme activity (38). Thus, sPLA2-IIA, via induction of COX-2 protein, may have broader effects on cell function than modulation of PG synthesis alone.

sPLA2-IIA amplifies TNF-α-induced PGE2 production by RSFs at concentrations that are found in the synovial fluid of patients with RA (26). Importantly, this amplification occurs over a range of TNF-α concentrations, suggesting that expression of sPLA2-IIA in synovium may sensitize synovial cells to produce PGs at low concentrations of TNF-α, thereby contributing to the severity of the PG-mediated synovial inflammatory response. This suggestion is supported by the observation that the spontaneous arthritis resulting from transgenic overexpression of TNF-α in mice (3) is exacerbated and is earlier in onset when human sPLA2-IIA is transgenically overexpressed in combination with TNF-α (39).

Our experiments with NLFs show that amplification of TNF-α-induced PGE2 production by sPLA2-IIA is not a general feature of human fibroblasts, even though a functional cytokine-inducible COX-2-dependent PGE2 production pathway is present in these cells. Further, sPLA2-IIA did not consistently amplify IL-1β-stimulated PGE2 production in RSFs, suggesting that IL-1β alone may be sufficient to maximally stimulate the cPLA2-α/COX-2 pathway in these cells. The relative roles of TNF-α and IL-1β in stimulating synovial inflammation is a matter of controversy. Notably, both cytokines are capable of inducing synovial inflammation, and there is data suggesting that each is dependent on the other. Also, TNF-α is important in the induction of synovial inflammation in RA, whereas IL-1β, although able to cause inflammation, is more consistently potent at mediating cartilage degradation (2). It is likely that there is significant variability in the cytokine profiles of patients with RA depending on both genetic and environmental factors. Consequently, sPLA2-IIA effects may also vary depending on the local synovial cytokine environment.

Both TNF-α-stimulated and sPLA2-IIA augmentation of TNF-α-stimulated PGE2 production by RSFs is COX-2-dependent. This finding is consistent with the effects of exogenous sPLA2 in studies in model rodent cell lines using other agonists such as nerve growth factor stimulation of rat mast cells (40, 41). In addition, in some, but not all cases where sPLA2 has been reported to augment agonist-stimulated PG production, COX-2 protein is also up-regulated (40, 41) as has been shown here. Our immunofluorescence studies show that sPLA2 is coexpressed with COX-2 in specific subpopulations of fibroblast and CD14-positive macrophages in rheumatoid synovium, supporting the relevance of our observations with synovial cells in culture to synovium. From these studies, it is likely that the selective COX-2 inhibitors now in clinical use (16) would effectively block sPLA2-IIA amplification of TNF-α-stimulated PG production in synovium.

The sPLA2 inhibitor LY311727, which binds in the active site channel of sPLA2 to block enzyme activity (28), suppressed both sPLA2-IIA-augmented PGE2 production and the concomitant up-regulation of COX-2, suggesting that regulation of COX-2 protein levels requires sPLA2 enzyme activity. However, an activity-independent mechanism of sPLA2-IIA-mediated up-regulation of COX-2 protein levels in RSFs cannot be ruled out in light of a recent report in rat serosal mast cells, where an “activity-dead” sPLA2-IIA mutant enzyme still augmented COX-2 up-regulation in response to nerve growth factor even though induction of COX-2 by nerve growth factor was suppressible by LY311727 (42). Importantly, in contrast to findings in a murine osteoblast cell line (41), the insensitivity of the TNF-α/sPLA2-IIA-mediated induction of COX-2 to NS-398 rules out an autocrine effect of PGE2 on this pathway in RSFs.

Cytokine induction of the genes encoding COX-2 and cPLA2-α is NF-κB-dependent in RSFs (12); however, sPLA2-IIA does not affect TNF-α-induced mobilization of NF-κB into the nucleus as determined by EMSAs, Western blots for nuclear p65, or IκB degradation assays (M. J. Bidgood, M. L. Taberner, and K. F. Scott, manuscript in preparation). COX-2 is also post-transcriptionally regulated via a p38-MAPK-dependent mechanism in human monocytes (43). It is known that TNF-α can signal through the p38 pathway in chondrocytes (5), and it is possible that sPLA2-IIA costimulates this pathway in RSFs. However, in our hands, sPLA2-IIA does not induce phosphorylation of p38 MAPK nor does it stimulate TNF-α-mediated phosphorylation of p38 MAPK in RSFs (M. J. Bidgood, M. L. Taberner, and K. F. Scott, manuscript in preparation). Consequently, the mechanism by which sPLA2-IIA amplifies COX-2 protein expression in RSFs remains to be established.

Our studies support the view that sPLA2-IIA is one factor that may amplify TNF-α-dependent pathways in rheumatoid synovium and that the level of expression of sPLA2-IIA in synovium, together with that of TNF-α, may contribute to the severity of the PG-mediated inflammatory response. Levels of sPLA2-IIA are also increased in several other immune-mediated conditions. It is possible that sPLA2-IIA may be a severity factor in these conditions also. Genome scanning studies have shown that several non-HLA loci contribute to disease susceptibility and severity in humans and in animal models of immune-mediated disease. Susceptibility loci identified in several autoimmune disorders, including RA (44), cluster to discrete chromosomal regions (45), suggesting that immune-mediated diseases may have genetic features that are shared, despite their diverse clinical manifestations. One of these loci (human chromosome 1p36-ter), identified as a susceptibility locus in multiple sclerosis and Crohn’s disease (45) and later in RA (44), overlap with the chromosomal location of the genes encoding sPLA2-IIA, -IID, -IIE, and -V, 1p35–36 (33, 34, 46). Thus, as suggested by our biochemical studies and by transgenic overexpression of sPLA2-IIA with TNF-α (39), sPLA2-IIA may be one among several candidate genes in this region in which mutations that lead to variation in expression may contribute to onset and/or severity of immune-mediated inflammatory diseases in humans.

We thank Derek van Dyk, Department of Biotechnology, University of New South Wales, for fermentation; Dr. Samuel Breit (Centre for Immunology, Darlinghurst) for the CCD34Lu cell line; Lilly Research (Indianapolis, IN) for LY311727; Pei Wen Lei and Chitra de Silva for excellent assistance; and Drs. Siiri Iismaa, Robert Graham, and Charles Mackay for critical review.

1

This work has been supported by National Health and Medical Research Council Grant 980263 and by grants from the Woodend Foundation, Rebecca Cooper Foundation, and Arthritis Foundation of Australia.

4

Abbreviations used in this paper: RA, rheumatoid arthritis; MAPK, mitogen-activated protein kinase; RSF, rheumatoid synovial fibroblast; cPLA2, cytosolic phospholipase A2; COX, cyclooxygenase; NSAID, nonsteroidal anti-inflammatory drug; sPLA2-IIA, human type IIA secretory phospholipase A2; NLF, neonatal lung fibroblast.

1
Feldmann, M., F. M. Brennan, R. N. Maini.
1996
. Role of cytokines in rheumatoid arthritis.
Annu. Rev. Immunol.
14
:
397
2
van den Berg, W. B..
1998
. Joint inflammation and cartilage destruction may occur uncoupled.
Springer Semin. Immunopathol.
20
:
149
3
Keffer, J., L. Probert, H. Cazlaris, S. Georgopoulos, E. Kazlaris, D. Kioussis, G. Kollias.
1991
. Transgenic mice expressing human tumour necrosis factor: a predictive genetic model of arthritis.
EMBO J.
10
:
4025
4
Hsu, H., H.-B. Shu, M.-G. Pan, D. V. Goeddel.
1996
. TRADD-TRAF2 and TRADD-FADD interactions define two distinct TNF receptor 1 signal transduction pathways.
Cell
84
:
299
5
Geng, Y., J. Valbracht, M. Lotz.
1996
. Selective activation of the mitogen-activated protein kinase subgroups c-Jun NH2 terminal kinase and p38 by IL-1 and TNF in human articular chondrocytes.
J. Clin. Invest.
98
:
2425
6
Miagkov, A. V., D. V. Kovalenko, C. E. Brown, J. R. Didsbury, J. P. Cogswell, S. A. Stimpson, A. S. Baldwin, S. S. Makarov.
1998
. NF-κB activation provides the potential link between inflammation and hyperplasia in the arthritic joint.
Proc. Natl. Acad. Sci. USA
95
:
13859
7
Shiozawa, S., K. Shimizu, K. Tanaka, K. Hino.
1997
. Studies on the contribution of c-fos/AP-1 to arthritic joint destruction.
J. Clin. Invest.
99
:
1210
8
Badger, A. M., D. E. Griswold, R. Kapadia, S. Blake, B. A. Swift, S. J. Hoffman, G. B. Stroup, E. Webb, D. J. Rieman, M. Gowen, et al
2000
. Disease-modifying activity of SB 242235, a selective inhibitor of p38 mitogen-activated protein kinase, in rat adjuvant-induced arthritis.
Arthritis Rheum.
43
:
175
9
Portanova, J. P., Y. Zhang, G. D. Anderson, S. D. Hauser, J. L. Masferrer, K. Seibert, S. A. Gregory, P. C. Isakson.
1996
. Selective neutralisation of prostaglandin E2 blocks inflammation, hyperalgesia, and interleukin-6 production in vivo.
J. Exp. Med.
184
:
883
10
Lader, C. S., A. M. Flanagan.
1998
. Prostaglandin E2, interleukin-1α, and tumor necrosis factor-α increase human osteoclast formation and bone resorption in vitro.
Endocrinology
139
:
3157
11
Hulkower, K. I., S. J. Wertheimer, W. Levin, J. W. Coffey, C. M. Anderson, T. Chen, D. L. DeWitt, R. M. Crowl, W. C. Hope, D. W. Morgan.
1994
. Interleukin-1β induces cytosolic phospholipase A2 and prostaglandin H synthase in rheumatoid synovial fibroblasts: evidence for their roles in the production of prostaglandin E2.
Arthritis Rheum.
37
:
653
12
Roshak, A. K., J. R. Jackson, K. McGough, M. Chabot-Fletcher, E. Mochan, L. A. Marshall.
1996
. Manipulation of distinct NFκB proteins alters interleukin-1β-induced human rheumatoid synovial fibroblast prostaglandin E2 formation.
J. Biol. Chem.
271
:
31496
13
Scott, K. F., K. J. Bryant, M. J. Bidgood.
1999
. Functional coupling and differential regulation of phospholipase A2-cyclooxygenase pathways in inflammation.
J. Leukocyte Biol.
66
:
535
14
Siegle, I., T. Klein, J. T. Backman, J. G. Saal, R. M. Nusing, P. Fritz.
1998
. Expression of cyclooxygenase 1 and cyclooxygenase 2 in human synovial tissue: differential elevation of cyclooxygenase 2 in inflammatory joint diseases.
Arthritis Rheum.
41
:
122
15
Anderson, G. D., S. D. Hauser, K. L. McGarity, M. E. Bremer, P. C. Isakson, S. A. Gregory.
1996
. Selective inhibition of cyclooxygenase (COX)-2 reverses inflammation and expression of COX-2 and interleukin-6 in rat adjuvant arthritis.
J. Clin. Invest.
97
:
2672
16
Hawkey, C. J..
1999
. COX-2 inhibitors.
Lancet
353
:
307
17
Jamal, O. S., P. G. Conaghan, A. M. Cunningham, P. M. Brooks, V. F. Munro, K. F. Scott.
1998
. Increased expression of human type IIa secretory phospholipase A2 antigen in arthritic synovium.
Ann. Rheum. Dis.
57
:
550
18
Seilhamer, J. J., W. Pruzanski, P. Vadas, S. Plant, J. A. Miller, J. Kloss, L. K. Johnson.
1989
. Cloning and recombinant expression of phospholipase A2 present in rheumatoid arthritic synovial fluid.
J. Biol. Chem.
264
:
5335
19
Minami, T., H. Tojo, Y. Shinomura, Y. Matsuzawa, M. Okamoto.
1994
. Increased group II phospholipase A2 in colonic mucosa of patients with Crohn’s disease and ulcerative colitis.
Gut
35
:
1593
20
Bowton, D. L., M. C. Seeds, M. B. Fasano, B. Goldsmith, D. A. Bass.
1997
. Phospholipase A2 and arachidonate increase in bronchoalveolar lavage fluid after inhaled antigen challenge in asthmatics.
Am. J. Resp. Crit. Care Med.
155
:
421
21
Green, J.-A., G. M. Smith, R. Buchta, R. Lee, K. Y. Ho, I. A. Rajkovic, K. F. Scott.
1991
. Circulating phospholipase A2 activity associated with sepsis and septic shock is indistinguishable from that associated with rheumatoid arthritis.
Inflammation
15
:
355
22
Smith, G. M., R. L. Ward, L. McGuigan, I. A. Rajkovic, K. F. Scott.
1992
. Measurement of human phospholipase A2 in arthritis plasma using a newly developed sandwich ELISA.
Br. J. Rheumatol.
31
:
175
23
Lin, M. K. S., V. Farewell, P. Vadas, A. A. S. Bookman, E. C. Keystone, W. Pruzanski.
1996
. Secretory phospholipase A2 as an index of disease activity in rheumatoid arthritis: prospective double-blind study of 212 patients.
J. Rheumatol.
23
:
1162
24
Bomalaski, J. S., P. Lawton, J. L. Browning.
1991
. Human extracellular recombinant phospholipase A2 induces an inflammatory response in rabbit joints.
J. Immunol.
146
:
3904
25
Grass, D. S., R. H. Felkner, M. Y. Chiang, R. E. Wallace, T. J. Nevalainen, C. F. Bennett, M. E. Swanson.
1996
. Expression of human group II PLA2 in transgenic mice results in epidermal hyperplasia in the absence of inflammatory infiltrate.
J. Clin. Invest.
97
:
2233
26
Pruzanski, W., K. Scott, G. Smith, I. Rajokovic, E. Stefanski, P. Vadas.
1992
. Enzymatic activity and immunoreactivity of extracellular phospholipase A2 in inflammatory synovial fluids.
Inflammation
16
:
451
27
Tseng, A., A. S. Inglis, K. F. Scott.
1996
. Native peptide inhibition: specific inhibition of type II phospholipases A2 by synthetic peptides derived from the primary sequence.
J. Biol. Chem.
271
:
23992
28
Schevitz, R. W., D. G. Bach, N. Y. Carlson, D. K. Chirgadze, R. D. Clawson, S. E. Dillard, L. W. Draheim, N. D. Hartley, E. D. Jones, J. L. Mihelich, et al
1995
. Structure-based design of the first potent and selective inhibitor of human non-pancreatic secretory phospholipase A2.
Nat. Struct. Biol.
2
:
458
29
Futaki, N., M. Takahashi, M. Yokoyama, I. Arai, S. Higuchi, S. Otomo.
1994
. NS-398, a new anti-inflammatory agent, selectively inhibits prostaglandin G/H synthase/cyclooxygenase (COX-2) activity in vitro.
Prostaglandins
47
:
55
30
Tibes, U., W. V. Scheuer, H.-J. Thierse, E. Burgermeister, S. Schramm, W.-G. Fribe, and E. Dietz. 1997. Role of cytosolic PLA2, secretory PLA2 and nitric oxide synthase in inflammation. In Phospholipase A2: Basic and Clinical Aspects in Inflammatory Diseases, Vol. 24. W. Uhl, T. J. Nevalainen, and M. W. Buchler, eds. Karger, Basel, p. 153.
31
Garcia-Pastor, P., A. Randazzo, L. Gomez-Paloma, M. J. Alcaraz, M. Paya.
1999
. Effects of petrosaspongiolide M, a novel phospholipase A2 inhibitor on acute and chronic inflammation.
J. Pharmacol. Exp. Ther.
289
:
166
32
Valentin, E., F. Ghomashchi, M. H. Gelb, M. Lazdunski, G. Lambeau.
2000
. Novel human secreted phospholipase A2 with homology to the group III bee venom enzyme.
J. Biol. Chem.
275
:
7492
33
Suzuki, N., J. Ishizaki, Y. Yokota, K.-I. Higashino, T. Ono, M. Ikeda, N. Fujii, K. Kawamoto, K. Hanasaki.
2000
. Structures, enzymatic properties, and expression of novel human and mouse secretory phospholipase A2s.
J. Biol. Chem.
275
:
5785
34
Valentin, E., R. S. Koduri, J.-C. Scimeca, G. Carle, M. H. Gelb, M. Lazdunski, G. Lambeau.
1999
. Cloning and recombinant expression of a novel mouse-secreted phospholipase A2.
J. Biol. Chem.
274
:
19152
35
Sawada, H., M. Murakami, A. Enomoto, S. Shimbara, I. Kudo.
1999
. Regulation of type V phospholipase A2 expression and function by proinflammatory stimuli.
Eur. J. Biochem.
263
:
826
36
Valentin, E., F. Ghomashchi, M. H. Gelb, M. Lazdunski, G. Lambeau.
1999
. On the diversity of secreted phospholipases A2: cloning, tissue distribution and functional expression of two novel mouse group II enzymes.
J. Biol. Chem.
274
:
31195
37
Jakobsson, P.-J., S. Thoren, R. Morgenstern, B. Samuelsson.
1999
. Identification of human prostaglandin E synthase: a microsomal, glutathione-dependent, inducible enzyme, constituting a potential novel drug target.
Proc. Natl. Acad. Sci. USA
96
:
7220
38
Trifan, O. C., R. M. Smith, B. D. Thompson, T. Hla.
1999
. Overexpression of cyclooxygenase-2 induces cell cycle arrest: evidence for a prostaglandin-independent mechanism.
J. Biol. Chem.
274
:
34141
39
Chapdelaine, J. M., V. B. Ciofalo, D. S. Grass, R. Felkner, R. E. Wallace, M. E. Swanson.
1996
. Human extracellular type II phospholipase A2 (PLA2) transgenic mice provide a tool to determine the role of PLA2 in inflammatory conditions.
Arthritis Rheum.
38
:
S293
40
Murakami, M., K. Tada, K. Nakajima, I. Kudo.
1997
. Cyclooxygenase-2-dependent delayed prostaglandin D2 generation is initiated by nerve growth factor in rat peritoneal mast cells: its augmentation by extracellular type II secretory phospholipase A2.
J. Immunol.
159
:
439
41
Murakami, M., H. Kuwata, Y. Amakasu, S. Shimbara, Y. Nakatani, G. Atsumi, I. Kudo.
1997
. Prostaglandin E2 amplifies cytosolic phospholipase A2- and cyclooxygenase-2-dependent delayed prostaglandin E2 generation in mouse osteoblastic cells: enhancement by secretory phospholipase A2.
J. Biol. Chem.
272
:
19891
42
Tada, K., M. Murakami, T. Kambe, I. Kudo.
1998
. Induction of cyclooxygenase-2 by secretory phospholipase A2 in nerve growth factor-stimulated rat serosal mast cells is facilitated by interaction with fibroblasts and mediated by a mechanism independent of their enzymatic functions.
J. Immunol.
161
:
5008
43
Dean, J. L. E., M. Brook, A. R. Clark, J. Saklatvala.
1999
. p38 mitogen-activated protein kinase regulates cyclooxygenase-2 mRNA stability and transcription in lipopolysaccharide-treated human monocytes.
J. Biol. Chem.
274
:
264
44
Cornelis, F., S. Faure, M. Martinez, J.-F. Prud’homme, P. Fritz, C. Dib, H. Alves, P. Barrera, N. de Vries, A. Balsa, et al
1998
. New susceptibility locus for rheumatoid arthritis suggested by a genome-wide linkage study.
Proc. Natl. Acad. Sci. USA
95
:
10746
45
Becker, K. G., R. M. Simon, J. E. Bailey-Wilson, B. Freidlin, W. E. Biddison, H. F. McFarland, J. M. Trent.
1998
. Clustering of non-major histocompatibility complex susceptibility candidate loci in human autoimmune diseases.
Proc. Natl. Acad. Sci. USA
95
:
9979
46
Tischfield, J. A., Y.-R. Xia, D. M. Shih, I. Klisak, J. Chen, S. J. Engle, A. N. Siakotos, M. V. Winstead, J. J. Seilhamer, V. Allamand, et al
1996
. Low-molecular-weight, calcium-dependent phospholipase A2 genes are linked and map to homologous chromosome regions in mouse and human.
Genomics
32
:
328