Changes in Eicosanoid and Tumour Necrosis Factor-α Production by Rat Peritoneal Macrophages During Carrageenin-Induced Peritonitis

Changes and correlations in cytokine and eicosanoid production by blood monocytes, non-purified and purified peritoneal cells during a carrageenin-induced peritonitis were investigated for a period of ten days. The cells were isolated and stimulated in vitro. Cytokine and eicosanoid production of the non-purified fraction increased steadily during peritonitis. During the whole episode of peritonitis the production capacity of granulocytes was very low and hardly any effect on the production capacity of macrophages (Mϕ) was observed. Cytokine and eicosanoid production of the non-purified fraction was mainly due to the presence of Mϕ. The production capacity of the peripheral blood monocytes was not similar to that of the peritoneal Mϕ.


Introduction
Inflammation is the reaction to an injury such as an invasion by infectious agents. The blood supply to the inflamed area increases, venular permeability increases and leucocytes migrate out of venules into the surrounding tissue. At the site of an acute inflammation polymorphonuclear granulocytes (PMNs) produce large amounts of reactive oxygen intermediates, release a variety of hydrolytic enzymes and phagocytose pathogens. By generating chemotactic factors, the migration of mononuclear cells (monocytes/M and lymphocytes) to the site of inflammation is stimulated. ',3 During the chronic phase of the inflammatory response, M, mononuclear phagocytes derived from blood monocytes, predominate over PMNs. In addition to acting as a nonspecific defence, activated M 5,6 will initiate and control the specific defence by presenting processed antigens to lymphocytes and by cytotoxity through direct cell-cell contact with tumour cells or infected cells, 8,9 during which they secrete cytokines and eicosanoids. 1,11 These inflammatory mediators also influence and regulate functions of other cells participating in the immune response, l'a Cytokines are proteins with regulatory functions. They can induce other cytokines with overlapping effects and form interactive networks with hormones and eicosanoids. M are a potent source of a wide variety of cytokines, 4 whereas granulocytes are reported to produce only a limited spectrum of cytokines. [15][16][17] ( 1994 Rapid Communications of Oxford Lid Eicosanoids are bioactive lipids derived from arachidonic acid (AA) by cyclooxygenase and lipoxygenase pathways. 8a9 Eicosa.noid production is species and tissue dependent, and often stimulus dependent. 2,21 The aim of this work was to study cytokine (tumour necrosis factor-m, TNF) and eicosanoid (leukotriene B4, LTB4; prostaglandin E2, PGE2; prostacycline, detected as 6-keto-prostaglandin Fire 6kPGF=; and thromboxane, TXB,) production capacity of peritoneal PMNs, M and blood monocytes after the induction of peritonitis with a carrageenin solution, a sulfated polygalactan which stimulates cell-mediated immunity. 22 It was of interest to determine whether, during an episode of peritonitis, changes occurred in (a) the differentiation of influxed cells; (b) the cytokine and eicosanoid production capacity of blood monocytes and peritoneal cells; (c) interactions between different types of peritoneal cells concerning their production capacity; and (d) if there was a correlation between the production of eicosanoids and TNF by blood monocytes and peritoneal M.

Materials and Methods
Animals and treatment: Young male Wistar rats (13 weeks, approximately 200 g, six rats per group) were injected intraperitoneally with 2 ml of a carrageenin solution (Marine Colloids Inc., USA, 1 mg/ml) on day 0. Six rats received an injection of saline (control group).
The peritoneal cavities, were washed twice with 20 ml PBS (pH 7.4, 4C, Oxoid, UK). The washings per rat were pooled, centrifuged (400 xg, 4C, 10 min) and suspended in 5 ml DMEM. 5 x 106 cells of this non-purified fraction were kept separate, the rest of this 'crude' fraction was separated on Percoll into a 'macrophage' and a 'granulocyte' fraction. These two purified fractions were washed three times with PBS (400 xg, 4C, 10 min), and suspended in DMEM (i x 106 cells/ml). A small sample of cells from each fraction was stained by Hemacolor (Merck, Germany) and the different cell types were counted using a microscope (Zeiss, standard 25, Germany). The viability of the cells was determined by trypan blue exclusion.
The cells were also incubated for 24 h in the absence or presence of lipopolysaccharide (LPS, 10 l.tg/ml final concentration in PBS, LPS from E. coli 0111:B4 in PBS (Sigma, USA), 37C, 7.5% COg). As a blank, PBS was added. At the end of the incubation the supernatant was centrifuged and kept at -80C until required for analysis.

Results
Concentration and differentiation of cells obtained from theperitoneal cavity. Induction of peritonitis in rats caused an increase of cells in the peritoneal cavity on day 1. The cell count dropped on the following days (Fig. 1A). A pronounced influx of granulocytes was seen on the first day which was followed by an influx of Me, resulting in a Me/ granulocyte ratio of 4:1 from day 3 to 10 ( Fig. 1B).
Cytokine and eicosanoid production by the crude fraction increased steadily during peritonitis. TNFc and TXB2 reached peak levels after 3 days, whereas PGE2 and 6kPGFI= reached a plateau 7 days after induction of the peritonitis. In comparison with levels of other eicosanoids the LTB4 production was low before and on the first day of the peritonitis and decreased significantly thereafter. By day 10 the production capacity returned to the initial level.
The production pattern of the cells from the macrophage fraction (Fig. 2B) was similar to that of cells from the crude fraction ( Fig. 2A). Inflammatory mediators were also produced by the granulocyte fraction (Fig. 2C), although the production capacity of these cells was very low in comparison with cells from the crude and macrophage fraction. The pattern of inflammatory mediator production of the granulocyte fraction did not significantly change in time. On the tenth day after induction of the peritonitis, the concentration of the cells in the granulocyte fraction was too low to permit in vitro incubations. The capacity of cells from the peripheral blood monocyte fraction to produce inflammatory mediators (Fig. 2D) was at least four times lower in comparison with both the crude and macrophage fractions. The production capacity of the cells from the monocyte fraction was about the same with or without stimulus. The patterns of the TNFz, PGE and 6kPGFI= production of cells from the monocyte fraction were different from those of the macrophage and crude fraction. After induction of peritonitis the TNF0t production of cells from the monocyte fraction increased significantly only on the tenth day, whereas PGE2 and TXB,. reached their highest level on the third day. The 6kPGFlu level did not change during peritonitis. The LTB4 production was negligible.
Correlations between inflammatory mediators produced in the same fraction: In the crude fraction ( Fig.   2A) there was a significant positive correlation between the production of PGE,. and 6kPGFI= and PGE2 and TXB,. (r= 0.6365 and r= 0.7145). Similar significant correlations were found in the macrophage fraction (Fig. 2B, r= 0.5580 and r= 0.6106). In the monocyte fraction (Fig. 2D) the only significant positive correlation between the inflammatory mediators was between PGE and WXB (r= 0.8881).
Correlation between inflammatory mediators in different fractions: In Fig. 2A and 2B the patterns of inflammatory mediators produced by the stimulated crude and macrophage cell fractions are presented. There was always a significant positive correlation between these two fractions. The correlation varied from r= 0.5671 for LTB to r 0.8586 for TNFo:. (PGE,., r= 0.6254; 6kPGFI=, r= 0.7536; and TXB2, r= 0.7502). This correlation remained significant when we assumed that production of the cell inflammation mediators in the crude and macrophage fraction were only derived from the Me (calculated to 100% macrophages, LTB4=0.4649; PGE2 0.5855; 6kPGFI= 0.5506; TXB2 0.7810; and TNFo: 0.8674).
There were, however, some significant differences between the calculated (to 100% Me) production levels of the crude and macrophage fractions. When separate days were considered the production levels of the calculated crude fraction in comparison with the calculated macrophage fraction was on day 1 lower for PGE2 and 6kPGF=, and on day 3 higher for gWB4, TXB2 and TNFo: (Fig. 3 A-E).
Correlation coefficient between inflammatory mediators from monocyte and macrophage fraction: Only the basal TNFz production of the cells from the monocyte and macrophage fraction were clearly correlated (r= 0.5421).

Discussion
Characterization of inflammatory mediators produced by granulocytes and monocytes/macrophages during induced inflammation in animal models, may be helpful to the understanding of inflammatory diseases and their treatment. The present study investigated the changes and correlations in cytokine and eicosanoid production capacity of peripheral blood monocytes, non-purified ('crude' fraction) and purified ('macrophage' and 'granulocyte' fraction) peritoneal cells during a carrageenin-induced peritonitis in rats.
In the present study, where an influx of leucocytes into the peritoneum was achieved after carrageenin injection, the number of PMNs increased in the acute phase (day 1) and thereafter decreased quickly. On day 0, before inducing the peritonitis, there was a reasonable number of Me present in the peritoneal cavity. The number of Me increased from day 1 until 7 days after induction of the peritonitis, resulting in a mainly Me cell population. Previous histological studies had similar results, '4,25 although a quantitative and kinetic method showed that both monocytes and PMNs from rat migrate rapidly to the inflamed site, the migration of PMNs, however, also declined long before the migration of monocytes started to decrease. The pattern of this migration can be explained with the function of the leucocytes in inflammation. In the initial phase both PMNs and Me are involved in nonspecific defence mechanisms. ,2 In addition, Me will also initiate and control the specific defence.  The mechanism of this selective migration is not clear. Leucocyte adhesion to the blood vessel endothelium, which is followed by transendothelial migration, is a multistep process in which several adhesion molecules are involved, z6,z7 Our experiment showed that the inflammatory mediator production capacity of the non-purified fraction, which was harvested from the inflamed peritoneal cavity, changed with time. The production of inflammatory mediators in the non-purified fraction was mainly caused by the presence of M) during carrageenin-induced peritonitis. This is, first of all, based on the significant correlation between the eicosanoid and TNFot production capacity of the macrophage fraction and the crude fraction, these correlations hardly changed when it was assumed that these fractions only contained M. Secondly, the production capacity of the granulocytes appeared to Changes in eicosanoid and TNF production be very low. Moreover the production capacity of the granulocyte fraction did not correlate with the crude fraction.
When the production levels of the non-purified fraction was compared with the macrophage fraction, assuming that the inflammatory mediators were only produced by M, it was observed that the production of PGE,. and 6kPGFI= had decreased on day I and the production of LTB4, TXB and TNF had increased on day 3. The results of these effects could be an amplification of the inflammation, due to the decrease of anti-inflammatory substances and the subsequent increase of pro-inflammatory substances.
The low production capacity of the crude fraction in the acute phase followed by an increasing production capacity in the chronic phase could also be explained by presence of the M. Blood monocytes and resident M transform from primed into active M at the side of the inflammation, after which profound changes occur in morphology and function of these cells. The M present at day 1 in the peritoneal cavity, probably were not primed in vivo at this time, which could be the reason that the eicosanoid and cytokine production capacity of these M were so low after stimulation in vitro with LPS. Once these M were primed in vivo, also the new monocytes that migrate into the peritoneal cavity, the production capacity of these monocytes/M steadily increased, showing in vitro a differentiated pattern for the eicosanoids and TNF. This pattern probably depends on the function of the metabolite.
As peripheral blood monocytes transform into peritoneal m, the correlation between these cells was investigated. The eicosanoid and TNF production capacity of blood monocytes was not similar to that of M during peritonitis. It was clearly shown that through influx and transformation of monocytes into M the synthesis of mediators completely changed.
In conclusion, M are the main source of the inflammatory mediators whose production is dependent on the episode of the carrageenin-induced peritonitis.Characterization of purified M in vitro in animal and human models will be helpful to the understanding of inflammatory diseases and their treatment. Peripheral blood monocytes do not reflect mediator production of M present at the inflammatory site.