Negative correlation between maorophage proooagulant and migraion ability in the course of inflammation

The importance of macrophage procoagulant activity (PCA) to cell migration is presumed. In this study we assayed the relationship between the two functions in guinea-pig peritoneal resident macrophages and cells elicited by a sterile inflammation induction, which lasted up to 6 days. The findings pointed to an in vivo induction of PCA in macrophages, which declined with time during inflammation. A clear negative correlation between PCA and random migration ability was demonstrated. Our results suggest that the local induction of coagulation by macrophages may immobilize the cells at the site of inflammation.


Introduction
between PCA and those cells' random migration.
Our results point to PCA induction at the inflam-Among the rich macrophage secretory repermatory site and its negative correlation with mactoire, their procoagulant activity (PCA) is one of rophage migration potency. the most interesting. 1'2 Mononuclear phagocyte cells may produce various coagulation factors Materials and Methods thus expressing a different secretory phenotype.
It is also demonstrated that monocytes/macrophages possess fibrinogen receptors, as well as Animals: Outbred guinea-pigs of both sexes, fibrinogen/fibrin linked in a granular or net-like weighing 350-450 g, 5-7 per group, were used. 4 form. The in vivo significance of macrophage All experiments were performed two to three PCA is still obscure. It is presumed that PCA times to ensure that the results were consistent contributes to the pathogenesis of some auto-and the representative experiments were shown. immune and other immunological diseases, to h 5 11 graft rejection and to delayed ypersensitivity. Peritoneal macrophages: To obtain macro-Besides, mononuclear phagocyte PCA is pre-phages, peritoneal cavities were washed with sumed to be a simple reproducible correlate of Hanks's balanced salt solution (HBSS) suppledelayed hypersensitivity in man and experimental mented with 0.5 IU/ml of heparin ('Galenika', animals. 12 '13 Belgrade, Yugoslavia). Resident macrophages It is also implied that the induction of extra-were harvested by the lavage of cavities of vascular coagulation by macrophages causes untreated animals, and elicited ones at days 3 (3-induration at the inflammatory locus. 4 Thus day macrophages)and 6 (6-day macrophages) fibrin deposition on one side probably limits the after induction of a sterile peritoneal inflammainflammatory lesion, and on the other side may tion with 10 ml of paraffin oil. According to arrest inflammatory cells. Therefore, the sig-neutral red staining 16 67-90% cells were macronificance of monocyte/macrophage PCA for their phages. The remainder of the cells consisted 15 own migration ability is supposed. The premainly of lymphocytes and 1-2% neutrophils. vious findings were mainly concerned with the Cell viability was assessed by trypan blue exclulocomotion of activated macrophages in delayed sion and was greater than 90%. hypersensitivity reactions. To our knowledge, the development of macrophage PCA during inflam-Preparation of serum and plasma: Blood was mation, especially in relation to macrophage prepared from normal, untreated guinea-pigs by motility, has never been examined, cardiac puncture. Blood samples were clotted 20 In this paper we tested the development of min at room temperature, incubated overnight at PCA in guinea-pig peritoneal macrophages during 4C, centrifuged, and pooled to obtain guinea-pig sterile inflammation, as well as the relationship serum. Nine volumes of blood were mixed with L Vilig and L. M. Prokig one volume of 3.8% sodium-citrate, centrifuged inhibition (p < 0.05, p < 0.01, p < 0.001, for 18 min at 2 000 rpm to obtain plasma, and respectively). The cells were also viewed by light then 15 min at 3000 rpm (at 4C) to eliminate microscopy (x 400; Amplival microscope, Carl platelets. The plasma samples were pooled and Zeiss, Jena, Germany), and photographs were stored at -70C for up to 1 month until use. In taken with Kodak 35 mm film. Migration chamsome experiments heat-inactivated plasma was bers were opened at the end of an incubation used (1 h at 56C).
period, and the quality of medium was examined (liquid/coagulated). Procoagulant activity: The ability of macrophages to shorten the coagulation time of recal-Macrophage preincubation with thrombin: After cified guinea-pig plasma was assayed either with collection, 6-day macrophages (5 x 10 cells/ml) freshly isolated cells, or after incubation (5 x measured after the addition of 0.1 ml of 0.025 M The substance had no effect on both random CaCI. PCA is expressed in seconds and/or millimigration and PCA. units (mU) of activity per 10 cells. A standard cueee of the plasma recalcified time was made Statistics: The statistical analysis was performed with decreasing logarithmic dilutions of rabbit using the two-tail Student's t-test and Spearman's brain thromboplastin (Blood Transfusion Insti-correlation test. tute, Belgrade, Yugoslavia; 20-0.1 l.tg/0.1 ml saline), added to 0.1 ml plasma. Six mg of Results thromboplastin was nominated as 1 000 mU of PCA. The standard thromboplastin curve was compared with that obtained by using dilutions Macrophage random migration in plasma: To of macrophage suspensions. The slopes of the investigate whether oil-elicited macrophages plot of clotting time versus rabbit thromboplastin possess PCA of any importance to their random dilutions were correlated, migration ability, the migration was at first estimated in the presence of 10% guinea-pig plasma. Random migration testing: The macrophage Plasma strongly inhibited motility of 6-day macrorandom migration was assayed by the capillary phages, compared with 10% serum ( Fig. 1; p < tube method. v Resident or elicited peritoneal 0.001). The analysis of migration kinetics (the exudate cells (PEC) were washed in HBSS increase of migration areas in time)showed that without heparin. Erythrocytes were eliminated by the migration areas did not increase in the prelysis with 0.83% NH4C1. The cells were then sence of plasma, in contrast to the characteristic resuspended in M 199 (40 x 10 cells/ml), successive enlargement in the presence of serum Migration chambers, containing two capillaries (results not presented). The migration inhibitory with a cell pellet, were immediately filled with factor(s) was thermolabile, since heat-inactivated medium supplemented with 10% of guinea-pig plasma did not reduce migration, and was serum, or plasma, or heat-inactivated plasma, equivalent to serum (results not presented). After After incubation from 4 to 44 h at 37C, the cell the opening of migration chambers, at the end of migration was determined by weighing the paper an incubation period, a complete or partial coagucut-outs of the area of the projected image. The lation of culture medium supplemented with average migration areas of four replicates (in plasma was observed. In the control chambers mg) were used for calculation of the migration containing either plasma without cells or serum index (MI): MI test area/control area x 100. plus cells, coagulation of medium was not seen. It was found that MI values under 80, i.e. 60 or Light microscopy and photos of macrophages 40, represent statistically significant migration from migration areas showed that they were pre- The influence of 10% guinea-pig plasma in vitro on the random migration of oil-elicited guinea-pig peritoneal 6-day macrophages. Migration areas obtained with 10% serum are taken as controls. All experiments were performed two to three times to ensure that the results were consistent and the representative experiment is shown (n 5-7 per group). Significant inhibition, p < 0.001. dominantly rounded in the presence of serum, whereas in plasma they displayed predominantly an elongated morphology (results not shown).
The appearance of 'fibrils' in which the cellular elements appeared to be 'trapped' was also seen in migration areas from chambers with plasma.
Eect o thrombin on macrophage random migration: To further demonstrate that PCA influences macrophage random migration potency, the 6-day macrophages were preincubated for 30 min with 1-10 U/ml of thrombin. Controls were preincubated in M 199 with or without 10% serum. Pre-treatment with thrombin caused a strong dose-dependent migration reduction (Fig. 2). At the highest dose, 90-100% inhibition of cell locomotion was measured at 4 h. The inhibition persisted throughout the cultivation period, except with the lowest dose, but gradually weakened with time. PCA of peritoneal macrophages: All types of peritoneal macrophages, i.e. resident and those harvested at days 3 and 6 of sterile inflammation, had a low basal PCA immediately after the cell collection (Fig. 3). However, the PCA was greater in 3-day macrophages compared with 6-day macrophages or resident ones. This pointed at an in vivo induction of PCA during the inflammatory process. In vitro cultivation of macrophages enhanced their PCA over basal level, to its maximum at 4-8 h. The in vitro PCA stimulation was more pronounced in resident and 3-day. macrophages (319 12 and 320 +__ 30 mU/10 cells), than in 6-day macrophages (147 -+_ 9 mU/ 10 cells). Thereafter PCA returned to the initial level at 20 h, except in resident cells in which a 25% stimulation persisted at the time (p < 0.05). The estimation of PCA kinetics (the change in time when 0 h level was set to control one) showed that resident macrophages had the greatest potency to PCA induction (6versus 2.5-fold in both 3-day and 6-day macrophages; results not presented).  Random migration of peritoneal macrophages during inflammation: The random migration was tested in parallel with PCA in the three types of macrophages (Table 1). Resident cells and mononuclear phagocytes from the earlier phase of inflammation had a lower migratory ability when compared with 6-day macrophages. It was observed on both resident and inflammatory cells, as we reported for 6-day macrophages only, 19'20 that random motility increased with time. Serum exerted its known positive chemokinetic influence, as well as prolonged cell survival up to 44 h. The subjection of the findings for PCA and random migration to the correlation test demonstrated at 4 and 20 h a significant negative relationship between the PCA and random migration capacity in all the three types of macrophages (p < 0.05 to p < 0.001).

Discussion
It was reported that PCA induction in activated macrophages significantly reduced their locomotion in vitro. In that paper such an influence was not attributed to elicited macrophages, in spite of evidence that they may also secrete some coagulation factors. 2 The testing of 6-day macrophage random migration in the presence of plasma seemed to us the most suitable and direct way to start the examination of the PCAmigration relationship. The results presented here show the change in shape of migrating cells, the strong reduction of the motility, as well as the coagulation of medium in such a culture condition. Because the medium coagulated only if macrophages were present in the migration chambers, it means that they started the process with their procoagulant products. The results are in line with findings that 5-50% of plasma in culture reduced macrophage motility for 27-100%. 22 The preincubation of elicited 6-day macro-274 Mediators of Inflammation Vol 5 1996 phages with thrombin demonstrated the significance of PCA to motility more directly, as it caused the powerful dose-dependent migration inhibition. Thus, with 5-10 U/ml of the agent, a 80-100% locomotion reduction appeared at 4 h. The substrate to thrombin action was presumably fibrinogen bound to macrophage surface. Indeed, it has been demonstrated that about 37-68% of the elicited peritoneal macrophages possess membrane-bound fibrinogen/ fibrin of plasma origin. 4 To our knowledge, the inhibitory effects of thrombin have been demonstrated only on activated guinea-pig macrophage chemotaxis, being estimated in a short period of time--usual for directed motility testing, a2 Contrary to that, in our model the influence of preincubation with thrombin on undirected migration was monitored up to 44 h. The gradual decline or even disappearance of the effect was presumably the consequences of the macrophage neutral proteases action. 12'19'23 Recently, it was reported that thrombin enhances monocyte adherence to endothelial cells in vivo. 24 Although PCA induction may be assumed through some indirect data, the examinations such as ours concern both the change of PCA in macrophages during inflammation and its significance for their own migration potency, even though this is rarely present. Thus, it is supposed that the macrophage disappearing reaction in delayed hypersensitivity is mediated by PCA induction.i-It is suggested that the loss of macrophages was due to the formation of clumps and adherence to the peritoneal mesothelium lining. The cells appearing in the peritoneal cavity, after some time, is attributed to the stimulation of their own opposite fibrinolytic activity. Our study with resident and inflammatory macrophages shows that all three types of macrophages had a small basal PCA immediately after being lavaged out of the peritoneum. The findings that PCA was the highest in 3-day macrophages and the lowest in resident ones, that resident cells exerted a better potency for PCA induction, and that coagulation inducibility was best in 'younger' inflammatory cells, speaks in favour of a gradual decrease of the PCA of mononuclear phagocytes in the course of inflammation. The parallel macrophage random motility testing further supported the view that PCA is important to cell mobility, as migration ability was negatively correlated with PCA. Some previous findings have also demonstrated the functional difference between macrophage populations. Farram et al. 25 reported that a mouse bone marrow macrophage cell line, maturating in vitro, expressed both higher PCA and sensitivity to macrophage procoagulant-inducing factor. The differences in responses to that factor were also demonstrated between different macrophage populations. 2 Resident, elicited and activated peritoneal murine macrophages were also found to differ in adherence and spreading, 26 and PCA. 27 Resident and in vivo activated macrophage migrated two-fold smaller out of capillary tubes than elicited ones. 22 A part of functional differences may originate from the observation that resident macrophages do not have membrane integrin LFA-1 but inflammatory cells do. 28 It should be kept in mind that elicited peritoneal macrophages are usually harvested after 3-4 days of induction, so that the data concerning PCA and migration are closer to our findings with 3-day macrophages than with 6-day macrophages.
The decrease of PCA in the course of both in vivo inflammation and in vitro cultivation may be the consequence of the enhanced opposite fibrinolytic activity, also attributed to mononuclear phagocytes. 19'29 Parallel tests of both functions point to the possible importance of a balance between procoagulant and fibrinolytic activities in the regulation of macrophage migration at sites of inflammation. 27'29'3 The presented results imply that activation of PCA may retain macrophages at the site of inflammation thus enabling them to accomplish appropriate biological activities.