Neutrophil recruitment inhibitory factor: a possible candidate for a novel cytokine

Inhibitory effect upon neutrophil migration to the inflammatory focus was previously detected in the cell-free incubation fluid of lipopolysaccharide (LPS)-stimulated macrophage monolayers. In the present study we showed that the neutrophil recruitment inhibitory activity from this supernatant was mainly detected in a fraction (P2) obtained by gel filtration chromatography on Sephacryl S-300. P2 fraction was able to inhibit ‘in vivo’ neutrophil emigration induced by different inflammatory stimuli, but it did not affect ‘in vitro’ neutrophil chemotaxis induced by FMLP. When injected intravenously, P2 inhibited oedema induced by carrageenin or immunological stimulus but not the oedema induced by dextran, thus affecting cell-dependent inflammatory responses. It was observed that P2 also induced neutrophil migration when injected locally in peritoneal cavities. This activity was significantly reduced by pretreatment of the animals with dexamethasone. Cytokines, such as IL-8 and TNF-α that are known to exhibit inhibitory effect upon neutrophil migration, were not detected in P2 fraction by highly sensitive assays. Overall the results suggest the existence of a novel cytokine exhibiting ‘in vivo’ neutrophil inhibitory activity, referred as NRIF.


Introduction
Cytokines, such as IL-6, IL-8 and TNF-, besides their pro-inflammatory effects 1-4 also exhibit antiinflammatory activities, such as inhibition of neutrophil adhesion to activated endothelial cells and 'in vivo' inhibition of neutrophil migration into inflammatory focus. (-8 These cytokines are released by various types of activated cells, such as those of phagocyte mononuclear system, 9'1 polymorphonuclear leucocytes, 1 endothelial cells and Tlymphocytes. 12 '3 In our laboratory we have examined the activity of supernatants from macrophage monolayers stimulated with LPS. It was found that this crude supernatant is able to induce neutrophil migration when injected into peritoneal cavities of dexamethasone-pretreated rats. This activity was referred as macrophage-derived neutrophil chemotactic factor (MNCF). .4 In contrast, neutrophil migration to peritoneal cavities induced by cytokines such as TNF-z and fl, IL-lz and gamma interferon 16 and IL-8 7 are inhibited by dexamethasone pretreatment. In addition, the same supernatant when intravenously (i.v.) injected was shown to inhibit in a dose dependent manner neutrophil migration to an inflamed site. 8 The factor responsible for this activity was named neutrophil recruitment inhibitory factor (NRIF).
Therefore the objective of this study was to characterize some biological properties of NRIF. For this purpose, we have submitted crude macrophage supernatant to gel filtration chromatography obtaining a fraction (P2) that retained the original inhibitory activity. We also observed that the i.v. administration of P2 was able to block cell dependent oedema. As IL-8 and TNF, cytokines for which neutrophil recruitment inhibitory activity has been described, <7 were not detected in P2 fraction, that in addition was eluted far from the elution volume expected for these cytokines, we suggest that NRIF may constitute a novel cytokine.

Material and Methods
Animals" Adult male Wistar rats weighing 150-180 g were housed in a temperature-controlled room and received water and food ad libitum until use. inhibitory activity, neutrophil suspensions (106 cells/ml) were incubated for 30 min at 37C, in a humidified incubator with 5% CO2, either in the absence or in the presence of the pooled chromatographic fractions of macrophage supernatant, now referred as fractions P1 to P10, prior to testing for chemotactic response to Nformyl-methionyl-L-leucyl-L-phenylalanine (FMLP, Sigma) in microchemotaxis assay. The chromatographic fractions were diluted in RPMI-BSA in order to give solutions containing the equivalent to the product released by 107 adherent macrophages/ml and were present throughout the assay. Chemotaxis assay. Chemotaxis was performed in 48-well chemotaxis chambers (Neuroprobe Inc., Cabin John, MD, USA). Polyvinylpirrolydone free polycarbonate filters (0.5/.tm) (Millipore) were placed between the upper and lower chambers.
Twenty-five #1 samples of either chemoattractant or RPMI-BSA were added to the wells of the lower chamber. FMLP was used at concentration of retaining maximal inhibitory activity was i.v. injected per animal, except when indicated in the Results section. The results are expressed as means -+-SEM of the number of neutrophils per ml of peritoneal wash.
Detection of MNCF activity. The S-300 chromatographic fractions were also assayed for the presence of MNCF activity as described elsewhere. 14 Briefly, the fractions were injected into peritoneal cavities (3 ml/cavity, equivalent to the material released by 5 106 macrophages) of dexamethasone-pretreated rats (0.5mgkg -1, s.c., 1 h before). Controls received sterile saline both s.c. and i.p. injected. After 4 h, neutrophil migration was measured as indicated for inhibition assay.
TNF activity assay: TNF content in macrophage supernatants and S-300 chromatographic fractions was measured by using a highly TNF-sensitive cell line, WEHI 164 clone 13, as described elsewhere. 2  incubated for an 20h, at 37C in a 5% CO2 incubator. Then, 10 #1 of a 3-4,5-dimethylthiazol-2yl-3,5-diphenylformazan solution (5 mg m1-1 PBS) (MTT, Sigma) were added to each well and the plates incubated for an additional 4 h. After that, 100/1 of isopropanol containing 0.04 N HC1 were added to each well. Fifteen min later, the degree of cell lysis was quantitated spectrophotometrically (630 nm) by using an enzyme-linked immunoassay analyser (Multiskan MCC/340 MKII, Flow Laboratories). Because of the unavailability of rat TNF, standard curves were done with mrTNF-(Genentech Inc.) and TNF content in assayed materials were calculated by comparison with the standard amounts. The results correspond to the means of data obtained in at least three different experiments. In some experiments, we use a rabbit anti-mrTNF antibody to study the involvement of TNF-like cytokine in our samples. Measurement of rat paw oedema: Oedema was measured plethysmographically. 2 The volume increase (A volume) of the inflamed paw was obtained by comparing the observed volume before and after the intra-plantar (i.pl.) administration of 100 #g/0.1 ml of carrageenin, dextran (Dextran 70, Pharmacia) or OVA. Animals were pretreated with either sterile PBS or P2 chromatographic fraction 15 min before i.pl. challenge. Oedema was mea-sured 2, 4 or 6 h after i.pl. challenge, except when indicated.

Results
Neutrophil migration activities of S-300 chromatographic fractions: The 'in vivo' inhibitory activity displayed by crude NRIF was detected in fraction P2 of S-300 chromatography of macrophage supernatants (Fig.  1, Panel A). The elution volume of P2 corresponded to an apparent molecular weight of 500 KDa.
When tested for TNF activity, crude supernatant was demonstrated to contain significant amounts (500 ng/0.2 ml), but only small amounts of such cytokine (1 pg m1-1) were detected in P2 fraction.
Antibodies raised against mouse recombinant TNFabolished the cytotoxicity of the mrTNF-0 and strongly reduced the activity of the crude supernatant (50-60% of inhibition). We have tested increasing doses of mrTNF-0 (0.3 pg-3/g) upon carrageenin-induced neutrophil migration. Only TNF-0 doses 106 times greater than that found in P2 were able to significantly inhibit neutrophil migration to rat peritoneal cavities (data not shown). In panel B of Fig. 1, it is shown that P2 fraction of S-300 chromatography did not inhibit 'in vitro' chemotaxis induced by FMLP, while it was found significant inhibition in fractions 5 to 10. When chromatographic fractions were tested for MNCF, the highest activity was eluted in the void volume (P1) and in a fraction corresponding to 30-60 KDa (P7) (data not shown). In addition, we have separately studied the effect of P2 injected into peritoneal cavities. Figure 2 shows that P2 caused neutrophil migration (left block of bars) that was significantly reduced by pretreatment of the animals with dexamethasone (right block of bars).
Besides its inhibitory effect upon neutrophil migration induced by carrageenin, P2 fraction also inhibited neutrophil recruitment induced by either LPS or FMLP into rat peritoneal cavities (Fig. 3).
Eect of P2 fraction upon oedema: Figure 4 shows that  LPS-stimulated macrophage supernatants. 18 This activity was referred to as neutrophil recruitment inhibitory factor (NRIF), since it was specific for polymorphonuclear leucocytes and in contrast with other substances, such as LPS, it did not inhibit mononuclear cell migration to rat peritoneal cavities. 22 This result indicates that NRIF is not contaminated with LPS. Further evidence that NRIF activity is not due to LPS contamination is based upon the fact that polymyxin B did not affect NRIF activity but abolished LPS neutrophil  Fig. 4. *p < 0.05 as compared to control. Student's t-test. recruitment inhibitory eEect. 18 Furthermore, resident macrophage monolayers release the same inhibitory activity without LPS stimulation. 23 In the present study we showed that NRIF activity was present in a fraction, referred as P2, obtained by gel filtration chromatography of the crude macrophage supernatant on Sephacryl S-300. This fraction was eluted in a volume corresponding to high molecular weight proteins (240-550 KDa) as deduced by standard calibration curve.
P2 fraction was effective against other inflammatory stimuli than carrageenin, such as EPS or FMLP, the latter being a direct chemotactic attractant 24 in 'in vivo' inhibitory assays. In contrast, P2 fraction did not inhibit neutrophil chemotaxis induced by FMLP in microchemotaxis assay. These results taken together, suggest that NRIF did not affect the ability of neutrophils to respond to chemotactic stimuli, thus supporting the suggestion that NRIF may be acting by blocking neutrophilendothelial adhesion mechanisms. 18 However, additional experiments with more purified material are required to clarify its precise mechanism of action.
Similar to NRIF, i.v. administration of TNF-0 has been shown to inhibit neutrophil accumulation at the site of inflammatory stimulus injection. Our results showed that crude macrophage supernatants contain significant TNF amounts (blocked by rabbit anti-mrTNF antibodies) that may contribute to inhibitory activity displayed by this material.
However, P2 fraction presented only traces of TNF, corresponding to a quantity 10 times slower than that necessary to present neutrophil recruitment inhibitory activity. Circumstantial evidence that NRIF inhibitory activity is not due to TNF is supported by the high apparent molecular weight of the P2 fraction.
IL-8 is another cytokine presently shown to inhibit neutrophil accumulation at sites of acute inflammation when administered intravenously. The possibility that NRIF activity may be contaminated by IL-8 is minimized by its high molecular weight elution volume and by the evidence that no traces of IL-8 was found in a sample of P2 fraction (ELISA, using a monoclonal antibody raised against human recombinant IL-8, sensitivity of detection smaller than 50 pg, see acknowledgements). It should be pointed out, however, that the absence of detection could be due to the low sensitivity of the test to rat IL-8.
However, in an on going experiment in our laboratory, it was found that antibodies raised against human IL-8 are able to inhibit either carrageenin as well as hlL-8-induced hyperalgesia in the rat.
Previous results obtained in our laboratory showed the presence of IL-1 in crude supernatant of LPS-stimulated macrophages. 18 However, as demonstrated by the authors, IL-1 does not block neutrophil migration to peritoneal cavities even if it is injected in doses higher than those detected in such supernatants.
Another inflammatory parameter, rat paw oedema induced by either carrageenin or antigen challenge, was significantly reduced by P2 fraction administration. This effect was not observed with dextran-induced oedema. Carrageenin and dextrans have been shown to induce increased vascular permeability by different mechanisms. While dextrans induce fluid accumulation due to mast cell degranulation with little protein and few neutrophils, carrageenin induces a protein-rich exudate containing large number of neutrophils. 25 Since NRIF did not present inhibitory activity upon dextran-induced oedema, we suggest that the factor only acts in polymorphonuclear-leucocyte dependent inflammatory responses. Thus, it is plausible that P2 is blocking immunologicallyinduced oedema by a similar mechanism. If this hypothesis is correct, we can envisage its usefulness in the control of immunological diseases such as bronchial asthma, where infiltration of inflammatory cells has been related to bronchial hyperresponsiveness. 26 Further evidence that NRIF activity is not due to IL-1 or TNF presence has been demonstrated by the results that i.v. reported to induce neutrophil migration. In contrast to MNCF, dexamethasone inhibited the activity of those known cytokines. 1>17 We have observed that 'in vivo' chemotactic activity displayed by P2 was significantly blocked by pretreatment of the animals with dexamethasone, thus indicating that there is a material which indirectly stimulates neutrophil emigration. This activity could result from the presence of traces of several cytokines or due to NRIF by itself. The residual emigration induced by P2 in rats pretreated with dexamethasone may result from the presence of a direct acting cytokine, such as MNCF since this activity was eluted in two peaks, one of them corresponding to the void volume fraction, eluted just before P2. Further studies are necessary to clarify this point.
In the light of the present results it is possible to assume NRIF as a candidate for a novel cytokine, capable of inhibiting neutrophil migration to an inflammatory site when present in the circulation, as might occur in endotoxemic shock.