Sequential release of TNFα and phospholipase A2 in a rat model of LPS-induced pleurisy

The levels of extracellular phospholipase A2 (sPLA2) and TNFα, and cell accumulation were measured in the pleural washings obtained at different times following the induction of Escherichia coli lipopolysaccharide (LPS, 100 μg/cavity) pleurisy in rats. TNFα peaked at 2 hours (3036 ± 160.3 units/ml) and decreased thereafter. Conversely, levels of sPLA2 peaked at 48 hours (1.97 ± 0.64 ng/ml) and were increased further (14.02 ± 4.16 ng/ml) by pretreatment with anti-TNFα antibody. Cell accumulation was not affected by antibody pretreatment. These data indicate that the sPLA2 enzyme is involved in LPS-induced pleurisy. The enzyme seems not to be stimulated by TNFα which may be involved in the downregulation of sPLA2 in this model of inflammation.


Introduction
The extracellular form of phospholipase A 2 has been shown to be involved in several in ammatory disorders, including those arising from Gram negative bacterial infections. 1,2 Phospholipase A 2 (PLA 2 ) activation is a rate limiting step in the generation of arachidonic acid, the substrate necessary for eicosanoid generation. The secreted form of PLA 2 (sPLA 2 group II) has been found at high levels in the serum and locally during in ammation in experimental animals and humans, 3 and there is in vitro evidence that its induction is under cytokine control. [4][5][6][7] Furthermore, sPLA 2 has been proposed to play a role in septic shock. To date, however, the relationship between cytokines and sPLA 2 levels in in ammatory exudates and in the serum is still controversial. 2,8 It has been shown that intrapleural injection of LPS into rats elicits an in ammatory response characterized by leucocyte accumulation. 9 Furthermore, challenge of isolated guinea pig lung with endotoxin generates TNFa 11 and the presence of TNFa in the bronchoalveolar lavage uid of experimental animals has been demonstrated following intratracheal administration of LPS. 12 It appears that tumour necrosis factor alpha (TNFa) is a cytokine playing a pivotal role in the development of endotoxin-induced in ammation. 10 We have examined whether intrathoracic injection of bacterial endotoxin into rats is capable of causing the release of sPLA 2 and TNFa, and the potential relationship between this enzyme and this cytokine in this model of in ammation.

Pleurisy
Pleurisy was induced in en urane anaesthetized male Wistar rats (250 ±300 g; Charles River, Milan, Italy) by intrathoracic injection of E. coli lipopolysaccharide (LPS, 0127:B8; 100 mg/ cavity) in a nal volume of 200 ml. Control animals received an equivalent volume of phosphate buffered saline (PBS). Different groups of animals were treated with a polyclonal anti-TNF a antibody (3 mg/ kg i.v.; TAB, London) or control IgG (3 mg/ kg i.v.) immediately before the induction of pleurisy. Animals were killed with CO 2 at 2, 6, 24, 48, or 96 hours following the injection of LPS or PBS. Groups of animals pretreated with anti-TNFa antibody, or control IgG, were killed after 2 and 48 hours, the time required to obtain peak TNFa and sPLA 2 levels respectively. Immediately after sacri ce the thoracic cavity was opened and washed with 2 ml of heparinized (5 UI/ ml) sterile saline. The uid volume was collected in graduated tubes and centrifuged at 125 3 g for 10 ±15 min. The pellet was suspended in 1 ml of PBS and total leucocyte count was evaluated. Supernatant was further centrifuged at 650 3 g for 10 ± 15 min, immediately frozen and kept at -208 C until assayed for TNFa and sPLA 2 .

Leucocyte count
Total leucocyte count was evaluated by optical microscopy in the cell suspension diluted with Turk's solution. Differential cell count was performed by counting air dried smears stained with Giemsa reagent under an oil immersion objective. Counts are reported as number of cells 3 10 6 / cavity.

TNFa determination
TNFa in the cell-free pleural washing was assessed on WEHI-164 cells by a bioassay using recombinant human TNFa (Sigma) as the reference standard, and rabbit anti-murine TNFa antiserum (Genzyme) which cross reacts with rat TNFa to assess the speci city of TNFa dependent cytolitic activity. 13 TNFa is expressed as units of TNFa/ ml, where one unit is de ned as the quantity required to lyse 50% of WEHI-164 cells.

PLA 2 assay
A sandwich assay for sPLA 2 was run according to the method described by Santos et al. 8 with some modi cations. Plates were coated overnight with a monoclonal anti-human recombinant sPLA 2 antibody (BA11), 10 mg/ ml, in 50 mM sodium bicarbonate buffer, pH 9.6. Six cycles of washing with 0.1% w / v Tween-20 in PBS were done at each step and all incubations were at 48 C. The plates were blocked overnight with 10 mg/ ml of gelatin in PBS. Human recombinant sPLA 2 was diluted in 50 nM Tris-HCl, pH 7.5, 0.1% w / v Tween-20, 1% v/ v foetal calf serum (FCS), to obtain the desired dilution, and 0.05 ml were added to the wells at 48 C for 1 hour (low temperature decreases non-speci c PLA 2 binding). Bound human recombinant sPLA 2 was detected by incubation for 90 min with 0.05 ml of a 1:2000 dilution of rabbit anti-human recombinant sPLA 2 (n.207) in assay buffer. Plates were washed and 0.1 ml of a solution containing 1 mg/ ml of tetramethylbenzidine in DMSO, mixed with 10 ml of acetate buffer (0.1 M, pH 5.0) and 0.125 ml of diluted hydrogen peroxide (0.2 ml 30% hydrogen peroxide and 7 ml of water), were added to each well. After 15 ± 20 min, 0.025 ml of stop solution (H 2 SO 4 , 2 M) was added and the yellow colour read at 450 nm by an ELISA reader. sPLA 2 in the cell-free pleural washing was expressed as ng/ ml.

Histological analysis
Lungs from control or LPS-injected animals were removed and xed in neutral buffered formalin before being embedded in paraf n. Sections (4 mm) were stained with hematoxylin and eosin.

Statistical analysis
Data are expressed as mean SEM and analysed with a computerized package for statistical comparison. 14 To compare groups of animals, unpaired two tailed Student's t test was used. Data on TNFa were analysed by the Mann ±Whitney test. A P , 0·05 was taken as signi cant.

Materials
Recombinant human sPLA 2 produced in mammalian cells, monoclonal anti-human recombinant sPLA 2 antibody (BA11) and rabbit polyclonal anti-human recombinant sPLA 2 antibody (n. 217) were a generous gift from Dr J. L. Browning (Biogen). En urane was purchased from Abbott (Italy). Lipopolysaccharide from E. coli (serotype 0127:B8), PBS and heparin were purchased from Sigma (Italy). Antibody to human TNFa was a generous gift from TAB London, and has been shown to neutralize rat TNFa in vitro (TAB London, unpublished data).

Leucocytes
Following intrathoracic injection of LPS, total leucocyte count was signi cantly increased only at 48 and 96 hours, while at 6 hours there was a signi cant reduction in the leucocyte count compared to control value. Polymorphonuclear count was signi cantly increased at 24 and 48 hours, signi cantly reduced at 6 hours and returned to control levels after 96 hours. Monocyte count was signi cantly increased at 96 hours, and signi cantly reduced below control levels at 6 and 24 hours (Fig. 1).

TNFa determination
TNFa in the cell-free pleural washing peaked at 2 hours following LPS injection. At the successive time points, TNFa levels were reduced progressively, becoming undetectable at 96 hours (Fig. 2).
Phospholipase A 2 levels in the cell-free pleural washing were detectable at 2 hours, reached their maximum at 48 hours and were stable up to 96 hours (Fig. 2).

Histological analysis
In the histological sections of lungs obtained from either PBS or LPS injected animals there was neither leukocyte in ltration, nor alteration in the cell lining of tissue.

Discussion
By using a highly speci c immunoenzymatic assay, we have demonstrated the presence of extracellular sPLA 2 group II in the cell-free pleural washing at different time points following LPS injection, providing direct evidence for a role of this enzyme in the pleurisy triggered by LPS. In our model, the appearance in the cell-free pleural washing of increased levels of sPLA 2 was delayed compared to TNFa levels. Indeed, sPLA 2 increased in the cell-free pleural washing only when TNFa levels started to decline, and it reached its maximum at 48 hours, when TNFa levels were very low. At the same time, cells signi cantly accumulated: polymorphonuclear leucocytes reached their maximum at 48 hours, and at 96 hours the pleural washing consisted almost totally of monocytes. When animals were pretreated with anti-TNFa antibody, TNFa levels were signi cantly reduced. However, cell accumulation did not parallel the increase in TNFa and pretreatment with anti-TNFa antibody did not affect total and differential cell count, suggesting that resident pulmonary cells, stimulated by LPS, may account for TNFa production and that TNFa is not involved in LPS-induced leucocyte margination in the pleural space. These data are consistent with data reported by others, who demonstrated that IL-1 and not TNFa is mostly responsible for LPS-induced neutrophil migration to an in ammatory site. 15 Moreover, it is known that adhesion molecules contribute greatly to LPS-induced leucocyte margination in the lung and their expression is under the control of multiple cytokines. 12 Other authors report that the ability of TNFa inhibitors to protect animals from ogogen-induced lung injury is not linked to a reduction in PMN sequestration. 16 In the light of this knowledge, it is not unexpected that the inhibition of TNFa alone did not affect leucocyte accumulation. Surprisingly, when we pretreated animals with the anti-TNFa antibody, we observed a strong increase in sPLA 2 levels in the cell-free pleural washing at 48 hours. Thus, it seems that TNFa does not exercise a positive control on sPLA 2 induction in vivo, rather its inhibition causes sPLA 2 levels to increase further, at least in this experimental model of pleurisy. Systemic administration of the antibody allows us to rule out the hypothesis of any interference of the antibody with ELISA assay reagents. Although there is much evidence that in vitro cytokines are the major inducers of sPLA 2 synthesis and secretion from several different cell types, there is no direct evidence that in vivo sPLA 2 secretion is under cytokine control. A positive correlation has been found between TNFa and sPLA 2 plasma levels both in human and in experimental animals challenged with bacterial endotoxin, as well as an increase in plasma sPLA 2 levels following TNFa injection in volunteers, 1,2 but this evidence has never been followed by clari cation of the in vivo role of the enzyme. In 1993 synthesis and secretion of a 14 kDa sPLA 2 from guinea pig alveolar macrophages was demonstrated and it was found that the enzyme was not involved in the release of arachidonic acid metabolites. 17 More recently the same research group has shown that expression of sPLA 2 in guinea pig alveolar macrophages is downregulated by an in ammatory stimulus such as FMLP, 18 and it has been found that the release of arachidonic acid metabolites from guinea pig alveolar macrophages in vivo is reduced below control levels 24 hours after pretreatment with LPS. 19 Up to now, the role of sPLA 2 in vivo has been unresolved; following these more recent ndings, it has been suggested that the secreted form of PLA 2 might have a protective role at site of injury. 17 -21 It is noteworthy that 6 hours following LPS administration there was a signi cant reduction in leucocyte number in the pleural space, paralleled by a reduction in both granulocyte and monocyte counts. Although there was neutrophil accumulation in the lung at 24 hours, the total leucocyte count did not increase since the monocyte count was still reduced below the control level. This reduction could be explained by an increased cell adhesivity which precedes cell accumulation. Indeed, recent data have demonstrated that injection of increasing doses of IL-1 or TNFa in to the pleural space of rats causes a bell-shaped cell accumulation curve. This nding has been shown to be linked to the increased cell adhesivity, evoked by high doses of IL-1 and TNFa. 22 In conclusion, our data demonstrate that following intrapleural injection of endotoxin into rats, TNFa and sPLA 2 group II appear sequentially in the pleural washings. The early appearance of TNFa is in agreement with data reported by others who describe TNFa as an early response cytokine. The high levels of sPLA 2 in the late phase of the in ammation further suggest a role for the extracellular form of the enzyme in in ammation triggered by bacterial endotoxins. TNFa is probably produced by resident pulmonary cells and does not seem to be directly involved in cell accumulation. TNFa does not stimulate the enzyme induction, rather there is evidence that its inhibition increases sPLA 2 levels, suggesting that this cytokine could participate in a mechanism of downregulation of the enzyme during in ammation; however, this hypothesis requires further clari cation.