Int J. Exp. Path. (I990) 71, 7I7-725
Recombinant gamma interferon causes neutrophil migration mediated by the release of a macrophage neutrophil chemotactic factor Ronaldo A. Ribeiro, Fernando Q. Cunha and Sergio H. Ferreira Department of Pharmacology, Faculty of Medicine of Ribeirdo Preto-USP-Ribeirao Preto, SP, Brazil
Received for publication 6 June I989 Accepted for publication i i May I990
Summary. A dose-dependent neutrophil migration was observed following the injection of purified (Hu IFN-y) or recombinant (rIFN-y) human gamma interferon into rat peritoneal cavities. This finding contrasts with their inability to cause chemotaxis in vitro in the Boyden chamber. Neutrophil migration into peritoneal cavities and subcutaneous air pouches induced by both preparations of interferon was abolished by pretreatment of the animals with dexamethasone. IFN-y-induced neutrophil migration was enhanced when the macrophage population of the peritoneal cavities was increased by previous injection of thioglycollate and reduced by peritoneal lavage. Macrophage monolayers pretreated either with rIFN-y or with lipopolysaccharide from E. coli release into the supernatant a factor that stimulates neutrophil recruitment in animals treated with dexamethasone. Dexamethasone blocked this release but did not affect the neutrophil recruitment induced by this factor. These results suggest that IFNy-induced neutrophil migration in vivo may be mediated by the release from resident macrophages of a neutrophil chemotactic factor and that dexamethasone blockade of neutrophil recruitment by IFN-y is due to inhibition of the release of this factor.
Keywords: recombinant gamma interferon, macrophages, neutrophil chemotactic factor Work from several laboratories has indicated that gamma interferon (IFN-y) is a macrophage activator, enhancing microbicidal activity (Nathan et al. I 983; Beaman I 98 7), and cytotoxicity against tumour target cells (Chen et al. I987), as well as the release of interleukin i (IL-i; I 985), tumour necrosis factor (TNF; Rook et al. I 98 7), leukotriene C4 (Saito et al. i988) and other products. IFN-y is involved in several pathological responses such as delayed-type hypersensitivity (Issekutz et al. I988), the Schwartzman reaction
(Billiau et al. I987), toxic shock syndrome (Jupin et al. I988), tuberculosis (Rook et al. I987), rheumatoid arthritis (Hopkins & Meager I988), sarcoidosis (Murray I988) and toxoplasmosis (Suzuki et al. I988). We decided to investigate whether IFN-y induces neutrophil migration since it is broadly accepted that IFN-y increases the host nonspecific defence mechanisms, of which neutrophil recruitment to an inflamed site is a crucial event. In this study we have investigated the ability of purified human IFN-y (Hu
Correspondence: Dr S.H. Ferreira, Departamento de Farmacologia, Faculdade de Medicina de Ribeirao Preto-USP, 14049-Ribeirao Preto-SP, Brazil.
717
R.A. Ribeiro et al. 7I8 IFN-y) and recombinant human IFN-y (rIFN- controlled rooms and received water and y) to induce neutrophil recruitment into the food ad libitum. subcutaneous air pouch and abdominal cavities of rats and in-vitro chemotaxis in the Drugs Boyden chamber. Several laboratories, including ours, have Human recombinant IFN-y (rIFN-y) derived shown that neutrophil migration into cavi- from E. coli (Gamma Interferon recDNACYS-TYR-CYS variety, 85/534, National ties and tissues induced by exogenous stimuli such as E. coli lipopolysaccharide (LPS), Institute for Biological Standards and Conlatex, bacteria, zymosan (Merril et al. 1978; trol, NIBSC, England. The ampoules of 2000 IU of recombinant IFN-y were contaminated Adamson & Bowden I982; Souza et al. I988) and by endogenous inflammatory with 25-50 IU of endotoxin, measured by the LAL assay, equivalent to 3.5-7 ng of E. mediators such as human purified IL-i (Cunha & Ferreira I986), rLL-i or rTNF coli endotoxin.). Human purified IFN-y (Hu (Faccioli et al. I989) is dependent on the IFN-y was prepared by induction of the release from macrophages of a neutrophil human leucocytes with phytohaemagglutinin-interferon, human leucocyte y, chemotactic factor (MNCF). Furthermore, 3000IU-82/587, NIBSC). Dexamethasone human monocytes are known to release a nseutrophil chemotactic factor which has (Decadronal) and indomethacin were purrecently been cloned (Yoshimura et al. chased from Merck Sharp & Dohme, Brazil. N- Formyl - L- methionyl - L - leucyl - L- phenylI 9 8 7a, Matsushima et al. I 98 8; Maestrelli et alanine (FMLP) was purchased from Sigma al. I 988). This factor is likely to belong to the Chemical Company, USA. Polymyxin B sulIL-8 superfamily (Van Damme et al. i989). Because we found in the early stages of phate was a gift from Wellcome Brazil. this study that IFN-y induces neutrophil recruitment in vivo but lacks chemotactic Experimental protocol activity in vitro, we decided to investigate whether neutrophil migration was mediated In-vivo neutrophil migration induced by IFN-y. Peritoneal cavities: rIFN-y in doses of 50 and via a chemotactic factor released by resident I 50 U (I U is equivalent to 0.5 ng) or Hu IFNmacrophages. This hypothesis was tested by verifying the ability of IFN-y to release from y in doses of 50 and I 50 U (i U is equivalent to 4.2 ng) were diluted in 3 ml of phosphatemacrophage monolayers a factor which induces neutrophil recruitment. Because of buffered saline (PBS), injected intraperitothe apparent similarities of action of LPS neally in normal rats. The highest dose (I 50 (Cunha & Ferreira I986) and IFN-y, we also U) was also injected in rats pretreated i h tested the inhibitory effect of dexamethasone earlier with dexamethasone (o. 5 mg/kg, on IFN-y-induced migration and on the s.c.). Control animals received 3ml of PBS i.p. release of this factor by IFN-y-stimulated To check the possibility of contamination macrophage monolayers. We conclude that with LPS, samples containing I50 U/3ml of both LPS and IFN-y causes neutrophil migrarIFN-y or 200 ng/3ml of LPS were incubated tion by a similar mechanism, i.e. by the during iomin with polymyxin B at final release of an endogenous chemotactic factor concentration of 3.5 ,ug/ml. Air pouch: Rat by resident cells, particularly macrophages. skin air pouches were produced as previously described by Edwards et al. (I 9 8 I). The backs Materials and methods of the rats were shaved and 20 ml of sterile air was injected subcutaneously. Three days Animals later, i oml of sterile air was again injected to Male Wistar rats weighing i80-200 g were maintain pouch patency. Six days after the used. Animals were housed in temperatureinitial injection of air, rIFN-y and Hu IFN-y
rIFN-y
causes
neutrophil migration by release of chemotactic factor
(I 50 U diluted in ml of PBS) were injected into the pouch of normal rats and rats i
pretreated (i h earlier) with dexamethasone (o.s mg/kg, s.c.). Controls received ml of PBS.
7I9
Table I. rIFN-y chemotactic activity in vitro. The results are means ± s.e.m. of N assays performed in duplicate
i
Chemotactic stimul.
Migration
(um)
Chemotactic index N
Collection of exudates
rIFN-y 5 U/ml
58.9±2.2
I.11±0.09
5
Four hours after injection of the test solutions, the animals were exsanguinated and the peritoneal or air pouch cells were harvested by injection of or 5 ml of PBS containing 5 U/ml of heparin, respectively.
50 U/ml FMLP (Io0' M) Random migration (RPMI)
61.3±2.2
i.i6±o.o8 5
79.8±6.5*
I.55±.03
5
io
Total and differential cell counts were performed as described elsewhere (Souza & Ferreira I985). The results were reported as the number of neutrophils per ml of fluid. Measurement of neutrophil migration in vitro (Boyden chamber)
rIFN-y and Hu IFN-y were diluted in full medium (RPMI I640, Difco) containing bovine serum albumin (i mg/ml) to obtain final concentrations of 5 and 50 U/ml. FMLP was diluted in the same medium and used at a concentration of o7 M. Neutrophils were obtained from rat peritoneal exudates 3h after challenge with carrageenin (0.3 mg/3 ml, i.p.). The cells were harvested with RPMI (pH 7.4), centrifuged, and resuspended in the same medium to give 5 106 neutrophils/ml. Each neutrophil suspension from a single peritoneal cavity was tested in duplicate for the chemotactic stimulus. IFN-y and FMLP were placed in the lower compartment of the Boyden chamber. The cell suspensions were placed in the upper compartment (io6 neutrophils). PMN chemotaxis was estimated go min after incubation at 370C, by the leading front method (Zigmond & Hirsch 1973; Wilkinson & Allan I978). For each stimulus, the chemotactic indexes represent the mean of the values obtained by dividing the migration estimated for each neutrophil suspension tested by the random migration observed with the incubating medium alone (Table i). i
X
54.0±4.5
5
*P90%), was followed by a significant de- Ferreira I986; Russo I980). IFN-y as shown crease in the neutrophil migration induced in this paper shares several chemotactic by rLFN-y (50 U). Similar results were also properties with LPS: (i) they did not show found with Hu IFN-y (data not shown). chemotactic activity in vitro, (2) their neuIncubation of macrophage monolayers trophil-induced migration is potentiated by with rIFN-y, as with LPS, caused the release the increase of the peritoneal population of into the supernatant of a factor which, when macrophages and reduced when the macroinjected intraperitoneally caused neutrophil phage population was decreased by peritomigration in rats pretreated with dexameth- neal lavage, (3) in-vivo neutrophil migration E
L
-r
_
R.A. Ribeiro et al.
722 2.0 r 6 T
6~~~
1.5 x en
-T) ._
a
0
1.0
6
.0
E z
0.5
0.0' C
LPS
3. Dexamethasone, but not indomethacin inhibits the release of a neutrophil chemotactic factor by macrophage monolayers incubated with rIFN-y. The bars represent neutrophil migration induced by intraperitoneal injection into animals pretreated with dexamethasone (0.5 mg/kg) of the supernatants (3 ml) obtained by incubation of macrophage monolayers with 0, RPMI-(C); LPS (io jug/ml); *, rIFN-y (io U/ml); rIFN-y (io U/ml) with dexamethasone (2 ,ug/ml), or El, rIFN-y (i U/ml) with indomethacin (2 Pg/ml). The results are means and s.e.m. of the number of animals indicated above each panel. The asterisks indicate a significant difference from the control value (Student's t-test, P
Recombinant gamma interferon causes neutrophil migration mediated by the release of a macrophage neutrophil chemotactic factor Ronaldo A. Ribeiro, Fernando Q. Cunha and Sergio H. Ferreira Department of Pharmacology, Faculty of Medicine of Ribeirdo Preto-USP-Ribeirao Preto, SP, Brazil
Received for publication 6 June I989 Accepted for publication i i May I990
Summary. A dose-dependent neutrophil migration was observed following the injection of purified (Hu IFN-y) or recombinant (rIFN-y) human gamma interferon into rat peritoneal cavities. This finding contrasts with their inability to cause chemotaxis in vitro in the Boyden chamber. Neutrophil migration into peritoneal cavities and subcutaneous air pouches induced by both preparations of interferon was abolished by pretreatment of the animals with dexamethasone. IFN-y-induced neutrophil migration was enhanced when the macrophage population of the peritoneal cavities was increased by previous injection of thioglycollate and reduced by peritoneal lavage. Macrophage monolayers pretreated either with rIFN-y or with lipopolysaccharide from E. coli release into the supernatant a factor that stimulates neutrophil recruitment in animals treated with dexamethasone. Dexamethasone blocked this release but did not affect the neutrophil recruitment induced by this factor. These results suggest that IFNy-induced neutrophil migration in vivo may be mediated by the release from resident macrophages of a neutrophil chemotactic factor and that dexamethasone blockade of neutrophil recruitment by IFN-y is due to inhibition of the release of this factor.
Keywords: recombinant gamma interferon, macrophages, neutrophil chemotactic factor Work from several laboratories has indicated that gamma interferon (IFN-y) is a macrophage activator, enhancing microbicidal activity (Nathan et al. I 983; Beaman I 98 7), and cytotoxicity against tumour target cells (Chen et al. I987), as well as the release of interleukin i (IL-i; I 985), tumour necrosis factor (TNF; Rook et al. I 98 7), leukotriene C4 (Saito et al. i988) and other products. IFN-y is involved in several pathological responses such as delayed-type hypersensitivity (Issekutz et al. I988), the Schwartzman reaction
(Billiau et al. I987), toxic shock syndrome (Jupin et al. I988), tuberculosis (Rook et al. I987), rheumatoid arthritis (Hopkins & Meager I988), sarcoidosis (Murray I988) and toxoplasmosis (Suzuki et al. I988). We decided to investigate whether IFN-y induces neutrophil migration since it is broadly accepted that IFN-y increases the host nonspecific defence mechanisms, of which neutrophil recruitment to an inflamed site is a crucial event. In this study we have investigated the ability of purified human IFN-y (Hu
Correspondence: Dr S.H. Ferreira, Departamento de Farmacologia, Faculdade de Medicina de Ribeirao Preto-USP, 14049-Ribeirao Preto-SP, Brazil.
717
R.A. Ribeiro et al. 7I8 IFN-y) and recombinant human IFN-y (rIFN- controlled rooms and received water and y) to induce neutrophil recruitment into the food ad libitum. subcutaneous air pouch and abdominal cavities of rats and in-vitro chemotaxis in the Drugs Boyden chamber. Several laboratories, including ours, have Human recombinant IFN-y (rIFN-y) derived shown that neutrophil migration into cavi- from E. coli (Gamma Interferon recDNACYS-TYR-CYS variety, 85/534, National ties and tissues induced by exogenous stimuli such as E. coli lipopolysaccharide (LPS), Institute for Biological Standards and Conlatex, bacteria, zymosan (Merril et al. 1978; trol, NIBSC, England. The ampoules of 2000 IU of recombinant IFN-y were contaminated Adamson & Bowden I982; Souza et al. I988) and by endogenous inflammatory with 25-50 IU of endotoxin, measured by the LAL assay, equivalent to 3.5-7 ng of E. mediators such as human purified IL-i (Cunha & Ferreira I986), rLL-i or rTNF coli endotoxin.). Human purified IFN-y (Hu (Faccioli et al. I989) is dependent on the IFN-y was prepared by induction of the release from macrophages of a neutrophil human leucocytes with phytohaemagglutinin-interferon, human leucocyte y, chemotactic factor (MNCF). Furthermore, 3000IU-82/587, NIBSC). Dexamethasone human monocytes are known to release a nseutrophil chemotactic factor which has (Decadronal) and indomethacin were purrecently been cloned (Yoshimura et al. chased from Merck Sharp & Dohme, Brazil. N- Formyl - L- methionyl - L - leucyl - L- phenylI 9 8 7a, Matsushima et al. I 98 8; Maestrelli et alanine (FMLP) was purchased from Sigma al. I 988). This factor is likely to belong to the Chemical Company, USA. Polymyxin B sulIL-8 superfamily (Van Damme et al. i989). Because we found in the early stages of phate was a gift from Wellcome Brazil. this study that IFN-y induces neutrophil recruitment in vivo but lacks chemotactic Experimental protocol activity in vitro, we decided to investigate whether neutrophil migration was mediated In-vivo neutrophil migration induced by IFN-y. Peritoneal cavities: rIFN-y in doses of 50 and via a chemotactic factor released by resident I 50 U (I U is equivalent to 0.5 ng) or Hu IFNmacrophages. This hypothesis was tested by verifying the ability of IFN-y to release from y in doses of 50 and I 50 U (i U is equivalent to 4.2 ng) were diluted in 3 ml of phosphatemacrophage monolayers a factor which induces neutrophil recruitment. Because of buffered saline (PBS), injected intraperitothe apparent similarities of action of LPS neally in normal rats. The highest dose (I 50 (Cunha & Ferreira I986) and IFN-y, we also U) was also injected in rats pretreated i h tested the inhibitory effect of dexamethasone earlier with dexamethasone (o. 5 mg/kg, on IFN-y-induced migration and on the s.c.). Control animals received 3ml of PBS i.p. release of this factor by IFN-y-stimulated To check the possibility of contamination macrophage monolayers. We conclude that with LPS, samples containing I50 U/3ml of both LPS and IFN-y causes neutrophil migrarIFN-y or 200 ng/3ml of LPS were incubated tion by a similar mechanism, i.e. by the during iomin with polymyxin B at final release of an endogenous chemotactic factor concentration of 3.5 ,ug/ml. Air pouch: Rat by resident cells, particularly macrophages. skin air pouches were produced as previously described by Edwards et al. (I 9 8 I). The backs Materials and methods of the rats were shaved and 20 ml of sterile air was injected subcutaneously. Three days Animals later, i oml of sterile air was again injected to Male Wistar rats weighing i80-200 g were maintain pouch patency. Six days after the used. Animals were housed in temperatureinitial injection of air, rIFN-y and Hu IFN-y
rIFN-y
causes
neutrophil migration by release of chemotactic factor
(I 50 U diluted in ml of PBS) were injected into the pouch of normal rats and rats i
pretreated (i h earlier) with dexamethasone (o.s mg/kg, s.c.). Controls received ml of PBS.
7I9
Table I. rIFN-y chemotactic activity in vitro. The results are means ± s.e.m. of N assays performed in duplicate
i
Chemotactic stimul.
Migration
(um)
Chemotactic index N
Collection of exudates
rIFN-y 5 U/ml
58.9±2.2
I.11±0.09
5
Four hours after injection of the test solutions, the animals were exsanguinated and the peritoneal or air pouch cells were harvested by injection of or 5 ml of PBS containing 5 U/ml of heparin, respectively.
50 U/ml FMLP (Io0' M) Random migration (RPMI)
61.3±2.2
i.i6±o.o8 5
79.8±6.5*
I.55±.03
5
io
Total and differential cell counts were performed as described elsewhere (Souza & Ferreira I985). The results were reported as the number of neutrophils per ml of fluid. Measurement of neutrophil migration in vitro (Boyden chamber)
rIFN-y and Hu IFN-y were diluted in full medium (RPMI I640, Difco) containing bovine serum albumin (i mg/ml) to obtain final concentrations of 5 and 50 U/ml. FMLP was diluted in the same medium and used at a concentration of o7 M. Neutrophils were obtained from rat peritoneal exudates 3h after challenge with carrageenin (0.3 mg/3 ml, i.p.). The cells were harvested with RPMI (pH 7.4), centrifuged, and resuspended in the same medium to give 5 106 neutrophils/ml. Each neutrophil suspension from a single peritoneal cavity was tested in duplicate for the chemotactic stimulus. IFN-y and FMLP were placed in the lower compartment of the Boyden chamber. The cell suspensions were placed in the upper compartment (io6 neutrophils). PMN chemotaxis was estimated go min after incubation at 370C, by the leading front method (Zigmond & Hirsch 1973; Wilkinson & Allan I978). For each stimulus, the chemotactic indexes represent the mean of the values obtained by dividing the migration estimated for each neutrophil suspension tested by the random migration observed with the incubating medium alone (Table i). i
X
54.0±4.5
5
*P90%), was followed by a significant de- Ferreira I986; Russo I980). IFN-y as shown crease in the neutrophil migration induced in this paper shares several chemotactic by rLFN-y (50 U). Similar results were also properties with LPS: (i) they did not show found with Hu IFN-y (data not shown). chemotactic activity in vitro, (2) their neuIncubation of macrophage monolayers trophil-induced migration is potentiated by with rIFN-y, as with LPS, caused the release the increase of the peritoneal population of into the supernatant of a factor which, when macrophages and reduced when the macroinjected intraperitoneally caused neutrophil phage population was decreased by peritomigration in rats pretreated with dexameth- neal lavage, (3) in-vivo neutrophil migration E
L
-r
_
R.A. Ribeiro et al.
722 2.0 r 6 T
6~~~
1.5 x en
-T) ._
a
0
1.0
6
.0
E z
0.5
0.0' C
LPS
3. Dexamethasone, but not indomethacin inhibits the release of a neutrophil chemotactic factor by macrophage monolayers incubated with rIFN-y. The bars represent neutrophil migration induced by intraperitoneal injection into animals pretreated with dexamethasone (0.5 mg/kg) of the supernatants (3 ml) obtained by incubation of macrophage monolayers with 0, RPMI-(C); LPS (io jug/ml); *, rIFN-y (io U/ml); rIFN-y (io U/ml) with dexamethasone (2 ,ug/ml), or El, rIFN-y (i U/ml) with indomethacin (2 Pg/ml). The results are means and s.e.m. of the number of animals indicated above each panel. The asterisks indicate a significant difference from the control value (Student's t-test, P
