- Email: [email protected]
Kupffer Cells Are a Dominant Site of Uncoupling Protein 2 Expression in Rat Liver
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS ARTICLE NO.
235, 760–764 (1997)
RC976852
Kupffer Cells Are a Dominant Site of Uncoupling Protein 2 Expression in Rat Liver Dominique Larrouy,*,1 Patrick Laharrague,* Georges Carrera,† Nathalie Viguerie-Bascands,* Corinne Levi-Meyrueis,‡ Christophe Fleury,‡ Claire Pecqueur,‡ Maryse Nibbelink,* Mireille Andre´,* Louis Casteilla,*,1 and Daniel Ricquier‡,1 *CNRS/UPRESA 5018 and †INSERM CJF 91-07, Institut Louis Bugnard, Universite´ Paul Sabatier, Hoˆpital Rangueil, 31054 Toulouse Cedex, France; and ‡CNRS UPR 9078/CEREMOD, 9, rue Jules Hetzel, 92190 Meudon, France
Received May 15, 1997
The mechanisms underlying thermogenesis in liver are not well understood. They may involve proteins related to the mitochondrial uncoupling protein (UCP1) of brown adipocytes. In this paper, it is demonstrated that UCP1 is not expressed in any liver cell type of rat while UCP2, a recently cloned homologue of UCP1, is expressed at a very high level in Kupffer cells but not in hepatocytes. This high level of expression of UCP2 in Kupffer cells allowed cross immunoreactivity with antibodies directed against UCP1. This cross reactivity was confirmed by the detection of UCP2 with anti-UCP1 antibody, in western blotting analysis of transfected yeasts expressing rat UCP2. The high level expression of UCP2 in Kupffer cells suggests a particular function of UCP2 in macrophages. q 1997 Academic Press
Cold induced thermogenesis in rodents is performed in part by brown adipose tissue. The mitochondrial uncoupling protein UCP1, specifically expressed in brown fat cells, provokes heat dissipation by uncoupling of oxidative phosphorylation from ATP synthesis, during sympathetic activation of the brown fat (1, 2). The exclusive occurence of UCP1 in brown adipose tissue was ascertained by many investigators using immunodetection (3-5) or mRNA analysis (6-9). A contribution of liver to cold-induced thermogenesis, was also demonstrated. Nevertheless, although it was proposed that an uncoupling of respiration of liver mitochondria might exist in cold-exposed animals, the mechanism of such an uncoupling remains unknown (10, 11). The detection of UCP1 mRNA in liver of newborn rats, or 1 Corresponding authors. (D.R.) Fax: /33 1 45 07 58 90. E-mail: [email protected]. (L.C.) Fax: /33 5 62 17 09 05. E-mail: casteil@ rangueil.inserm.fr. Abbreviation: UCP2, uncoupling protein 2.
0006-291X/97 $25.00
adult rats exposed to the cold (12) was in agreement with this hypothesis, but was later on described as artefactual (13). Until now, the search for UCP1 and/or UCP1 mRNA in tissues other than brown adipose tissue, was performed mainly on homogenized tissues. Such experiments could give rise to misleading results if UCP1 was expressed only in a particular and sparse cell type present in a tissue. Indeed, liver is a combination of at least 12 different cell types (14). Hence, the initial aim of this study was to definitively establish if UCP1 is expressed in certain liver cell types using a combination of immunological, immuno-histochemical, and molecular methods. Immunological methods used in the present study, led to contradictory results, suggesting that UCP1 antibodies cross-reacted with an other mitochondrial protein expressed in non-parenchymal cells present in liver. Recently, a new protein belonging to the family of mitochondrial anion carriers was cloned from a mouse muscle cDNA library (15). This protein, named uncoupling protein 2 or UCP2, seems to have properties similar to UCP1 and is putatively involved in heat production, since it was shown to be able to uncouple the mitochondrial respiration from ATP synthesis. Moreover, the wide tissular expression of UCP2 agrees with a role in energy expenditure. We report here that the UCP1 immunoreactivity of rat liver revealed the presence of very high amount of UCP2 in non parenchymal cells and in particular in Kupffer cells. MATERIALS AND METHODS Animals. Male Wistar rats were housed at 257C with free access to food (65% carbohydrate, 11% fat and 24% protein w/w; UAR, Villemoisson France) and water. Antibodies. Sheep and rabbit polyclonal antibodies against rat UCP were previously described (4, 16). Donkey anti-sheep IgG alkaline phosphatase-conjugated was purchased from Sigma Chemical (St. Louis, MO). Pig anti-rabbit IgG alkaline phosphatase-conjugated
760
Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
AID
BBRC 6852
/
6930$$1501
06-13-97 14:08:22
bbrcg
AP: BBRC
Vol. 235, No. 3, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
was obtained from Dako corporation (Carpinteria, CA). Monoclonal antibody against macrophage (ED2) was from Serotec (Oxford). Antimouse IgG FITC conjugated was from Sigma Chemical Co. Preparation and analysis of RNA. Total RNA were prepared according to the single step method of Chomczynski and Sacchi (17). Northern blot analysis was performed as previously described (18). Rat UCP1 (19), mouse UCP2 (15) and rat albumin (20) cDNA probes were labelled with [a32P]dCTP using a megaprime labelling system kit (Boehringer, Mannheim Germany). Blots were analyzed using a phosphorimager (Molecular Dynamics). RT-PCR amplification of UCP1 mRNA was carried out as previously described (13, 18). RTPCR amplification of UCP2 mRNA was made using 5*-AATCGAATTCATGGTTGGTTTCAAGGCC-3* sense primer, and 5*-GGAGGAGCTCAGAAAGGTGCCTCCCGA-3* antisense primer. Moloney murine-leukemia-virus RT was obtained from Life Technology (Gaithersburg, MD), ribonuclease inhibitor from Pharmacia (Uppsala, Sweden) and Taq DNA polymerase from Appligene (Illkirch, France). Immunohistochemistry. Livers were fixed in Dubosq’Brasil immediately after removal, dehydrated in ethanol and paraffin embedded. 5 mm sections were deparaffinized, rehydrated and incubated in medium containing purified sheep IgG (1:2000 dilution) raised against rat UCP1 (4, 16) during 1 hour at room temperature, rinsed and incubated in presence of rabbit antisheep IgG (1:100 dilution) coupled to alkaline phosphatase. Alkaline phosphatase activity was revealed using Sigma Fast Red TR/Naphtol AS-MX tablets as substrate and counterstained with Harris’ hematoxylin. Control experiments were performed using purified sheep IgG (Sigma) and yielded no staining. Western blot analysis. Mitochondrial fractions were prepared by differential centrifugation of liver homogenate as described by Casteilla et al (21). 5-30mg of mitochondrial fraction or total homogenate liver proteins were electrophoresed in 10 % polyacrylamide SDS-gels, transferred onto nitrocellulose and incubated with sheep antiserum raised against rat UCP1 (4) (1:2,000 dilution) and rabbit IgG conjugated to peroxydase. Peroxydase activity was revealed using 4-chloro-1-naphtol and hydrogen peroxyde as substrate. Western analysis of yeast proteins was carried out using the same UCP1antibodies at a 1:20,000 dilution and the Renaissance system (Dupont de Nemours, Les Ulis). Expression in yeast. The expression of mouse UCP2 in diploid yeast (S. cerevisiae) strain W303 using pYeDP-UCP2 plasmid was previously described by Fleury et al. (15). Mitochondria were isolated from this yeast strain, from the same S. cerevisiae strain transfected by an empty pYeDP expression vector and from a yeast strain transfected by a pYeDP-UCP1 vector (21). Isolation of cells. Isolated hepatocytes and total non-parenchymal cells (NPCs) were obtained from rat liver using the two-step perfusion collagenase method (22). The resulting cell suspension was centrifuged at 50 g for 30 sec to separate the pellet of hepatocytes from the supernatant containing all the NPCs. For the studies of hepatocytes and Kupffer cells, the liver was first perfused by the two-step perfusion method with 0.03 % collagenase A (Boehringer Mannheim Gmbh, Mannheim, Germany) during 10 min. Then, the liver was excised and minced into 1-2 mm cubes using a razor blade. These fragments were incubated in 100 ml of Gey’s balanced salt solution (GBSS) containing 0.1 % pronase E (MerckCle´venot S.A., Nogent-sur-Marne, France) and 0.5 mg of deoxyribonuclease (Sigma Chemical Co., St. Louis, USA) for 40 min at 377 C and pH 7.4 with continuous stirring. The cells were filtered through nylon gauze, and centrifuged at 50 g for 30 sec to separate the pellet of hepatocytes from the supernatant containing the NPCs. The NPCs suspension was centrifuged for 10 min at 550 g and the resulting pellet washed two times using resuspension in 20 ml of GBSS and centrifugation. The Kupffer cells were separated from the other NPCs first by centrifugation in presence of Metrizamide (2-(3-acetamido-5-N-methylacetamido-2,4,6-triiodobenzamido)-2-deoxy-D-
glucose, Nycomed Pharma AS, Oslo, Norway), followed by centrifugal elutriation (23). The percentage of Kupffer cells in the resulting suspension was determined on a cytospin preparation by immunochemistry using ED2 (Serotec, Oxford, England) as primary antibody and an anti-mouse IgG FITC conjugate (Sigma Chemical Co) as secondary antibody.
RESULTS UCP1 Immunoreactivity of Rat Liver Detection of UCP1 immunoreactivity was performed using both histological and electrophoretical methods. Immunohistochemistry experiments made on liver slices and using two different antibodies against rat UCP1, easily revealed a positive signal in non-parenchymal cells but never in hepatocytes (fig 1). For the most part, positive cells could be morphologically identified as luminal monocytes or Kupffer cells. Molecular weight and subcellular location of the signal were determined using western blotting of crude extracts and of mitochondria enriched fraction from rat livers (fig 2). The detection of UCP1 immunoreactivity was very weak using total liver protein extracts, but was significantly enhanced in the fraction enriched in mitochondria. This signal appeared to be weak in comparison with the signal obtained in brown fat mitochondria, but the apparent molecular mass of the protein was similar in both tissues. Lack of UCP1 mRNA in Rat Liver To test wether UCP1 was really present in liver, UCP1 mRNA expression was investigated using Northern and RT-PCR experiments. None of these techniques allowed the detection of UCP1 mRNA in hepatocytes or in liver non-parenchymal cells (data not shown). The sensitivity of RT-PCR procedure was tested by detecting the UCP1 mRNA in a mixture of RNA made from 2 ng of brown fat RNA diluted into 500 ng of liver RNA (data not shown). Together, these results clearly indicated that UCP1 is not present in any type of rat liver cells. It was concluded that the signal detected in non-parenchymal cells using antibody against UCP1 identified another protein, putatively related to UCP1, UCP2 being a good candidate. Liver UCP2 mRNA Detection by Northern Experiments To test wether the signal detected in liver using UCP1 antibody was due to a cross reactivity with UCP2, expression of UCP2 mRNA was investigated in different types of liver cell using Northern analysis. As shown in fig 3, UCP2 mRNA was nearly undetectable in analysis of 20 mg purified hepatocyte RNA (integrity of hepatocyte mRNA was verified using the albumin cDNA probe), whereas a weak signal was detected in the non-parenchymal cell fraction of the liver. This fraction contains mainly sinusoid endothelial and Kup-
761
AID
BBRC 6852
/
6930$$1502
06-13-97 14:08:22
bbrcg
AP: BBRC
Vol. 235, No. 3, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Immunohistochemical detection of UCP in liver slices. (1) Control experiment with sheep non-immunserum. (2, 3, 4) Immunostaining using sheep anti-rat UCP1 IgG and anti-sheep Ig G alkaline phosphatase-conjugated antibody. The positive cells are colored in red. Original magnifications: (1 and 2) 1200 (Bar Å 30 mm); (3) 1400 (Bar Å 10mm); (4) 11000 (Bar Å 10mm). This figure is representative of several independant experiments.
ffer cells (14). After isolation by centrifuge elutriation, it was possible to obtain an enriched population of Kupffer cells. It was verified, using the macrophage-specific ED2 antibody that this population contained up to 85% of Kupffer cells (24). As indicated in fig 3, Kupffer cells expressed a very high amount of UCP2 mRNA.
FIG. 2. Western analysis of liver proteins with anti-UCP1 antibody. M, BAT (brown adipose tissue) or liver mitochondrial proteins (10mg); H, liver homogenate proteins (10mg); the analysis of liver proteins from two rats is shown. The position of the band corresponds to a molecular weight of 33 kD.
The specificity of the signal detected in such Northern experiments was confirmed by RT-PCR experiments using a set of UCP2-specific oligonucleotides (fig. 4). Sequencing of RT-PCR products confirmed their absolute homology with UCP2 (data not shown). Immunodetection of Yeast Heterolog UCP2 The presence of UCP2 mRNA and UCP1 immunoreactivity together with the lack of UCP1 mRNA in Kupffer cells seemed to indicate that the antibody initially obtained against rat brown adipose tissue UCP1, was also able to detect UCP2. To test this hypothesis, Western blotting experiments were performed using mitochondrial fraction of Yeast expressing mouse UCP2. As shown in fig 5, using an excess (80 mg) of yeast mitochondrial protein, a positive signal was obtained indicating that the antiserum raised against UCP1 was indeed able to detect UCP2. DISCUSSION This study demonstrates that UCP2 (15), a newly described homologue of the mitochondrial UCP1 of
762
AID
BBRC 6852
/
6930$$1502
06-13-97 14:08:22
bbrcg
AP: BBRC
Vol. 235, No. 3, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 4. RT-PCR amplification of UCP2 mRNA from mouse liver. RT-PCR analysis of UCP2 mRNA was carried out from 200 ng mouse macrophage RNA (Raw 264.7 cell line, ATCC - TIB 71) or from 400 ng mouse liver RNA (Liver), in absence (0) or presence (/) of reverse transcriptase (RT). Ethidium bromide staining of PCR products is shown. The left-hand side lane corresponds to molecular weight markers. The identity of PCR products (901bp) was confirmed by sequencing (data not shown). No PCR product was detected when oligonucleotides specific for UCP1 mRNA were used (data not shown).
FIG. 3. Detection of UCP2 mRNA in cell subtypes of rat liver by Northern analysis. RNA was extracted from liver cells fractions (H, hepatocytes; NPC, non parenchymal cells; K, Kupffer cells) and from BAT. 10mg RNA was used in every lane. (A) Hybridization with mouse UCP2 cDNA probe (1.7 kb mRNA). (B) Hybridization with rat albumin cDNA probe. (C) Cytospin of the Kupffer cell enriched fraction, left: phase contrast, right: immunodetection of Kupffer cells with ED2 antibody.
brown adipocytes, is present in liver and is mainly restricted to monocyte/macrophage cells; in addition it was confirmed that UCP1 is not present in liver. Using purified antibodies specifically reacting against a particular domain of UCP1, as well as RTPCR analysis, we previously established the absence of UCP1 or UCP1 mRNA in rat liver (13). Therefore, the labelling of certain rat liver cells using two different antisera against UCP1 (Figure 1), could not be interpreted as the presence of UCP1 in these cells. The recent cloning of UCP2 which is 59% identical to UCP1 and alters the mitochondrial membrane potentiel (15), prompted us to analyze UCP2 expression in liver, and demonstrate that the signal identified using antisera against UCP1, was UCP2. Such a detection of UCP2 by UCP1-antiserum might result from epitopes shared
by the two proteins. Nevertheless, another hypothesis which is the presence of anti-UCP2 antibodies in UCP1-antisera can be proposed. This hypothesis is supported by the the likely contamination of UCP1 by
FIG. 5. Western blot analysis of recombinant yeasts expressing UCP2. Yeast strains transfected with empty, mouse UCP2, or rat UCP1, pYeDP expression vectors were isolated. Proteins were prepared from total homogenate of yeasts (H) or from mitochondrial fraction (M, M1 and M2 refer to two different mitochondrial preparations). 8 mg protein were used for yeasts expressing UCP1 whereas 80 mg protein were used for other yeast strains. The proteins were incubated in presence of antibodies raised against rat brown adipose tissue UCP1; the same batch of antibodies was used in immunohistochemistry experiments (Figure 1).
763
AID
BBRC 6852
/
6930$$1502
06-13-97 14:08:22
bbrcg
AP: BBRC
Vol. 235, No. 3, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
UCP2, also present in brown fat mitochondria (15) when UCP1 was purified and subsequently used to generate antisera. A particular aspect of the data reported here, is the weakness of UCP2 expression in hepatocytes and its high abundance in Kupffer cells. In a first report, it was proposed that UCP2 mRNA was widely expressed in mammalian tissues (15). However, in this first description of UCP2, no analysis of particular cell type was made. It is noteworthy that, in addition to adipose tissues and muscles, high levels of UCP2 were encountered in all macrophage-containing tissues such as spleen, thymus, bone marrow (15) suggesting that UCP2 is abundant not only in Kupffer cells but also in other types of macrophages and also in monocyte/ macrophage cell line as Raw cells. A systematic study based on immunohistochemistry using specific antiUCP2 antibodies, and in-situ hybridization should be carried out to precisely determine which cells express this mitochondrial protein in other organs than the liver. The precise function and biological importance of UCP2 in Kupffer cells and macrophages has not been determined yet. It has been shown that mouse UCP2 expressed in yeasts can uncouple respiration (15) and therefore limit ADP phosphorylation, which promotes thermogenesis. Liver thermogenesis undoubtedly contributes to resting metabolic rate and thermogenesis. Among cell types present in liver, the cells of the sinusoı¨dal area received little attention in comparison with hepatocytes, and their contribution to energetic metabolism is assumed to be weak. Indeed, mitochondria from Kupffer cells account for a low percentage of liver mitochondria (25). The data presented here may stimulate new research on macrophage mitochondria, their role in inflammatory response as well as new aspects of macrophage biology linked to the UCP2 expression. ACKNOWLEDGMENTS We thank Drs. J. F. Decaux and J. L. Danan for providing us with rat albumin cDNA, and Dr. J. F. Petit for the gift of Raw 264.7 cells. We thank Dr. Luc Pe´nicaud for stimulating discussion and for help in writing the manuscript, Dr. Fre´de´ric Bouillaud for helpful discussion, and Dr. Philippe Djian for reading the manuscript. This research was supported by Centre National de la Recherche Scientifique, Direction des Recherches, Etudes et Techniques, and Association de Recherches sur le Cancer. C.F. was supported by a Ph.D. thesis fellowship of the Direction des Recherches, Etudes et Techniques.
REFERENCES 1. Ricquier, D., Casteilla, L., and Bouillaud, F. (1991) FASEB J. 5, 2237–2242. 2. Himms-Hagen, J. (1990) FASEB J. 4, 2890–2898. 3. Lean, M. E., and James, W. P. (1983) FEBS Lett. 163, 235–240. 4. Ricquier, D., Barlet, J. P., Garel, J. M., Combes-George, M., and Dubois, M. P. (1983) Biochem. J. 210, 859–866. 5. Hansen, E. S., Nedergaard, J., Cannon, B., and Knudsen, J. (1984) Comp. Biochem. Physiol. [b] 79, 441–445. 6. Bouillaud, F., Ricquier, D., Thibault, J., and Weissenbach, J. (1985) Proc. Nat. Acad. Sci. USA 82, 445–448. 7. Jacobsson, A., Stadler, U., Glotzer, M. A., and Kosak, L. P. (1985) J. Biol. Chem. 260, 16250–16254. 8. Ricquier, D., Bouillaud, F., Toumelin, P., Mory, G., Bazin, R., Arch, J., and Pe´nicaud, L. (1986) J. Biol. Chem. 261, 13905– 13910. 9. Boyer, B. B., and Kozak, L. P. (1991) Mol. Cell. Biol. 11, 4147– 4156. 10. Lanni, A., Martins, R., Ambid, L., and Goglia, F. (1990) Comp. Biochem. Physiol. 97 B, 809–813. 11. Goglia, F., Liverini, G., Lanni, A., Iossa, S., and Barletta, A. (1988) Biochem. Biophys. Res. Com. 151, 1241–1249. 12. Shinohara, Y., Shima, A., Kamida, M., and Terada, H. (1991) FEBS Lett. 293, 173–174. 13. Ricquier, D., Raimbault, S., Champigny, O., Miroux, B., and Bouillaud, F. (1992) FEBS Lett. 303, 103–106. 14. Alpini, G., Phillips, J. O., Vroman, B., and LaRusso, N. F. (1994) Hepatology 20, 494–514. 15. Fleury, C., Neverova, M., Collins, C., Raimbault, S., Champigny, O., Levi-Meyrueis, C., Bouillaud, F. S., M. F., Surwit, R. S., Ricquier, D., and Warden, G. H. (1997) Nature Genetics 13, 269– 272. 16. Desautels, M. (1985) Am. J. Physiol. E 249, 99–106. 17. Chomczynski, P., and Sacchi, N. (1987) Anal. Biochem. 162, 156– 159. 18. Viguerie-Bascands, N., Bousquet-Me¨lou, A., Galitzky, J., Larrouy, D., Ricquier, D., Berlan, M., and Casteilla, L. (1995) J. Clin. Endocrinol. Metab. 81, 19. Bouillaud, F., Weissenbach, J., and Ricquier, D. (1986) J. Biol. Chem. 261, 1487–1490. 20. Sargent, T. D., Wu, J. R., Salat-Trepat, J. M., Wallace, R. B., Reyes, A. A., and Bonner, J. (1979) Proc. Natl. Acad. Sci. USA 76, 3256–3260. 21. Casteilla, L., Forest, C., Robelin, J., Ricquier, D., Lombet, A., and Ailhaud, G. (1987) Am. J. Physiol. 252, E627–E636. 22. Seglen, P. O. (1976) Methods Cell Biol. 13, 29–83. 23. Knook, D. L., and Sleyster, E. C. (1976) Exp. Cell Res. 99, 444– 449. 24. Dijkstra, C. D., Do¨pp, E. A., van den Berg, T. K., and Damoiseaux, J. G. M. C. (1994) J. Immunol. Methods 174, 21–23. 25. Blouin, A., Bolender, R. P., and Weibel, E. R. (1977) J. Cell. Biol. 72, 441–455.
764
AID
BBRC 6852
/
6930$$1502
06-13-97 14:08:22
bbrcg
AP: BBRC
235, 760–764 (1997)
RC976852
Kupffer Cells Are a Dominant Site of Uncoupling Protein 2 Expression in Rat Liver Dominique Larrouy,*,1 Patrick Laharrague,* Georges Carrera,† Nathalie Viguerie-Bascands,* Corinne Levi-Meyrueis,‡ Christophe Fleury,‡ Claire Pecqueur,‡ Maryse Nibbelink,* Mireille Andre´,* Louis Casteilla,*,1 and Daniel Ricquier‡,1 *CNRS/UPRESA 5018 and †INSERM CJF 91-07, Institut Louis Bugnard, Universite´ Paul Sabatier, Hoˆpital Rangueil, 31054 Toulouse Cedex, France; and ‡CNRS UPR 9078/CEREMOD, 9, rue Jules Hetzel, 92190 Meudon, France
Received May 15, 1997
The mechanisms underlying thermogenesis in liver are not well understood. They may involve proteins related to the mitochondrial uncoupling protein (UCP1) of brown adipocytes. In this paper, it is demonstrated that UCP1 is not expressed in any liver cell type of rat while UCP2, a recently cloned homologue of UCP1, is expressed at a very high level in Kupffer cells but not in hepatocytes. This high level of expression of UCP2 in Kupffer cells allowed cross immunoreactivity with antibodies directed against UCP1. This cross reactivity was confirmed by the detection of UCP2 with anti-UCP1 antibody, in western blotting analysis of transfected yeasts expressing rat UCP2. The high level expression of UCP2 in Kupffer cells suggests a particular function of UCP2 in macrophages. q 1997 Academic Press
Cold induced thermogenesis in rodents is performed in part by brown adipose tissue. The mitochondrial uncoupling protein UCP1, specifically expressed in brown fat cells, provokes heat dissipation by uncoupling of oxidative phosphorylation from ATP synthesis, during sympathetic activation of the brown fat (1, 2). The exclusive occurence of UCP1 in brown adipose tissue was ascertained by many investigators using immunodetection (3-5) or mRNA analysis (6-9). A contribution of liver to cold-induced thermogenesis, was also demonstrated. Nevertheless, although it was proposed that an uncoupling of respiration of liver mitochondria might exist in cold-exposed animals, the mechanism of such an uncoupling remains unknown (10, 11). The detection of UCP1 mRNA in liver of newborn rats, or 1 Corresponding authors. (D.R.) Fax: /33 1 45 07 58 90. E-mail: [email protected]. (L.C.) Fax: /33 5 62 17 09 05. E-mail: casteil@ rangueil.inserm.fr. Abbreviation: UCP2, uncoupling protein 2.
0006-291X/97 $25.00
adult rats exposed to the cold (12) was in agreement with this hypothesis, but was later on described as artefactual (13). Until now, the search for UCP1 and/or UCP1 mRNA in tissues other than brown adipose tissue, was performed mainly on homogenized tissues. Such experiments could give rise to misleading results if UCP1 was expressed only in a particular and sparse cell type present in a tissue. Indeed, liver is a combination of at least 12 different cell types (14). Hence, the initial aim of this study was to definitively establish if UCP1 is expressed in certain liver cell types using a combination of immunological, immuno-histochemical, and molecular methods. Immunological methods used in the present study, led to contradictory results, suggesting that UCP1 antibodies cross-reacted with an other mitochondrial protein expressed in non-parenchymal cells present in liver. Recently, a new protein belonging to the family of mitochondrial anion carriers was cloned from a mouse muscle cDNA library (15). This protein, named uncoupling protein 2 or UCP2, seems to have properties similar to UCP1 and is putatively involved in heat production, since it was shown to be able to uncouple the mitochondrial respiration from ATP synthesis. Moreover, the wide tissular expression of UCP2 agrees with a role in energy expenditure. We report here that the UCP1 immunoreactivity of rat liver revealed the presence of very high amount of UCP2 in non parenchymal cells and in particular in Kupffer cells. MATERIALS AND METHODS Animals. Male Wistar rats were housed at 257C with free access to food (65% carbohydrate, 11% fat and 24% protein w/w; UAR, Villemoisson France) and water. Antibodies. Sheep and rabbit polyclonal antibodies against rat UCP were previously described (4, 16). Donkey anti-sheep IgG alkaline phosphatase-conjugated was purchased from Sigma Chemical (St. Louis, MO). Pig anti-rabbit IgG alkaline phosphatase-conjugated
760
Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
AID
BBRC 6852
/
6930$$1501
06-13-97 14:08:22
bbrcg
AP: BBRC
Vol. 235, No. 3, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
was obtained from Dako corporation (Carpinteria, CA). Monoclonal antibody against macrophage (ED2) was from Serotec (Oxford). Antimouse IgG FITC conjugated was from Sigma Chemical Co. Preparation and analysis of RNA. Total RNA were prepared according to the single step method of Chomczynski and Sacchi (17). Northern blot analysis was performed as previously described (18). Rat UCP1 (19), mouse UCP2 (15) and rat albumin (20) cDNA probes were labelled with [a32P]dCTP using a megaprime labelling system kit (Boehringer, Mannheim Germany). Blots were analyzed using a phosphorimager (Molecular Dynamics). RT-PCR amplification of UCP1 mRNA was carried out as previously described (13, 18). RTPCR amplification of UCP2 mRNA was made using 5*-AATCGAATTCATGGTTGGTTTCAAGGCC-3* sense primer, and 5*-GGAGGAGCTCAGAAAGGTGCCTCCCGA-3* antisense primer. Moloney murine-leukemia-virus RT was obtained from Life Technology (Gaithersburg, MD), ribonuclease inhibitor from Pharmacia (Uppsala, Sweden) and Taq DNA polymerase from Appligene (Illkirch, France). Immunohistochemistry. Livers were fixed in Dubosq’Brasil immediately after removal, dehydrated in ethanol and paraffin embedded. 5 mm sections were deparaffinized, rehydrated and incubated in medium containing purified sheep IgG (1:2000 dilution) raised against rat UCP1 (4, 16) during 1 hour at room temperature, rinsed and incubated in presence of rabbit antisheep IgG (1:100 dilution) coupled to alkaline phosphatase. Alkaline phosphatase activity was revealed using Sigma Fast Red TR/Naphtol AS-MX tablets as substrate and counterstained with Harris’ hematoxylin. Control experiments were performed using purified sheep IgG (Sigma) and yielded no staining. Western blot analysis. Mitochondrial fractions were prepared by differential centrifugation of liver homogenate as described by Casteilla et al (21). 5-30mg of mitochondrial fraction or total homogenate liver proteins were electrophoresed in 10 % polyacrylamide SDS-gels, transferred onto nitrocellulose and incubated with sheep antiserum raised against rat UCP1 (4) (1:2,000 dilution) and rabbit IgG conjugated to peroxydase. Peroxydase activity was revealed using 4-chloro-1-naphtol and hydrogen peroxyde as substrate. Western analysis of yeast proteins was carried out using the same UCP1antibodies at a 1:20,000 dilution and the Renaissance system (Dupont de Nemours, Les Ulis). Expression in yeast. The expression of mouse UCP2 in diploid yeast (S. cerevisiae) strain W303 using pYeDP-UCP2 plasmid was previously described by Fleury et al. (15). Mitochondria were isolated from this yeast strain, from the same S. cerevisiae strain transfected by an empty pYeDP expression vector and from a yeast strain transfected by a pYeDP-UCP1 vector (21). Isolation of cells. Isolated hepatocytes and total non-parenchymal cells (NPCs) were obtained from rat liver using the two-step perfusion collagenase method (22). The resulting cell suspension was centrifuged at 50 g for 30 sec to separate the pellet of hepatocytes from the supernatant containing all the NPCs. For the studies of hepatocytes and Kupffer cells, the liver was first perfused by the two-step perfusion method with 0.03 % collagenase A (Boehringer Mannheim Gmbh, Mannheim, Germany) during 10 min. Then, the liver was excised and minced into 1-2 mm cubes using a razor blade. These fragments were incubated in 100 ml of Gey’s balanced salt solution (GBSS) containing 0.1 % pronase E (MerckCle´venot S.A., Nogent-sur-Marne, France) and 0.5 mg of deoxyribonuclease (Sigma Chemical Co., St. Louis, USA) for 40 min at 377 C and pH 7.4 with continuous stirring. The cells were filtered through nylon gauze, and centrifuged at 50 g for 30 sec to separate the pellet of hepatocytes from the supernatant containing the NPCs. The NPCs suspension was centrifuged for 10 min at 550 g and the resulting pellet washed two times using resuspension in 20 ml of GBSS and centrifugation. The Kupffer cells were separated from the other NPCs first by centrifugation in presence of Metrizamide (2-(3-acetamido-5-N-methylacetamido-2,4,6-triiodobenzamido)-2-deoxy-D-
glucose, Nycomed Pharma AS, Oslo, Norway), followed by centrifugal elutriation (23). The percentage of Kupffer cells in the resulting suspension was determined on a cytospin preparation by immunochemistry using ED2 (Serotec, Oxford, England) as primary antibody and an anti-mouse IgG FITC conjugate (Sigma Chemical Co) as secondary antibody.
RESULTS UCP1 Immunoreactivity of Rat Liver Detection of UCP1 immunoreactivity was performed using both histological and electrophoretical methods. Immunohistochemistry experiments made on liver slices and using two different antibodies against rat UCP1, easily revealed a positive signal in non-parenchymal cells but never in hepatocytes (fig 1). For the most part, positive cells could be morphologically identified as luminal monocytes or Kupffer cells. Molecular weight and subcellular location of the signal were determined using western blotting of crude extracts and of mitochondria enriched fraction from rat livers (fig 2). The detection of UCP1 immunoreactivity was very weak using total liver protein extracts, but was significantly enhanced in the fraction enriched in mitochondria. This signal appeared to be weak in comparison with the signal obtained in brown fat mitochondria, but the apparent molecular mass of the protein was similar in both tissues. Lack of UCP1 mRNA in Rat Liver To test wether UCP1 was really present in liver, UCP1 mRNA expression was investigated using Northern and RT-PCR experiments. None of these techniques allowed the detection of UCP1 mRNA in hepatocytes or in liver non-parenchymal cells (data not shown). The sensitivity of RT-PCR procedure was tested by detecting the UCP1 mRNA in a mixture of RNA made from 2 ng of brown fat RNA diluted into 500 ng of liver RNA (data not shown). Together, these results clearly indicated that UCP1 is not present in any type of rat liver cells. It was concluded that the signal detected in non-parenchymal cells using antibody against UCP1 identified another protein, putatively related to UCP1, UCP2 being a good candidate. Liver UCP2 mRNA Detection by Northern Experiments To test wether the signal detected in liver using UCP1 antibody was due to a cross reactivity with UCP2, expression of UCP2 mRNA was investigated in different types of liver cell using Northern analysis. As shown in fig 3, UCP2 mRNA was nearly undetectable in analysis of 20 mg purified hepatocyte RNA (integrity of hepatocyte mRNA was verified using the albumin cDNA probe), whereas a weak signal was detected in the non-parenchymal cell fraction of the liver. This fraction contains mainly sinusoid endothelial and Kup-
761
AID
BBRC 6852
/
6930$$1502
06-13-97 14:08:22
bbrcg
AP: BBRC
Vol. 235, No. 3, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Immunohistochemical detection of UCP in liver slices. (1) Control experiment with sheep non-immunserum. (2, 3, 4) Immunostaining using sheep anti-rat UCP1 IgG and anti-sheep Ig G alkaline phosphatase-conjugated antibody. The positive cells are colored in red. Original magnifications: (1 and 2) 1200 (Bar Å 30 mm); (3) 1400 (Bar Å 10mm); (4) 11000 (Bar Å 10mm). This figure is representative of several independant experiments.
ffer cells (14). After isolation by centrifuge elutriation, it was possible to obtain an enriched population of Kupffer cells. It was verified, using the macrophage-specific ED2 antibody that this population contained up to 85% of Kupffer cells (24). As indicated in fig 3, Kupffer cells expressed a very high amount of UCP2 mRNA.
FIG. 2. Western analysis of liver proteins with anti-UCP1 antibody. M, BAT (brown adipose tissue) or liver mitochondrial proteins (10mg); H, liver homogenate proteins (10mg); the analysis of liver proteins from two rats is shown. The position of the band corresponds to a molecular weight of 33 kD.
The specificity of the signal detected in such Northern experiments was confirmed by RT-PCR experiments using a set of UCP2-specific oligonucleotides (fig. 4). Sequencing of RT-PCR products confirmed their absolute homology with UCP2 (data not shown). Immunodetection of Yeast Heterolog UCP2 The presence of UCP2 mRNA and UCP1 immunoreactivity together with the lack of UCP1 mRNA in Kupffer cells seemed to indicate that the antibody initially obtained against rat brown adipose tissue UCP1, was also able to detect UCP2. To test this hypothesis, Western blotting experiments were performed using mitochondrial fraction of Yeast expressing mouse UCP2. As shown in fig 5, using an excess (80 mg) of yeast mitochondrial protein, a positive signal was obtained indicating that the antiserum raised against UCP1 was indeed able to detect UCP2. DISCUSSION This study demonstrates that UCP2 (15), a newly described homologue of the mitochondrial UCP1 of
762
AID
BBRC 6852
/
6930$$1502
06-13-97 14:08:22
bbrcg
AP: BBRC
Vol. 235, No. 3, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 4. RT-PCR amplification of UCP2 mRNA from mouse liver. RT-PCR analysis of UCP2 mRNA was carried out from 200 ng mouse macrophage RNA (Raw 264.7 cell line, ATCC - TIB 71) or from 400 ng mouse liver RNA (Liver), in absence (0) or presence (/) of reverse transcriptase (RT). Ethidium bromide staining of PCR products is shown. The left-hand side lane corresponds to molecular weight markers. The identity of PCR products (901bp) was confirmed by sequencing (data not shown). No PCR product was detected when oligonucleotides specific for UCP1 mRNA were used (data not shown).
FIG. 3. Detection of UCP2 mRNA in cell subtypes of rat liver by Northern analysis. RNA was extracted from liver cells fractions (H, hepatocytes; NPC, non parenchymal cells; K, Kupffer cells) and from BAT. 10mg RNA was used in every lane. (A) Hybridization with mouse UCP2 cDNA probe (1.7 kb mRNA). (B) Hybridization with rat albumin cDNA probe. (C) Cytospin of the Kupffer cell enriched fraction, left: phase contrast, right: immunodetection of Kupffer cells with ED2 antibody.
brown adipocytes, is present in liver and is mainly restricted to monocyte/macrophage cells; in addition it was confirmed that UCP1 is not present in liver. Using purified antibodies specifically reacting against a particular domain of UCP1, as well as RTPCR analysis, we previously established the absence of UCP1 or UCP1 mRNA in rat liver (13). Therefore, the labelling of certain rat liver cells using two different antisera against UCP1 (Figure 1), could not be interpreted as the presence of UCP1 in these cells. The recent cloning of UCP2 which is 59% identical to UCP1 and alters the mitochondrial membrane potentiel (15), prompted us to analyze UCP2 expression in liver, and demonstrate that the signal identified using antisera against UCP1, was UCP2. Such a detection of UCP2 by UCP1-antiserum might result from epitopes shared
by the two proteins. Nevertheless, another hypothesis which is the presence of anti-UCP2 antibodies in UCP1-antisera can be proposed. This hypothesis is supported by the the likely contamination of UCP1 by
FIG. 5. Western blot analysis of recombinant yeasts expressing UCP2. Yeast strains transfected with empty, mouse UCP2, or rat UCP1, pYeDP expression vectors were isolated. Proteins were prepared from total homogenate of yeasts (H) or from mitochondrial fraction (M, M1 and M2 refer to two different mitochondrial preparations). 8 mg protein were used for yeasts expressing UCP1 whereas 80 mg protein were used for other yeast strains. The proteins were incubated in presence of antibodies raised against rat brown adipose tissue UCP1; the same batch of antibodies was used in immunohistochemistry experiments (Figure 1).
763
AID
BBRC 6852
/
6930$$1502
06-13-97 14:08:22
bbrcg
AP: BBRC
Vol. 235, No. 3, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
UCP2, also present in brown fat mitochondria (15) when UCP1 was purified and subsequently used to generate antisera. A particular aspect of the data reported here, is the weakness of UCP2 expression in hepatocytes and its high abundance in Kupffer cells. In a first report, it was proposed that UCP2 mRNA was widely expressed in mammalian tissues (15). However, in this first description of UCP2, no analysis of particular cell type was made. It is noteworthy that, in addition to adipose tissues and muscles, high levels of UCP2 were encountered in all macrophage-containing tissues such as spleen, thymus, bone marrow (15) suggesting that UCP2 is abundant not only in Kupffer cells but also in other types of macrophages and also in monocyte/ macrophage cell line as Raw cells. A systematic study based on immunohistochemistry using specific antiUCP2 antibodies, and in-situ hybridization should be carried out to precisely determine which cells express this mitochondrial protein in other organs than the liver. The precise function and biological importance of UCP2 in Kupffer cells and macrophages has not been determined yet. It has been shown that mouse UCP2 expressed in yeasts can uncouple respiration (15) and therefore limit ADP phosphorylation, which promotes thermogenesis. Liver thermogenesis undoubtedly contributes to resting metabolic rate and thermogenesis. Among cell types present in liver, the cells of the sinusoı¨dal area received little attention in comparison with hepatocytes, and their contribution to energetic metabolism is assumed to be weak. Indeed, mitochondria from Kupffer cells account for a low percentage of liver mitochondria (25). The data presented here may stimulate new research on macrophage mitochondria, their role in inflammatory response as well as new aspects of macrophage biology linked to the UCP2 expression. ACKNOWLEDGMENTS We thank Drs. J. F. Decaux and J. L. Danan for providing us with rat albumin cDNA, and Dr. J. F. Petit for the gift of Raw 264.7 cells. We thank Dr. Luc Pe´nicaud for stimulating discussion and for help in writing the manuscript, Dr. Fre´de´ric Bouillaud for helpful discussion, and Dr. Philippe Djian for reading the manuscript. This research was supported by Centre National de la Recherche Scientifique, Direction des Recherches, Etudes et Techniques, and Association de Recherches sur le Cancer. C.F. was supported by a Ph.D. thesis fellowship of the Direction des Recherches, Etudes et Techniques.
REFERENCES 1. Ricquier, D., Casteilla, L., and Bouillaud, F. (1991) FASEB J. 5, 2237–2242. 2. Himms-Hagen, J. (1990) FASEB J. 4, 2890–2898. 3. Lean, M. E., and James, W. P. (1983) FEBS Lett. 163, 235–240. 4. Ricquier, D., Barlet, J. P., Garel, J. M., Combes-George, M., and Dubois, M. P. (1983) Biochem. J. 210, 859–866. 5. Hansen, E. S., Nedergaard, J., Cannon, B., and Knudsen, J. (1984) Comp. Biochem. Physiol. [b] 79, 441–445. 6. Bouillaud, F., Ricquier, D., Thibault, J., and Weissenbach, J. (1985) Proc. Nat. Acad. Sci. USA 82, 445–448. 7. Jacobsson, A., Stadler, U., Glotzer, M. A., and Kosak, L. P. (1985) J. Biol. Chem. 260, 16250–16254. 8. Ricquier, D., Bouillaud, F., Toumelin, P., Mory, G., Bazin, R., Arch, J., and Pe´nicaud, L. (1986) J. Biol. Chem. 261, 13905– 13910. 9. Boyer, B. B., and Kozak, L. P. (1991) Mol. Cell. Biol. 11, 4147– 4156. 10. Lanni, A., Martins, R., Ambid, L., and Goglia, F. (1990) Comp. Biochem. Physiol. 97 B, 809–813. 11. Goglia, F., Liverini, G., Lanni, A., Iossa, S., and Barletta, A. (1988) Biochem. Biophys. Res. Com. 151, 1241–1249. 12. Shinohara, Y., Shima, A., Kamida, M., and Terada, H. (1991) FEBS Lett. 293, 173–174. 13. Ricquier, D., Raimbault, S., Champigny, O., Miroux, B., and Bouillaud, F. (1992) FEBS Lett. 303, 103–106. 14. Alpini, G., Phillips, J. O., Vroman, B., and LaRusso, N. F. (1994) Hepatology 20, 494–514. 15. Fleury, C., Neverova, M., Collins, C., Raimbault, S., Champigny, O., Levi-Meyrueis, C., Bouillaud, F. S., M. F., Surwit, R. S., Ricquier, D., and Warden, G. H. (1997) Nature Genetics 13, 269– 272. 16. Desautels, M. (1985) Am. J. Physiol. E 249, 99–106. 17. Chomczynski, P., and Sacchi, N. (1987) Anal. Biochem. 162, 156– 159. 18. Viguerie-Bascands, N., Bousquet-Me¨lou, A., Galitzky, J., Larrouy, D., Ricquier, D., Berlan, M., and Casteilla, L. (1995) J. Clin. Endocrinol. Metab. 81, 19. Bouillaud, F., Weissenbach, J., and Ricquier, D. (1986) J. Biol. Chem. 261, 1487–1490. 20. Sargent, T. D., Wu, J. R., Salat-Trepat, J. M., Wallace, R. B., Reyes, A. A., and Bonner, J. (1979) Proc. Natl. Acad. Sci. USA 76, 3256–3260. 21. Casteilla, L., Forest, C., Robelin, J., Ricquier, D., Lombet, A., and Ailhaud, G. (1987) Am. J. Physiol. 252, E627–E636. 22. Seglen, P. O. (1976) Methods Cell Biol. 13, 29–83. 23. Knook, D. L., and Sleyster, E. C. (1976) Exp. Cell Res. 99, 444– 449. 24. Dijkstra, C. D., Do¨pp, E. A., van den Berg, T. K., and Damoiseaux, J. G. M. C. (1994) J. Immunol. Methods 174, 21–23. 25. Blouin, A., Bolender, R. P., and Weibel, E. R. (1977) J. Cell. Biol. 72, 441–455.
764
AID
BBRC 6852
/
6930$$1502
06-13-97 14:08:22
bbrcg
AP: BBRC