- Email: [email protected]
The change in leukotrienes and lipoxins in activated mouse peritoneal macrophages
Biochimica et Biophysica Acta 1484 (2000) 87^92 www.elsevier.com/locate/bba
Rapid report
The change in leukotrienes and lipoxins in activated mouse peritoneal macrophages Lido Calorini *, Antonella Mannini, Francesca Bianchini, Gabriele Mugnai, Salvatore Ruggieri Department of Experimental Pathology and Oncology, University of Florence, Viale G.B. Morgagni 50, 50134 Florence, Italy Received 19 November 1999; received in revised form 27 January 2000; accepted 1 February 2000
Abstract The aim of this study was to investigate to what extent the generation of leukotrienes (LTs) and lipoxins (LXs) was affected by the expression of definite levels of macrophage activation. We used a system of murine peritoneal macrophages at different states of activation consisting in resident macrophages and FCS-, thioglycollate- or Corynebacterium parvumelicited macrophages. The profile of lipoxygenase metabolites in resident macrophages was characterized by the presence of high levels of 12-HETE, followed by 15-HETE, 5-HETE, LTB4 and 6-trans-LTB4 , 6-trans-12-epi-LTB4 . A comparable pattern was also found in FCS-elicited macrophages which appeared not to be responsive to the challenge with interferon Q plus LPS, as measured by the generation of NO and tumor necrosis factor K. Resident as well as FCS-elicited macrophages also generated appreciable quantities of LXs (A4 and B4 ). Thioglycollate-elicited macrophages, which expressed a state of `responsive' macrophages, showed a block of the LT and LX synthesis. This block was also present in C. parvum-elicited macrophages which expressed a fully `activated' phenotype, reflected by their capacity of releasing NO and tumor necrosis factor K even though they were not challenged. These results provide the first evidence that the level of `responsive' as well as `activated' macrophages was associated with of a simultaneous block of LTB4 and LXs. ß 2000 Elsevier Science B.V. All rights reserved. Keywords: Resident macrophage; Fetal calf serum-elicited macrophage; Thioglycollate-elicited macrophage; Corynebacterium parvumelicited macrophage; Prostaglandin; Leukotriene; Lipoxin ; 15-Hydroxy-eicosatetraenoic acid; 12-Hydroxy-eicosatetraenoic acid; 5-Hydroxy-eicosatetraenoic acid
Under the in£uence of various conditions that prevail in a tissue microenvironment, macrophages express di¡erent states of activation characterized by the acquisition of speci¢c activities [1] which have been explored in macrophages stimulated by various agents (e.g., thioglycollate broth, peptone, fetal calf serum (FCS), Corynebacterium parvum, bacillus
* Corresponding author. Fax: +39-55-416908; E-mail: [email protected]¢.it
Calmette-Gue¨rin (BCG)) [2^4]. Among these activities, the synthesis of arachidonic acid metabolites (prostaglandins (PGs) and leukotrienes (LTs)) has been particularly investigated in view of the role of these metabolites in the various biological responses in£uenced by macrophages [5]. The general trend that emerged from the investigations regarding PGs was a reduction of PGE2 and PGI2 synthesis during macrophage activation, as studied in murine peritoneal macrophages elicited with thioglycollate, C. parvum, or BCG [6,7], and with Listeria monocytogenes
1388-1981 / 00 / $ ^ see front matter ß 2000 Elsevier Science B.V. All rights reserved. PII: S 1 3 8 8 - 1 9 8 1 ( 0 0 ) 0 0 0 1 5 - 9
BBAMCB 50421 30-3-00
88
L. Calorini et al. / Biochimica et Biophysica Acta 1484 (2000) 87^92
[8]. A reduction of PGs (E2 , D2 , F2K and I2 ) was also reported for rabbit alveolar macrophages stimulated by BCG [9]. Data on the behavior of lipoxygenase products in activated macrophages were, on the other hand, somewhat contradictory. In fact, the synthesis of LTC4 was diminished in murine peritoneal macrophages elicited with C. parvum [7] or L. monocytogenes [8] and in chicken peritoneal macrophages elicited with Sephadex [10], but markedly increased in murine peritoneal thioglycollate-elicited macrophages stimulated with lipopolysaccharide (LPS) [11]. Moreover, the synthesis of LTB4 was increased in alveolar macrophages from rabbits treated with BCG or Freund's complete adjuvant [12], but decreased in alveolar macrophages collected from rats treated with Freund's complete adjuvant [13] and in peritoneal macrophages derived from rats treated with thioglycollate [14]. The inconsistency of the trend derived from these data might be related to di¡erences in the agents used for inducing macrophage activation, origin of macrophages, and animal species. Besides the lack of consensus of the data on the LTs produced by activated macrophages, there is also a lack of information regarding lipoxins (LXs) produced by macrophages at di¡erent levels of activation, although there are a few studies on LX synthesis in resident macrophages of a di¡erent origin [15^18]. LXs, a distinct series of lipoxygenase metabolites that contain a conjugated tetraene structure, are synthesized through the cooperation of di¡erent lipoxygenase (LO) activities [18]. LXs antagonize the stimulatory e¡ects of LTs on human polymorphonuclear neutrophils (PMNs) [18], but cooperate with LTs in promoting chemotaxis and adherence in human monocytes [19^21]. Using mouse peritoneal resident macrophages and macrophages elicited in vivo with FCS, thioglycollate or C. parvum, we investigated whether di¡erent levels of macrophage activation were associated with speci¢c pro¢les of LTs and LXs. According to Adams and Hamilton [1], the levels of activation expressed by the various macrophage populations were established on the basis of their responsiveness to interferon Q (IFNQ) and/or LPS, as measured by the generation of nitric oxide (NO) and tumor necrosis factor K (TNFK), two typical secretory products of activated macrophages [22,23].
Macrophage cultures were established from peritoneal exudates collected by lavage with 8 ml of icecold phosphate bu¡ered saline (PBS) from 6^8-weekold pathogen-free C57BL/6 female mice injected intraperitoneally with 1 ml of FCS (Boehringer Mannheim, Germany) containing endotoxin (0.1 ng/ml), with 1 ml of 3% thioglycollate (Sigma), or with 700 Wg of C. parvum (Coparwax, Wellcome Foundation, UK) suspended in 1 ml of PBS. The mice injected with C. parvum were killed 6 days later, and those injected with thioglycollate broth and FCS were killed 4 days later. The exudates were pooled into plastic tubes, and centrifuged at 250Ug for 10 min. The pelletted exudate cells were resuspended in DMEM 4500 containing 250 Wg/ml bovine serum albumin, seeded on 100 mm Falcon dishes, and incubated for 2 h at 37³C in a 10% CO2 humidi¢ed atmosphere to permit the adherence of macrophages. The dishes were washed several times with PBS in order to remove the non-adherent cells. The above described procedure was also used for establishing cultures of peritoneal resident macrophages. The morphological examination of monolayers stained with a May^Gru«nwald^Giemsa solution revealed that 85^95% of the adherent cells were macrophages, both resident and elicited by FCS, thioglycollate or C. parvum. In di¡erent preparations, the numbers of resident macrophages, and macrophages elicited with FCS, thioglycollate, and C. parvum were 0.5^ 0.6U106 , 0.7^1.2U106 , 5^8U106 and 1.0^1.2U106 per mouse, respectively. The analyses of lipoxygenase metabolites were performed on macrophage monolayers incubated for 20 min at 37³C in a pre-warmed PBS solution at pH 7.45 containing Ca2 (1.0 mM) and Mg2 (0.8 mM) chlorides, arachidonic acid (20 WM) (Cascade Biochem, UK), and calcium ionophore A23187 (5 WM) (Sigma). The arachidonic acid and the calcium ionophore A23187 were solubilized in ethanol, which did not exceed a 0.5% concentration in the incubation mixtures. The reaction was stopped by adding 2 vol. of cold methanol containing 50^ 100 ng of PGB2 used as an internal standard. This mixture was transferred into plastic tubes and centrifuged twice at 1600Ug for 10 min at 4³C in order to precipitate the proteinaceous material. The supernatant was transferred into round-bottom £asks and evaporated under vacuum. The residue was resus-
BBAMCB 50421 30-3-00
L. Calorini et al. / Biochimica et Biophysica Acta 1484 (2000) 87^92
89
Fig. 1. NO3 2 (A) and TNFK (B) released into growth media by murine peritoneal resident and FCS-, thioglycollate (TG)-, or C. parvum (Cp)-elicited macrophages. Macrophage cultures were grown for 18 h in the absence (unchallenged macrophages (black bars)), or in the presence (challenged macrophages) of LPS (10 ng/ml, dark gray bars), IFNQ (50 U/ml, light gray bars) or LPS plus IFNQ (white bars). Values are means þ S.E.M. Macrophage NO synthase activity was determined by measuring NOÿ 2 concentration in macrophage growth media. 500 Wl aliquots of macrophage growth media (a phenol red-free DMEM medium containing 250 Wg/ml bovine serum albumin) were incubated with an equal volume of the Griess reagent (250 Wl of 1% sulfanilamide diluted in 5% H3 PO4 plus 250 Wl of 0.1% naphthyl-ethylene diamine dihydrochloride) at room temperature for 15 min [24]. The absorbance of the reaction mixture was measured at 546 nm in a Uvikon spectrophotometer (Kontron, Switzerland). The NOÿ 2 content was determined by the use of a calibration curve constructed with a sodium nitrite standard. TNFK was assayed using the L-929 cell cytotoxicity test in the presence of 1 Wg/ml of actinomycin D [25], and is reported as percentage of cytotoxicity.
6
pended in 10 ml of methanol^H2 O (1:45, v/v) by vortexing and then was rapidly loaded into a C18 reverse phase (RP) cartridge (C18 Plus Sep-Pak, Waters Associates) after acidi¢cation to pH 3.5 with 0.1 N HCl. The cartridge was eluted with n-hexane (10 ml) followed by methyl formate (8 ml) and methanol (10 ml). The material within the methyl formate fraction (LTB4 , lipoxins, and monoHETEs) was concentrated under a stream of nitrogen and chromatographed on a LiChrospher 100 RP-18 column (4.6 mmU25 cm, 5 Wm particle size, LiChroCART, Merck) mounted in a HPLC apparatus (Bynary LC Pump, Model 250, Perkin Elmer).
Chromatography was performed at room temperature using a 19 min isocratic elution in solvent A (methanol:H2 O:acetic acid; 65:35:0.01, v/v), followed by a 1 min linear gradient from 100% solvent A to 100% solvent B (methanol:H2 O:acetic acid; 75:25:0.01, v/v), and a 40 min isocratic elution in solvent B. The £ow rate was 1.0 ml/min during the whole run. The e¥uents were analyzed with a photodiode array rapid spectral detector (Model 235, Perkin Elmer). The material eluted in the methanol fraction from C18 RP cartridges (peptidoleukotrienes) was analyzed on a LiChrospher 100 RP-18 column using an isocratic elution in methanol:H2 O:acetic acid (65:35:0.01, v/v, pH 5.7) at a £ow rate of 1.0 ml/min. The spectral detector was set at 280 nm. The lipoxygenase metabolites in the chromatograms were identi¢ed by comparison with authentic standards and veri¢ed by spectra analysis (see Fig. 2A). The levels of lipoxygenase metabolites was determined using standard curves constructed with authentic standards elaborated by a Personal Integrator, Nelson Model 1020 (Perkin Elmer); the values were corrected through the recovery of the PGB2 internal standard. The detection limit of the various lipoxygenase metabolites was 5 ng. The macrophage lipoxygenase products were also metabolically radiolabeled with [1-14 C]arachidonic acid (0.1 WCi/ml) (speci¢c activity 58 mCi/mmol, Amersham Life Sci-
BBAMCB 50421 30-3-00
90
L. Calorini et al. / Biochimica et Biophysica Acta 1484 (2000) 87^92
ence), chromatographed by RP-HPLC as above, and analyzed with an on line radioactivity detector set for 14 C (Flo-one/L, Series A-500, Radiomatic Camberra Packard). The radiochromatograms served to con¢rm that the peaks revealed with a UV detector derived from arachidonic acid. As shown in Fig. 1, unchallenged resident and FCS-elicited murine peritoneal macrophages released rather limited quantities of NOÿ 2 which remained unchanged even after being challenged with LPS or IFNQ, but slightly stimulated by LPS plus IFNQ. Unchallenged resident as well as FCS-elicited macrophages also secreted small amounts of TNFK into the growth media. These amounts were not modi¢ed after stimulation with IFNQ or LPS plus IFNQ. Unchallenged thioglycollate-elicited macrophages secreted the same amounts of NOÿ 2 and TNFK as those found in the growth media of resident macrophages; these amounts, however, were signi¢cantly enhanced after stimulation with LPS plus IFNQ, a behavior re£ecting a state of `responsive' macrophages. Comparable results were also observed by Ding et al. [26] and Drapier et al. [27]. C. parvum-elicited macrophages may be considered a macrophage population expressing a `fully activated' phenotype due to their capacity of releasing high quantities of NOÿ 2 and TNFK even in the absence of stimulation by LPS and/or IFNQ. As shown in Table 1, in resident macrophages, lipoxygenase metabolites showed a rather complex pro¢le composed of 12-HETE, followed by 15HETE and 5-HETE, a pro¢le also reported by other
laboratories [28,29]. Resident macrophages also produced LTB4 at a level comparable to that found by Humes et al. [30] in mouse peritoneal macrophages. The chromatograms of lipoxygenase metabolites derived from resident macrophages showed the two isomeric metabolites derived from the non-enzymatic hydrolysis of LTA4 (6-trans-LTB4 , 6-trans-12-epiLTB4 ), while g-oxidation products were absent. Our study also demonstrated for the ¢rst time that mouse peritoneal resident macrophages secrete LXs, A4 and B4 , in appreciable quantities (Fig. 2B). The production of LXs by resident macrophages might re£ect their high level of 12-LO, an activity which converts LTA4 into LXs [18,31,32], indicating a 5-LO/12-LO cooperation for the synthesis of LXs in mouse peritoneal macrophages. This metabolic cooperation inside a single cell represents an alternative to the already established transcellular metabolic pathway which leads to the production of LXs by combining the PMN 5-LO with the platelet 12-LO [18,31,32]. The FCS-elicited macrophages showed a low level of activation comparable to that of resident macrophages and exhibited a pro¢le of lipoxygenase products identical to that of resident macrophages. Once a `responsive' competence was acquired by the macrophages, as in thioglycollate-elicited macrophages, the generation of LTs and LXs was blocked and 15- and 12-HETE declined dramatically. These changes in LT and LX production were also present in the `fully activated' macrophages represented in our model system by C. parvum-elicited macro-
Table 1 Lipoxygenase products of peritoneal murine resident and FCS-, thioglycollate- or C. parvum-elicited macrophagesa Lipoxygenase products
Macrophages Resident
FCS-elicited
Thioglycollate-elicited
C. parvum-elicited
LXsb 6-trans-LTB4 +6-trans-12-epi-LTB4 LTB4 15-HETE 12-HETE 5-HETE
18.0 þ 3.8 31.9 þ 14.1 21.0 þ 8.4 60.9 þ 14.9 382.7 þ 113.8 27.2 þ 10.6
12.0 þ 4.4 14.0 þ 2.3 12.6 þ 5.6 67.3 þ 14.2 318.8 þ 124.6 20.8 þ 4.2
NDc ND ND 21.5 þ 9.2* 56.2 þ 21.9* 8.8 þ 3.8
ND ND ND 13.7 þ 8.2* 24.2 þ 3.5* 20.9 þ 3.1
*Signi¢cantly di¡erent at P 6 0.05 (Mann-Whitney test) from resident macrophages. a Values, expressed as ng/106 macrophages, are means þ S.E.M. of 3^4 separate experiments. Small quantities of LTC4 were also found in resident macrophages, although not consistently in di¡erent experiments. b LXs represent the sum of LXA4 , LXB4 and their trans isomers. c ND, not detected under the analytical conditions used.
BBAMCB 50421 30-3-00
L. Calorini et al. / Biochimica et Biophysica Acta 1484 (2000) 87^92
91
LTs and LXs might be instrumental in promoting the expression of functional properties which are characteristic of `activated macrophages' [5], as a result of the gene modulatory role of these arachidonic acid metabolites [38^42]. In view of the general counter-regulatory role attributed to LXs [18], it is possible that LXs produced by non-activated macrophages (resident and FCS-elicited macrophages in our study) exert an inhibitory in£uence on the expression of biological properties typical of an activated phenotype. These properties will be expressed only after the synthesis of LXs is blocked as shown in the `responsive' or `fully activated' phenotype. Indeed, the acute dermal in£ammation occurring in LTB4 receptor transgenic mice was found to be attenuated by LXs [43]. The authors wish to thank Prof. Alberto Fonnesu for his interest in this work. This study was supported by grants from CNR (ACRO project), MURST 40% Co¢n 1997 and MURST ex 60%.
References
Fig. 2. Spectra of 12-HETE (a), LTB4 (b) and LXB4 (c) eluted from HPLC columns (A) and a representative HPLC chromatogram (in a zoom evaluation mode) of LXs generated by resident murine peritoneal macrophages (B). The material eluted under peaks 1^4 showed spectra consistent with conjugated tetraene chromophores, and co-eluted with authentic standards of LXB4 , LXA4 and LX trans isomers.
phages. The block of the LT and LX generation in `responsive' and `activated' macrophages might be related to the inhibition of lipoxygenases caused by monoHETEs [33], which are released by macrophages stimulated by strongly in£ammatory agents [7,34]. PGE2 , and NO have also been reported to inhibit generation of LTs and LXs in rat alveolar macrophages [35,36] and mouse peritoneal macrophages [37]. The simultaneous down-regulation of
[1] D.O. Adams, T.A. Hamilton, Annu. Rev. Immunol. 2 (1984) 283^318. [2] J.B. Hibbs, R.R. Taintor, Methods Enzymol. 132 (1986) 508^520. [3] D.O. Adams, Methods Enzymol. 58 (1979) 494^506. [4] W.J. Johnson, P.A. Marino, R.D. Schreiber, D.O. Adams, J. Immunol. 131 (1983) 1038^1043. [5] R. Takemura, Z. Werb, Am. J. Physiol. 246 (1984) C1^ C9. [6] J.L. Humes, S. Burger, F.A. Kuehl Jr., P.D. Wightman, M.E. Dahlgren, P. Davies, R.J. Bonney, J. Immunol. 124 (1980) 2110^2116. [7] W.A. Scott, N.A. Pawlowski, H.W. Murray, M. Andreach, J. Zrike, Z.A. Cohn, J. Exp. Med. 155 (1982) 1148^1160. [8] C.S. Tripp, K.M. Leahy, P. Needleman, J. Clin. Invest. 76 (1985) 898^901. [9] W. Hsueh, R. Lumb, F. Gonzales-Crussi, Biochim. Biophys. Acta 710 (1982) 406^414. [10] J. Whelan, K.A. Golemboski, K.S. Broughton, J.E. Kinsella, R.R. Dietert, Prostaglandins Leukotrienes Essent. Fatty Acids 56 (1997) 41^49. [11] N.E. Hubbard, K.L. Erikson, Cell. Immunol. 160 (1995) 115^122. [12] W. Hsueh, F.F. Sun, S. Henderson, Biochim. Biophys. Acta 835 (1985) 92^97. [13] M. Shamsuddin, W. Hsueh, L.J. Smith, J. Lab. Clin. Med. 120 (1992) 434^443. [14] M. Abe, T. Goya, N. Nagata, H. Takahashi, H. Muranishi,
BBAMCB 50421 30-3-00
92
[15] [16] [17] [18] [19] [20] [21] [22]
[23] [24] [25] [26] [27] [28] [29]
L. Calorini et al. / Biochimica et Biophysica Acta 1484 (2000) 87^92 N. Shigematsu, Prostaglandins Leukotrienes Essent. Fatty Acids 45 (1992) 59^65. S.J. Kim, Biochem. Biophys. Res. Commun. 150 (1988) 870^ 876. T.R. Pettitt, A.F. Rowley, S.E. Barrow, A.I. Mallet, C.J. Secombes, J. Biol. Chem. 266 (1991) 8720^8726. M.E. Surette, J. Whelan, G.-P. Lu, J.E. Kinsella, Lipids 28 (1993) 1131^1134. C.N. Serhan, Biochim. Biophys. Acta 1212 (1994) 1^25. J.F. Maddox, C.N. Serhan, J. Exp. Med. 183 (1996) 137^ 146. J.F. Maddox, M. Hachicha, T. Takano, N.A. Petasis, V.V. Fokin, C.N. Serhan, J. Biol. Chem. 272 (1997) 6972^6978. J.F. Maddox, S.P. Colgan, C.L. Clish, N.A. Petasis, V.V. Kokin, C.N. Serhan, FASEB J. 12 (1998) 487^494. E.A. Carswell, L.J. Old, R.L. Kassel, S. Green, N. Fiore, B. Williamson, Proc. Natl. Acad. Sci. USA 72 (1975) 3666^ 3670. D.J. Stuehr, M.A. Marletta, Proc. Natl. Acad. Sci. USA 82 (1985) 7738^7742. B.B. Aggarwal, K. Mehta, Methods Enzymol. 269 (1996) 166^171. J.E. Belizario, C.A. Dinarello, Cancer Res. 51 (1991) 2379^ 2385. A.H. Ding, C.F. Nathan, D.J. Stuehr, J. Immunol. 141 (1988) 2407^2412. J.-C. Drapier, J. Wietzerbin, J.B. Hibbs Jr., Eur. J. Immunol. 18 (1988) 1587^1592. M. Rigaud, J. Durand, J.C. Breton, Biochim. Biophys. Acta 573 (1979) 408^412. H. Rabinovitch, J. Durand, M. Rigaud, F. Mendy, J.-C. Breton, Lipids 16 (1981) 518^524.
[30] J.L. Humes, S. Sadowski, M. Galavage, M. Goldenberg, E. Subers, R.J. Bonney, F.A. Kuehl Jr., J. Biol. Chem. 257 (1982) 1591^1594. [31] C.N. Serhan, K.-A. Sheppard, J. Clin. Invest. 85 (1990) 772^ 780. [32] M. Romano, C.N. Serhan, Biochemistry 31 (1992) 8269^ 8277. [33] J.L. Humes, E.E. Opas, M. Galavage, D. Soderman, R.J. Bonney, Biochem. J. 233 (1986) 199^206. [34] P. Borgeat, M. Nadeau, H. Salari, P. Poubelle, B. Fruteau de Laclos, Adv. Lipid Res. 21 (1985) 47^77. [35] G. Brunn, C. Hey, I. Wessler, K. Racke, Eur. J. Pharmacol. 326 (1997) 53^60. [36] B.W. Christman, J.W. Christman, R. Dworski, I.A. Blair, C. Prakash, J. Immunol. 151 (1993) 2096^2104. [37] P.D. Wightman, A. Dallob, J. Biol. Chem. 265 (1990) 9176^ 9180. [38] M. Rola-Pleszczynski, J. Stankova, Blood 80 (1992) 1004^ 1011. [39] J. Stankova, M. Rola-Pleszczynski, Biochem. J. 282 (1992) 625^629. [40] M.A. Brach, S. de Vos, C. Arnold, H.-J. Gruss, R. Mertelsmann, F. Herrmann, Eur. J. Immunol. 22 (1992) 2705^2711. [41] M. Rola-Pleszczynski, M. Thivierge, N. Gagnon, C. Lacasse, J. Stankova, J. Lipid Med. 6 (1993) 175^181. [42] C.N. Serhan, J.Z. Haeggstrom, C.C. Leslie, FASEB J. 10 (1996) 1147^1158. [43] N. Chiang, K. Gronert, C.B. Clish, J.A. O'Brien, M.W. Freeman, C.N. Serhan, J. Clin. Invest. 104 (1999) 309^ 316.
BBAMCB 50421 30-3-00
Rapid report
The change in leukotrienes and lipoxins in activated mouse peritoneal macrophages Lido Calorini *, Antonella Mannini, Francesca Bianchini, Gabriele Mugnai, Salvatore Ruggieri Department of Experimental Pathology and Oncology, University of Florence, Viale G.B. Morgagni 50, 50134 Florence, Italy Received 19 November 1999; received in revised form 27 January 2000; accepted 1 February 2000
Abstract The aim of this study was to investigate to what extent the generation of leukotrienes (LTs) and lipoxins (LXs) was affected by the expression of definite levels of macrophage activation. We used a system of murine peritoneal macrophages at different states of activation consisting in resident macrophages and FCS-, thioglycollate- or Corynebacterium parvumelicited macrophages. The profile of lipoxygenase metabolites in resident macrophages was characterized by the presence of high levels of 12-HETE, followed by 15-HETE, 5-HETE, LTB4 and 6-trans-LTB4 , 6-trans-12-epi-LTB4 . A comparable pattern was also found in FCS-elicited macrophages which appeared not to be responsive to the challenge with interferon Q plus LPS, as measured by the generation of NO and tumor necrosis factor K. Resident as well as FCS-elicited macrophages also generated appreciable quantities of LXs (A4 and B4 ). Thioglycollate-elicited macrophages, which expressed a state of `responsive' macrophages, showed a block of the LT and LX synthesis. This block was also present in C. parvum-elicited macrophages which expressed a fully `activated' phenotype, reflected by their capacity of releasing NO and tumor necrosis factor K even though they were not challenged. These results provide the first evidence that the level of `responsive' as well as `activated' macrophages was associated with of a simultaneous block of LTB4 and LXs. ß 2000 Elsevier Science B.V. All rights reserved. Keywords: Resident macrophage; Fetal calf serum-elicited macrophage; Thioglycollate-elicited macrophage; Corynebacterium parvumelicited macrophage; Prostaglandin; Leukotriene; Lipoxin ; 15-Hydroxy-eicosatetraenoic acid; 12-Hydroxy-eicosatetraenoic acid; 5-Hydroxy-eicosatetraenoic acid
Under the in£uence of various conditions that prevail in a tissue microenvironment, macrophages express di¡erent states of activation characterized by the acquisition of speci¢c activities [1] which have been explored in macrophages stimulated by various agents (e.g., thioglycollate broth, peptone, fetal calf serum (FCS), Corynebacterium parvum, bacillus
* Corresponding author. Fax: +39-55-416908; E-mail: [email protected]¢.it
Calmette-Gue¨rin (BCG)) [2^4]. Among these activities, the synthesis of arachidonic acid metabolites (prostaglandins (PGs) and leukotrienes (LTs)) has been particularly investigated in view of the role of these metabolites in the various biological responses in£uenced by macrophages [5]. The general trend that emerged from the investigations regarding PGs was a reduction of PGE2 and PGI2 synthesis during macrophage activation, as studied in murine peritoneal macrophages elicited with thioglycollate, C. parvum, or BCG [6,7], and with Listeria monocytogenes
1388-1981 / 00 / $ ^ see front matter ß 2000 Elsevier Science B.V. All rights reserved. PII: S 1 3 8 8 - 1 9 8 1 ( 0 0 ) 0 0 0 1 5 - 9
BBAMCB 50421 30-3-00
88
L. Calorini et al. / Biochimica et Biophysica Acta 1484 (2000) 87^92
[8]. A reduction of PGs (E2 , D2 , F2K and I2 ) was also reported for rabbit alveolar macrophages stimulated by BCG [9]. Data on the behavior of lipoxygenase products in activated macrophages were, on the other hand, somewhat contradictory. In fact, the synthesis of LTC4 was diminished in murine peritoneal macrophages elicited with C. parvum [7] or L. monocytogenes [8] and in chicken peritoneal macrophages elicited with Sephadex [10], but markedly increased in murine peritoneal thioglycollate-elicited macrophages stimulated with lipopolysaccharide (LPS) [11]. Moreover, the synthesis of LTB4 was increased in alveolar macrophages from rabbits treated with BCG or Freund's complete adjuvant [12], but decreased in alveolar macrophages collected from rats treated with Freund's complete adjuvant [13] and in peritoneal macrophages derived from rats treated with thioglycollate [14]. The inconsistency of the trend derived from these data might be related to di¡erences in the agents used for inducing macrophage activation, origin of macrophages, and animal species. Besides the lack of consensus of the data on the LTs produced by activated macrophages, there is also a lack of information regarding lipoxins (LXs) produced by macrophages at di¡erent levels of activation, although there are a few studies on LX synthesis in resident macrophages of a di¡erent origin [15^18]. LXs, a distinct series of lipoxygenase metabolites that contain a conjugated tetraene structure, are synthesized through the cooperation of di¡erent lipoxygenase (LO) activities [18]. LXs antagonize the stimulatory e¡ects of LTs on human polymorphonuclear neutrophils (PMNs) [18], but cooperate with LTs in promoting chemotaxis and adherence in human monocytes [19^21]. Using mouse peritoneal resident macrophages and macrophages elicited in vivo with FCS, thioglycollate or C. parvum, we investigated whether di¡erent levels of macrophage activation were associated with speci¢c pro¢les of LTs and LXs. According to Adams and Hamilton [1], the levels of activation expressed by the various macrophage populations were established on the basis of their responsiveness to interferon Q (IFNQ) and/or LPS, as measured by the generation of nitric oxide (NO) and tumor necrosis factor K (TNFK), two typical secretory products of activated macrophages [22,23].
Macrophage cultures were established from peritoneal exudates collected by lavage with 8 ml of icecold phosphate bu¡ered saline (PBS) from 6^8-weekold pathogen-free C57BL/6 female mice injected intraperitoneally with 1 ml of FCS (Boehringer Mannheim, Germany) containing endotoxin (0.1 ng/ml), with 1 ml of 3% thioglycollate (Sigma), or with 700 Wg of C. parvum (Coparwax, Wellcome Foundation, UK) suspended in 1 ml of PBS. The mice injected with C. parvum were killed 6 days later, and those injected with thioglycollate broth and FCS were killed 4 days later. The exudates were pooled into plastic tubes, and centrifuged at 250Ug for 10 min. The pelletted exudate cells were resuspended in DMEM 4500 containing 250 Wg/ml bovine serum albumin, seeded on 100 mm Falcon dishes, and incubated for 2 h at 37³C in a 10% CO2 humidi¢ed atmosphere to permit the adherence of macrophages. The dishes were washed several times with PBS in order to remove the non-adherent cells. The above described procedure was also used for establishing cultures of peritoneal resident macrophages. The morphological examination of monolayers stained with a May^Gru«nwald^Giemsa solution revealed that 85^95% of the adherent cells were macrophages, both resident and elicited by FCS, thioglycollate or C. parvum. In di¡erent preparations, the numbers of resident macrophages, and macrophages elicited with FCS, thioglycollate, and C. parvum were 0.5^ 0.6U106 , 0.7^1.2U106 , 5^8U106 and 1.0^1.2U106 per mouse, respectively. The analyses of lipoxygenase metabolites were performed on macrophage monolayers incubated for 20 min at 37³C in a pre-warmed PBS solution at pH 7.45 containing Ca2 (1.0 mM) and Mg2 (0.8 mM) chlorides, arachidonic acid (20 WM) (Cascade Biochem, UK), and calcium ionophore A23187 (5 WM) (Sigma). The arachidonic acid and the calcium ionophore A23187 were solubilized in ethanol, which did not exceed a 0.5% concentration in the incubation mixtures. The reaction was stopped by adding 2 vol. of cold methanol containing 50^ 100 ng of PGB2 used as an internal standard. This mixture was transferred into plastic tubes and centrifuged twice at 1600Ug for 10 min at 4³C in order to precipitate the proteinaceous material. The supernatant was transferred into round-bottom £asks and evaporated under vacuum. The residue was resus-
BBAMCB 50421 30-3-00
L. Calorini et al. / Biochimica et Biophysica Acta 1484 (2000) 87^92
89
Fig. 1. NO3 2 (A) and TNFK (B) released into growth media by murine peritoneal resident and FCS-, thioglycollate (TG)-, or C. parvum (Cp)-elicited macrophages. Macrophage cultures were grown for 18 h in the absence (unchallenged macrophages (black bars)), or in the presence (challenged macrophages) of LPS (10 ng/ml, dark gray bars), IFNQ (50 U/ml, light gray bars) or LPS plus IFNQ (white bars). Values are means þ S.E.M. Macrophage NO synthase activity was determined by measuring NOÿ 2 concentration in macrophage growth media. 500 Wl aliquots of macrophage growth media (a phenol red-free DMEM medium containing 250 Wg/ml bovine serum albumin) were incubated with an equal volume of the Griess reagent (250 Wl of 1% sulfanilamide diluted in 5% H3 PO4 plus 250 Wl of 0.1% naphthyl-ethylene diamine dihydrochloride) at room temperature for 15 min [24]. The absorbance of the reaction mixture was measured at 546 nm in a Uvikon spectrophotometer (Kontron, Switzerland). The NOÿ 2 content was determined by the use of a calibration curve constructed with a sodium nitrite standard. TNFK was assayed using the L-929 cell cytotoxicity test in the presence of 1 Wg/ml of actinomycin D [25], and is reported as percentage of cytotoxicity.
6
pended in 10 ml of methanol^H2 O (1:45, v/v) by vortexing and then was rapidly loaded into a C18 reverse phase (RP) cartridge (C18 Plus Sep-Pak, Waters Associates) after acidi¢cation to pH 3.5 with 0.1 N HCl. The cartridge was eluted with n-hexane (10 ml) followed by methyl formate (8 ml) and methanol (10 ml). The material within the methyl formate fraction (LTB4 , lipoxins, and monoHETEs) was concentrated under a stream of nitrogen and chromatographed on a LiChrospher 100 RP-18 column (4.6 mmU25 cm, 5 Wm particle size, LiChroCART, Merck) mounted in a HPLC apparatus (Bynary LC Pump, Model 250, Perkin Elmer).
Chromatography was performed at room temperature using a 19 min isocratic elution in solvent A (methanol:H2 O:acetic acid; 65:35:0.01, v/v), followed by a 1 min linear gradient from 100% solvent A to 100% solvent B (methanol:H2 O:acetic acid; 75:25:0.01, v/v), and a 40 min isocratic elution in solvent B. The £ow rate was 1.0 ml/min during the whole run. The e¥uents were analyzed with a photodiode array rapid spectral detector (Model 235, Perkin Elmer). The material eluted in the methanol fraction from C18 RP cartridges (peptidoleukotrienes) was analyzed on a LiChrospher 100 RP-18 column using an isocratic elution in methanol:H2 O:acetic acid (65:35:0.01, v/v, pH 5.7) at a £ow rate of 1.0 ml/min. The spectral detector was set at 280 nm. The lipoxygenase metabolites in the chromatograms were identi¢ed by comparison with authentic standards and veri¢ed by spectra analysis (see Fig. 2A). The levels of lipoxygenase metabolites was determined using standard curves constructed with authentic standards elaborated by a Personal Integrator, Nelson Model 1020 (Perkin Elmer); the values were corrected through the recovery of the PGB2 internal standard. The detection limit of the various lipoxygenase metabolites was 5 ng. The macrophage lipoxygenase products were also metabolically radiolabeled with [1-14 C]arachidonic acid (0.1 WCi/ml) (speci¢c activity 58 mCi/mmol, Amersham Life Sci-
BBAMCB 50421 30-3-00
90
L. Calorini et al. / Biochimica et Biophysica Acta 1484 (2000) 87^92
ence), chromatographed by RP-HPLC as above, and analyzed with an on line radioactivity detector set for 14 C (Flo-one/L, Series A-500, Radiomatic Camberra Packard). The radiochromatograms served to con¢rm that the peaks revealed with a UV detector derived from arachidonic acid. As shown in Fig. 1, unchallenged resident and FCS-elicited murine peritoneal macrophages released rather limited quantities of NOÿ 2 which remained unchanged even after being challenged with LPS or IFNQ, but slightly stimulated by LPS plus IFNQ. Unchallenged resident as well as FCS-elicited macrophages also secreted small amounts of TNFK into the growth media. These amounts were not modi¢ed after stimulation with IFNQ or LPS plus IFNQ. Unchallenged thioglycollate-elicited macrophages secreted the same amounts of NOÿ 2 and TNFK as those found in the growth media of resident macrophages; these amounts, however, were signi¢cantly enhanced after stimulation with LPS plus IFNQ, a behavior re£ecting a state of `responsive' macrophages. Comparable results were also observed by Ding et al. [26] and Drapier et al. [27]. C. parvum-elicited macrophages may be considered a macrophage population expressing a `fully activated' phenotype due to their capacity of releasing high quantities of NOÿ 2 and TNFK even in the absence of stimulation by LPS and/or IFNQ. As shown in Table 1, in resident macrophages, lipoxygenase metabolites showed a rather complex pro¢le composed of 12-HETE, followed by 15HETE and 5-HETE, a pro¢le also reported by other
laboratories [28,29]. Resident macrophages also produced LTB4 at a level comparable to that found by Humes et al. [30] in mouse peritoneal macrophages. The chromatograms of lipoxygenase metabolites derived from resident macrophages showed the two isomeric metabolites derived from the non-enzymatic hydrolysis of LTA4 (6-trans-LTB4 , 6-trans-12-epiLTB4 ), while g-oxidation products were absent. Our study also demonstrated for the ¢rst time that mouse peritoneal resident macrophages secrete LXs, A4 and B4 , in appreciable quantities (Fig. 2B). The production of LXs by resident macrophages might re£ect their high level of 12-LO, an activity which converts LTA4 into LXs [18,31,32], indicating a 5-LO/12-LO cooperation for the synthesis of LXs in mouse peritoneal macrophages. This metabolic cooperation inside a single cell represents an alternative to the already established transcellular metabolic pathway which leads to the production of LXs by combining the PMN 5-LO with the platelet 12-LO [18,31,32]. The FCS-elicited macrophages showed a low level of activation comparable to that of resident macrophages and exhibited a pro¢le of lipoxygenase products identical to that of resident macrophages. Once a `responsive' competence was acquired by the macrophages, as in thioglycollate-elicited macrophages, the generation of LTs and LXs was blocked and 15- and 12-HETE declined dramatically. These changes in LT and LX production were also present in the `fully activated' macrophages represented in our model system by C. parvum-elicited macro-
Table 1 Lipoxygenase products of peritoneal murine resident and FCS-, thioglycollate- or C. parvum-elicited macrophagesa Lipoxygenase products
Macrophages Resident
FCS-elicited
Thioglycollate-elicited
C. parvum-elicited
LXsb 6-trans-LTB4 +6-trans-12-epi-LTB4 LTB4 15-HETE 12-HETE 5-HETE
18.0 þ 3.8 31.9 þ 14.1 21.0 þ 8.4 60.9 þ 14.9 382.7 þ 113.8 27.2 þ 10.6
12.0 þ 4.4 14.0 þ 2.3 12.6 þ 5.6 67.3 þ 14.2 318.8 þ 124.6 20.8 þ 4.2
NDc ND ND 21.5 þ 9.2* 56.2 þ 21.9* 8.8 þ 3.8
ND ND ND 13.7 þ 8.2* 24.2 þ 3.5* 20.9 þ 3.1
*Signi¢cantly di¡erent at P 6 0.05 (Mann-Whitney test) from resident macrophages. a Values, expressed as ng/106 macrophages, are means þ S.E.M. of 3^4 separate experiments. Small quantities of LTC4 were also found in resident macrophages, although not consistently in di¡erent experiments. b LXs represent the sum of LXA4 , LXB4 and their trans isomers. c ND, not detected under the analytical conditions used.
BBAMCB 50421 30-3-00
L. Calorini et al. / Biochimica et Biophysica Acta 1484 (2000) 87^92
91
LTs and LXs might be instrumental in promoting the expression of functional properties which are characteristic of `activated macrophages' [5], as a result of the gene modulatory role of these arachidonic acid metabolites [38^42]. In view of the general counter-regulatory role attributed to LXs [18], it is possible that LXs produced by non-activated macrophages (resident and FCS-elicited macrophages in our study) exert an inhibitory in£uence on the expression of biological properties typical of an activated phenotype. These properties will be expressed only after the synthesis of LXs is blocked as shown in the `responsive' or `fully activated' phenotype. Indeed, the acute dermal in£ammation occurring in LTB4 receptor transgenic mice was found to be attenuated by LXs [43]. The authors wish to thank Prof. Alberto Fonnesu for his interest in this work. This study was supported by grants from CNR (ACRO project), MURST 40% Co¢n 1997 and MURST ex 60%.
References
Fig. 2. Spectra of 12-HETE (a), LTB4 (b) and LXB4 (c) eluted from HPLC columns (A) and a representative HPLC chromatogram (in a zoom evaluation mode) of LXs generated by resident murine peritoneal macrophages (B). The material eluted under peaks 1^4 showed spectra consistent with conjugated tetraene chromophores, and co-eluted with authentic standards of LXB4 , LXA4 and LX trans isomers.
phages. The block of the LT and LX generation in `responsive' and `activated' macrophages might be related to the inhibition of lipoxygenases caused by monoHETEs [33], which are released by macrophages stimulated by strongly in£ammatory agents [7,34]. PGE2 , and NO have also been reported to inhibit generation of LTs and LXs in rat alveolar macrophages [35,36] and mouse peritoneal macrophages [37]. The simultaneous down-regulation of
[1] D.O. Adams, T.A. Hamilton, Annu. Rev. Immunol. 2 (1984) 283^318. [2] J.B. Hibbs, R.R. Taintor, Methods Enzymol. 132 (1986) 508^520. [3] D.O. Adams, Methods Enzymol. 58 (1979) 494^506. [4] W.J. Johnson, P.A. Marino, R.D. Schreiber, D.O. Adams, J. Immunol. 131 (1983) 1038^1043. [5] R. Takemura, Z. Werb, Am. J. Physiol. 246 (1984) C1^ C9. [6] J.L. Humes, S. Burger, F.A. Kuehl Jr., P.D. Wightman, M.E. Dahlgren, P. Davies, R.J. Bonney, J. Immunol. 124 (1980) 2110^2116. [7] W.A. Scott, N.A. Pawlowski, H.W. Murray, M. Andreach, J. Zrike, Z.A. Cohn, J. Exp. Med. 155 (1982) 1148^1160. [8] C.S. Tripp, K.M. Leahy, P. Needleman, J. Clin. Invest. 76 (1985) 898^901. [9] W. Hsueh, R. Lumb, F. Gonzales-Crussi, Biochim. Biophys. Acta 710 (1982) 406^414. [10] J. Whelan, K.A. Golemboski, K.S. Broughton, J.E. Kinsella, R.R. Dietert, Prostaglandins Leukotrienes Essent. Fatty Acids 56 (1997) 41^49. [11] N.E. Hubbard, K.L. Erikson, Cell. Immunol. 160 (1995) 115^122. [12] W. Hsueh, F.F. Sun, S. Henderson, Biochim. Biophys. Acta 835 (1985) 92^97. [13] M. Shamsuddin, W. Hsueh, L.J. Smith, J. Lab. Clin. Med. 120 (1992) 434^443. [14] M. Abe, T. Goya, N. Nagata, H. Takahashi, H. Muranishi,
BBAMCB 50421 30-3-00
92
[15] [16] [17] [18] [19] [20] [21] [22]
[23] [24] [25] [26] [27] [28] [29]
L. Calorini et al. / Biochimica et Biophysica Acta 1484 (2000) 87^92 N. Shigematsu, Prostaglandins Leukotrienes Essent. Fatty Acids 45 (1992) 59^65. S.J. Kim, Biochem. Biophys. Res. Commun. 150 (1988) 870^ 876. T.R. Pettitt, A.F. Rowley, S.E. Barrow, A.I. Mallet, C.J. Secombes, J. Biol. Chem. 266 (1991) 8720^8726. M.E. Surette, J. Whelan, G.-P. Lu, J.E. Kinsella, Lipids 28 (1993) 1131^1134. C.N. Serhan, Biochim. Biophys. Acta 1212 (1994) 1^25. J.F. Maddox, C.N. Serhan, J. Exp. Med. 183 (1996) 137^ 146. J.F. Maddox, M. Hachicha, T. Takano, N.A. Petasis, V.V. Fokin, C.N. Serhan, J. Biol. Chem. 272 (1997) 6972^6978. J.F. Maddox, S.P. Colgan, C.L. Clish, N.A. Petasis, V.V. Kokin, C.N. Serhan, FASEB J. 12 (1998) 487^494. E.A. Carswell, L.J. Old, R.L. Kassel, S. Green, N. Fiore, B. Williamson, Proc. Natl. Acad. Sci. USA 72 (1975) 3666^ 3670. D.J. Stuehr, M.A. Marletta, Proc. Natl. Acad. Sci. USA 82 (1985) 7738^7742. B.B. Aggarwal, K. Mehta, Methods Enzymol. 269 (1996) 166^171. J.E. Belizario, C.A. Dinarello, Cancer Res. 51 (1991) 2379^ 2385. A.H. Ding, C.F. Nathan, D.J. Stuehr, J. Immunol. 141 (1988) 2407^2412. J.-C. Drapier, J. Wietzerbin, J.B. Hibbs Jr., Eur. J. Immunol. 18 (1988) 1587^1592. M. Rigaud, J. Durand, J.C. Breton, Biochim. Biophys. Acta 573 (1979) 408^412. H. Rabinovitch, J. Durand, M. Rigaud, F. Mendy, J.-C. Breton, Lipids 16 (1981) 518^524.
[30] J.L. Humes, S. Sadowski, M. Galavage, M. Goldenberg, E. Subers, R.J. Bonney, F.A. Kuehl Jr., J. Biol. Chem. 257 (1982) 1591^1594. [31] C.N. Serhan, K.-A. Sheppard, J. Clin. Invest. 85 (1990) 772^ 780. [32] M. Romano, C.N. Serhan, Biochemistry 31 (1992) 8269^ 8277. [33] J.L. Humes, E.E. Opas, M. Galavage, D. Soderman, R.J. Bonney, Biochem. J. 233 (1986) 199^206. [34] P. Borgeat, M. Nadeau, H. Salari, P. Poubelle, B. Fruteau de Laclos, Adv. Lipid Res. 21 (1985) 47^77. [35] G. Brunn, C. Hey, I. Wessler, K. Racke, Eur. J. Pharmacol. 326 (1997) 53^60. [36] B.W. Christman, J.W. Christman, R. Dworski, I.A. Blair, C. Prakash, J. Immunol. 151 (1993) 2096^2104. [37] P.D. Wightman, A. Dallob, J. Biol. Chem. 265 (1990) 9176^ 9180. [38] M. Rola-Pleszczynski, J. Stankova, Blood 80 (1992) 1004^ 1011. [39] J. Stankova, M. Rola-Pleszczynski, Biochem. J. 282 (1992) 625^629. [40] M.A. Brach, S. de Vos, C. Arnold, H.-J. Gruss, R. Mertelsmann, F. Herrmann, Eur. J. Immunol. 22 (1992) 2705^2711. [41] M. Rola-Pleszczynski, M. Thivierge, N. Gagnon, C. Lacasse, J. Stankova, J. Lipid Med. 6 (1993) 175^181. [42] C.N. Serhan, J.Z. Haeggstrom, C.C. Leslie, FASEB J. 10 (1996) 1147^1158. [43] N. Chiang, K. Gronert, C.B. Clish, J.A. O'Brien, M.W. Freeman, C.N. Serhan, J. Clin. Invest. 104 (1999) 309^ 316.
BBAMCB 50421 30-3-00