Effects of Genetic Background on Thermoregulation and Fatty Acid-induced Uncoupling of Mitochondria in UCP1-deficient Mice
2001; Elsevier BV; Volume: 276; Issue: 15 Linguagem: Inglês
10.1074/jbc.m100466200
ISSN1083-351X
AutoresWolfgang Hofmann, Xiaotuan Liu, Christie M. Bearden, Mary‐Ellen Harper, Leslie P. Kozak,
Tópico(s)Muscle metabolism and nutrition
ResumoAn interaction between free fatty acids and UCP1 (uncoupling protein-1) leading to de-energization of mitochondria was assumed to be a key event for triggering heat production in brown fat. Recently, Matthiaset al., finding indistinguishable de-energization of isolated brown fat mitochondria by fatty acids in UCP1-deficient mice and control mice, challenged this assumption (Matthias, A., Jacobsson, A., Cannon, B., and Nedergaard, J. (1999) J. Biol. Chem. 274, 28150–28160). Since their results were obtained using UCP1-deficient and control mice on an undefined genetic background, we wanted to determine unambiguously the phenotype of UCP1 deficiency with the targeted Ucp1 allele on congenic C57BL/6J and 129/SvImJ backgrounds. UCP1-deficient congenic mice have a very pronounced cold-sensitive phenotype; however, deficient mice on the F1 hybrid background were resistant to cold. We propose that heterosis provides a mechanism to compensate for UCP1 deficiency. Contrary to the results of Matthias et al., we found a significant loss of fatty acid-induced de-energization, as reflected by membrane potential and oxygen consumption, in brown fat mitochondria from UCP1-deficient mice. Unlike cold sensitivity, fatty acid-induced uncoupling of mitochondria was independent of the genetic background of UCP1-deficient mice. We propose that intracellular free fatty acids directly regulate uncoupling activity of UCP1 in a manner consistent with models described in the literature. An interaction between free fatty acids and UCP1 (uncoupling protein-1) leading to de-energization of mitochondria was assumed to be a key event for triggering heat production in brown fat. Recently, Matthiaset al., finding indistinguishable de-energization of isolated brown fat mitochondria by fatty acids in UCP1-deficient mice and control mice, challenged this assumption (Matthias, A., Jacobsson, A., Cannon, B., and Nedergaard, J. (1999) J. Biol. Chem. 274, 28150–28160). Since their results were obtained using UCP1-deficient and control mice on an undefined genetic background, we wanted to determine unambiguously the phenotype of UCP1 deficiency with the targeted Ucp1 allele on congenic C57BL/6J and 129/SvImJ backgrounds. UCP1-deficient congenic mice have a very pronounced cold-sensitive phenotype; however, deficient mice on the F1 hybrid background were resistant to cold. We propose that heterosis provides a mechanism to compensate for UCP1 deficiency. Contrary to the results of Matthias et al., we found a significant loss of fatty acid-induced de-energization, as reflected by membrane potential and oxygen consumption, in brown fat mitochondria from UCP1-deficient mice. Unlike cold sensitivity, fatty acid-induced uncoupling of mitochondria was independent of the genetic background of UCP1-deficient mice. We propose that intracellular free fatty acids directly regulate uncoupling activity of UCP1 in a manner consistent with models described in the literature. Brown adipose tissue (BAT)1 plays an important role in heat production and is considered to contribute to energy balance (5Smith R.E. Horwitz B.A. Physiol. Rev. 1969; 49: 330-425Crossref PubMed Scopus (518) Google Scholar, 6Rothwell N.J. Stock M.J. Nature. 1979; 281: 31-35Crossref PubMed Scopus (1198) Google Scholar). Stimulation by the sympathetic nervous system causes an up-regulation in the metabolic rate of BAT that is reflected in an increase of heat production (7Foster D.O. Frydman M.L. Can. J. Physiol. Pharmacol. 1978; 56: 110-122Crossref PubMed Scopus (493) Google Scholar, 8Foster D.O. Frydman M.L. Can. J. Physiol. Pharmacol. 1979; 57: 257-270Crossref PubMed Scopus (394) Google Scholar). The ability to generate heat is attributed to the high number of mitochondria containing UCP1 (uncoupling protein-1) (9Nicholls D.G. Locke R.M. Physiol. Rev. 1984; 64: 1-64Crossref PubMed Scopus (1344) Google Scholar). This transmembrane protein is thought to be an ion carrier that uncouples respiration from ATP synthesis, allowing mitochondria to produce heat. The importance of UCP1 for thermogenesis was proven by the observation that UCP1-deficient mice are cold-sensitive (10Enerback S. Jacobsson A. Simpson E.M. Guerra C. Yamashita H. Harper M.-E. Kozak L.P. Nature. 1997; 387: 90-94Crossref PubMed Scopus (1082) Google Scholar). However, the fact that some UCP1-deficient mice are resistant to cold and that adiposity is not increased led us to postulate that additional thermogenic mechanisms can compensate for UCP1 deficiency. In this respect, UCP1-deficient mice can help not only to understand the mechanisms that control BAT thermogenesis, but can also be used to identify alternative pathways for heat production. Although it is known that the sympathetic nervous system controls heat production of brown adipocytes (11Prusiner S.B. Cannon B. Ching T.M. Lindberg O. Eur. J. Biochem. 1968; 7: 51-57Crossref PubMed Scopus (62) Google Scholar, 12Thomas S.A. Palmiter R.D. Nature. 1997; 387: 94-97Crossref PubMed Scopus (221) Google Scholar), the intracellular signaling pathway remains unclear. It has been proposed that free fatty acids (FFAs), released by the action of hormone-sensitive lipase, serve both as an energy substrate and as an activator of the proton carrier function of UCP1, thereby triggering heat production. It has also been shown that FFAs can increase respiration of isolated brown adipocytes (13Bukowiecki L.J. Follea N. Lupien J. Paradis A. J. Biol. Chem. 1981; 256: 12840-12848Abstract Full Text PDF PubMed Google Scholar) and that isolated BAT mitochondria are more prone to de-energization by FFAs compared with liver mitochondria (14Wojtczak L. Schonfeld P. Biochim. Biophys. Acta. 1993; 1183: 41-57Crossref PubMed Scopus (317) Google Scholar). Finally, experiments with reconstituted UCP1 in proteoliposomes (4Winkler E. Klingenberg M. J. Biol. Chem. 1994; 269: 2508-2515Abstract Full Text PDF PubMed Google Scholar) and with ectopic expression of UCP1 in yeast (15Gonzalez-Barroso M.M. Fleury C. Arechaga I. Zaragoza P. Levi-Meyrueis C. Raimbault S. Ricquier D. Bouillaud F. Rial E. Eur. J. Biochem. 1996; 239: 350-445Crossref Scopus (56) Google Scholar) demonstrate that FFAs are necessary for UCP1 to act as a proton carrier. The model for the activation of UCP1 by FFAs has recently been challenged by Matthias et al. (1Matthias A. Jacobsson A. Cannon B. Nedergaard J. J. Biol. Chem. 1999; 274: 28150-28160Abstract Full Text Full Text PDF PubMed Scopus (102) Google Scholar) by the analysis of the bioenergetics of isolated BAT mitochondria from UCP1-deficient mice. In agreement with the findings of Monemdjou et al. (16Monemdjou S. Kozak L.P. Harper M.-E. Am. J. Physiol. 1999; 276: E1073-E1082PubMed Google Scholar), Matthias et al. found a high mitochondrial membrane potential in UCP1-deficient mice that was independent of GDP addition. This confirmed that GDP inhibited UCP1-dependent proton flux. Matthias et al. also presented provocative findings that the ability of FFAs to uncouple BAT does not depend on UCP1. Their data suggested that wild-type and UCP1-deficient BAT mitochondria had the same ratio of de-energization after the addition of the same amounts of fatty acid. In a second study, Matthias et al.(17Matthias A. Ohlson K.B. Fredriksson J.M. Jacobsson A. Nedergaard J. Cannon B. J. Biol. Chem. 2000; 275: 25073-25081Abstract Full Text Full Text PDF PubMed Scopus (281) Google Scholar) showed that the stimulation of oxygen consumption by noradrenaline and fatty acid in isolated brown adipocytes depended on Ucp1expression. Matthias et al. concluded that the uncoupling effect of FFA is not directly mediated by UCP1; rather an alternative activator such as a metabolite of FFAs enhances the proton transport activity of UCP1. Considering the long history of biochemical studies pointing toward an interaction between UCP1 and FFAs in the activation of proton flux, we reevaluated UCP1-independent FFA uncoupling in UCP1-deficient mice that have defined genetic backgrounds. Our results show that FFA-induced uncoupling of respiration in isolated BAT mitochondria is strongly reduced in UCP1-deficient mice. The basis of the difference between our results and those of the Matthias et al. (1Matthias A. Jacobsson A. Cannon B. Nedergaard J. J. Biol. Chem. 1999; 274: 28150-28160Abstract Full Text Full Text PDF PubMed Scopus (102) Google Scholar, 17Matthias A. Ohlson K.B. Fredriksson J.M. Jacobsson A. Nedergaard J. Cannon B. J. Biol. Chem. 2000; 275: 25073-25081Abstract Full Text Full Text PDF PubMed Scopus (281) Google Scholar) is not clear since FFA-induced uncoupling is reduced in UCP1-deficient mice irrespective of the genetic background of the targeted mice. On the other hand, differences in genetic background are extremely important for thermoregulation in UCP1-deficient mice since targeted mice on congenic backgrounds are extremely sensitive to cold, whereas mice on a F1 hybrid background are resistant to cold. Clearly some, as yet to be identified, thermogenic mechanisms allow the latter mice to resist cold. Ucp tm1 knockout mice on a mixed 129/SvPas and C57BL/6J background were backcrossed to 129/SvImJ and C57BL/6J for 10 generations. At the 10th backcross generation, each line was typed with 135 microsatellite markers to establish its genetic purity. These mice were intercrossed to obtain mice homozygous for theUcp1 tm1 allele. The absence of Ucp1mRNA was confirmed by real-time reverse transcription-polymerase chain reaction using a Taqman® probe. Littermates with two wild-type Ucp1 alleles were used as control mice. 129/SvImJ × C57BL/6J hybrid mice were obtained by crossing homozygous Ucp1 tm1 knockout mice from the two congenic lines. All mice were kept in a 12-h light cycle in 27 °C and fed a 4.5% fat chow diet. Animals of both genders between the ages of 8 and 12 weeks were used. Cold sensitivity was estimated using the Vital View Data Acquisition System® (Mini Mitter Co., Inc., Sunriver, OR). Temperature probes were implanted in the peritoneal cavity 1 week prior to the experiment. Mice were single-caged and exposed to 4 °C on ER-4000 receiver units. The body temperatures were monitored online with a sampling rate of 12/h. Cold exposure was maintained for 18 h, and all hypothermic mice reaching a body core temperature of <30 °C were removed from the cold environment. Mitochondria were isolated in a manner similar to that described by Cannon and Lindberg (18Cannon B. Lindberg O. Methods Enzymol. 1979; 55: 65-78Crossref PubMed Scopus (190) Google Scholar). Each step was performed on ice or at 4 °C. Mice were killed by cervical dislocation, and interscapular BAT was removed. The fat pads were homogenized immediately in homogenization medium (250 mm sucrose, 1 mm HEPES (pH 7.2), and 0.2 mm EGTA) in the presence of 0.5% BSA in a Potter-Elvehjem homogenizer. The suspension was centrifuged at 900 ×g for 10 min and filtered through a 100-μm nylon mesh filter. The mitochondrial pellet was obtained by centrifugation at 6000 × g for 16 min and resuspended in homogenization medium with 0.1% albumin. The mitochondria were then centrifuged at 6000 × g and resuspended in high potassium medium (100 mm KCl, 2 mm HEPES (pH 7.2), and 1 mm EGTA). Resuspension directly in respiration buffer (125 mm sucrose, 4 mm HEPES (pH 7.2), 1 mm EGTA, 2 mm MgCl2, 4 mm Tris phosphate, 20 mm KCl, 0.1% BSA, 5 μm rotenone, and 6 μg/ml oligomycin) caused a higher fluorescent quench, but did not alter the final results. Mitochondrial protein content was determined by the Lowry method (35Lowry D.H. Rosebrough N.J. Farr A.L. Randall R.J. J. Biol. Chem. 1951; 193: 265-275Abstract Full Text PDF PubMed Google Scholar), and the suspensions were kept on ice for no longer than 4 h for the experiments. Biochemical properties of mitochondria were characterized by determination of activities for mitochondrial glycerol-3-phosphate dehydrogenase (19Gardner R.S. Anal. Biochem. 1974; 59: 272-276Crossref PubMed Scopus (54) Google Scholar) and cytochrome c oxidase (20Capaldi R.A. Marusich M.F. Taanman J.W. Methods Enzymol. 1995; 260: 117-132Crossref PubMed Scopus (125) Google Scholar). All measurements were performed at 37 °C with constant stirring. A Clark-type oxygen electrode from Hansatech was used for monitoring oxygen consumption. Mitochondria were incubated in respiration buffer at a concentration of 0.05 mg/ml protein. Glycerol phosphate was added to a final concentration of 5 mm. The UCP1-mediated proton leak across the inner mitochondrial membrane was inhibited by the addition of GDP (1 mm final concentration). De-energization of mitochondria was achieved by the addition of FFA. A concentration of 1 nm FCCP was sufficient to disrupt the electrochemical gradient and to observe maximum mitochondrial respiration. Fluorescent quench of the cationic dye rhodamine 123 was used for estimation of protonmotive force (21Nedergaard J. Eur. J. Biochem. 1983; 133: 185-191Crossref PubMed Scopus (33) Google Scholar, 22Emaus R.K. Grunwald R. Lemasters J.J. Biochim. Biophys. Acta. 1986; 850: 436-448Crossref PubMed Scopus (723) Google Scholar). The measurements were performed under the same conditions as the oxygen measurements, except that 0.13 μm fluorescent dye was added. Fluorescence was monitored in a PerkinElmer Life Sciences LS 50B spectrofluorometer. According to our spectral data, we chose wavelengths of 500 and 525 nm for excitation and emission, respectively. The fluorescence quench was converted to membrane potential using the Nernst equation. The conversion was calibrated using different potassium concentrations of the incubation medium and generation of diffusion potential by the addition of 29 ng/ml valinomycin, a potassium ionophore. Liver mitochondria were treated identically to BAT mitochondria, except that succinate at 10 mm was used as a substrate. RNA was isolated using TRI®-Reagent and subjected to quantitative real-time reverse transcription-polymerase chain reaction using the ABI Prism 7700 sequence detection system (23Koza R.A. Hohmann S.M. Guerra C. Rossmeisl M. Kozak L.P. J. Biol. Chem. 2000; 275: 34486-34492Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar). Polymerase chain reaction primers and Taqman® probes were designed corresponding to cDNA sequences of Ucp1, Ucp2, and Ucp3(GenBankTM/EBI accession numbers U63419, AB012159, andAF032902, respectively). All chemicals were purchased from Sigma. GDP anddl-glycerol 3-phosphate were ordered as Tris and di(monocyclohexylammonium) salts, respectively. Only charcoal-treated fatty acid-free BSA was used. Lauric acid was used as a Tris salt in water; all other fatty acids were dissolved in ethanol. UCP1-deficient mice are sensitive to cold, showing the pivotal function of UCP1 for thermogenesis (10Enerback S. Jacobsson A. Simpson E.M. Guerra C. Yamashita H. Harper M.-E. Kozak L.P. Nature. 1997; 387: 90-94Crossref PubMed Scopus (1082) Google Scholar). However, in our initial description of the mice, we reported that ∼15% of the individuals were resistant to cold, suggesting that variation in genetic background could alter cold sensitivity. To address this issue, we established UCP1-deficient mice on inbred 129/SvImJ and C57BL/6J backgrounds. Incipient mice developed normally with regard to growth and development. As listed in Table I for the 129/SvImJ background, UCP1-deficient mice showed an increased BAT depot weight, a comparable content of mitochondrial protein, decreased cytochrome c oxidase activity, and increased mitochondrial glycerol-3-phosphate dehydrogenase activity compared with wild-type mice.Table IParameters of interscapular brown fat pad and isolated mitochondria in congenic 129/SvImJ miceGenotypeMouse weight1-aValues for male mice only.BAT weight1-aValues for male mice only.COXmGPDHgmgμm/min/mgμm/min/mgUcp1(+/+) 129/SvImJ23.0 ± 1.553 ± 76.3 ± 0.62.4 ± 0.6Ucp1(−/−) 129/SvImJ22.5 ± 1.087 ± 131-bp < 0.01 compared with wild-type mitochondria.4.7 ± 0.71-bp < 0.01 compared with wild-type mitochondria.3.6 ± 1.41-bp < 0.01 compared with wild-type mitochondria.Cytochrome c oxidase (COX) and mitochondrial glycerol-3-phosphate dehydrogenase (mGPDH) assays were performed as described under "Experimental Procedures," specific activities are expressed per milligram of protein. n ≥ 14, animals of both genders and age of 65 ± 5 days. Values are expressed as means ± S.D.1-a Values for male mice only.1-b p < 0.01 compared with wild-type mitochondria. Open table in a new tab Cytochrome c oxidase (COX) and mitochondrial glycerol-3-phosphate dehydrogenase (mGPDH) assays were performed as described under "Experimental Procedures," specific activities are expressed per milligram of protein. n ≥ 14, animals of both genders and age of 65 ± 5 days. Values are expressed as means ± S.D. Profound differences in sensitivity to cold were observed as the offspring approached the 10th backcross generation. To keep homozygous postnatal mice alive, the breeding room temperature was increased to 27 °C. In adult mice, the ability to produce heat under cold stress was characterized with cold sensitivity tests. Fig.1 shows the traces of body temperature of mice during cold exposure. 129/SvImJ and C57BL/6J wild-type mice could maintain their body temperature in an environment at 4 °C for an indefinite time (Fig. 1 A). UCP1-deficient mice congenic on 129/SvImJ and C57BL/6J backgrounds lost body core temperature rapidly (Fig. 1, B and C). The average rate of temperature loss of 4.9° and 3.9°/h was similar for both backgrounds. We also analyzed UCP1-deficient mice on a (129/SvImJ × C57BL/6J) F1 hybrid background (Fig. 1 D). These hybrid mice were generated by out-crossing parental lines ofUcp1−/− mice. Remarkably, these mice were resistant to cold, despite the absence of UCP1. We observed also that ∼1 out of 10 mice of the cold-resistant groups lost body temperature (Fig. 1,A and D). It is not unexpected to occasionally find a cold-sensitive mouse in a cold-resistant group since anything that affects the health of a mouse impairs cold resistance. It is quite possible that the surgery to implant temperature probes affected the health of the mice. FFAs, released by the action of hormone-sensitive lipase, play a key role in thermogenesis as an energy substrate for mitochondrial β-oxidation. Additionally, they are also thought to stimulate respiration by direct activation of UCP1. To establish the validity of this interaction, we investigated the ability of fatty acids to de-energize BAT mitochondria, as evidenced by mitochondrial membrane potential. UCP1-deficient mice and control mice on different genetic backgrounds were kept at 27 °C prior to dissection, and the protonmotive force of isolated BAT mitochondria was measured. Fig.2 shows representative traces for wild-type (A, C, and E) and UCP1-deficient (B, D, and F) mitochondria on 129/SvImJ (A and B), C57BL/6J (C and D), and F1 hybrid (E andF) backgrounds. As expected, BAT mitochondria in the control mice showed very little increase in protonmotive force after the addition of substrate, but a large increase after purine nucleotide addition (Fig. 2 A, C, and E). As previously shown (1Matthias A. Jacobsson A. Cannon B. Nedergaard J. J. Biol. Chem. 1999; 274: 28150-28160Abstract Full Text Full Text PDF PubMed Scopus (102) Google Scholar, 16Monemdjou S. Kozak L.P. Harper M.-E. Am. J. Physiol. 1999; 276: E1073-E1082PubMed Google Scholar), mitochondria in the mutant mice became maximally charged following the addition of substrate only (Fig. 2,B, D, and F) and showed no further increase upon the addition of 1 mm GDP. After achievement of maximum protonmotive force, BAT mitochondria were subjected to de-energization by stepwise addition of FFA. The final concentrations of lauric acid were 33, 67, and 100 μm. In this phase, UCP1-deficient mitochondria (Fig. 2, B, D, andF) were substantially less responsive to the addition of the same amount of FFA compared with BAT mitochondria in control mice (A, C, and E). The last step of each experiment was always the total de-energization with FCCP, a nonspecific uncoupler. Overall, the traces in Fig. 2 show an impaired fatty acid-induced de-energization in UCP1-deficient BAT mitochondria that is independent of the genetic background. The quantitative effects of fatty acids on membrane potential were determined using mitochondria from mice on a 129/SvImJ background. The experiments were carried out as described above, and de-energization was expressed in millivolts. Fig. 3 shows dose-response curves for linoleic acid in control BAT mitochondria and UCP1-deficient mitochondria from BAT and liver. The concentration of unbound linoleic acid in the presence of 0.1% BSA was calculated according to Richieri et al. (24Richieri G.V. Anel A. Kleinfeld A.M. Biochemistry. 1993; 32: 7574-7580Crossref PubMed Scopus (336) Google Scholar). Values were fitted by graphs similar to Michaelis-Menten kinetics, and initial de-energization rates were estimated. The de-energization in UCP1-deficient BAT mitochondria was 35% of that in wild-type BAT mitochondria. For comparison, liver mitochondria with 0.6 mV/nm are still less prone to de-energization compared with UCP1-deficient BAT mitochondria with 1.6 mV/nm. The abilities of lauric, linoleic, and stearic acids to de-energize mitochondria are summarized in Table II. All measurements were carried out in the presence of 0.1% BSA; therefore, the concentrations of unbound fatty acids are much lower than the nominal concentrations. Because correction coefficients for calculating the concentration of unbound molecules are not known for some fatty acids, the nominal concentrations are presented. As already shown by Matthias et al. (1Matthias A. Jacobsson A. Cannon B. Nedergaard J. J. Biol. Chem. 1999; 274: 28150-28160Abstract Full Text Full Text PDF PubMed Scopus (102) Google Scholar), the maximum membrane potential was slightly increased in UCP1-deficient mitochondria. Contradictory to the findings of this group, we observed that with a nominal concentration of 66 μm lauric acid, a 36% reduction in membrane potential occurred in wild-type mice, compared with a 16% reduction in UCP1-deficient mice. The ability of linoleic and stearic acids to de-energize was also reduced in UCP1-deficient BAT mitochondria. We measured reductions from 30 to 14% and from 31 to 15% for the 27 μm nominal concentration of linoleic and stearic acids, respectively.Table IIProton-motive force and de-energization by fatty acids in BAT mitochondria from Ucp1 knockout mice and control miceGenotypeΔpDe-energizationLauric acid (66 μm)Linoleic acid (27 μm)Stearic acid (27 μm)mV%Ucp1(+/+)−145 ± 1736 ± 532 ± 634 ± 8Ucp1(−/−)−161 ± 1015 ± 62-ap < 0.001 compared with wild-type mitochondria.12 ± 32-ap < 0.001 compared with wild-type mitochondria.13 ± 72-ap < 0.001 compared with wild-type mitochondria.The experiments were performed as described for Fig. 2 for both genotypes in the presence of 1 mm GDP. De-energization is defined as the ratio between the reduction of Δp and the maximum of Δp. Concentrations of FFA are given as nominal concentrations in the presence of 0.1% BSA. n ≥ 9, animals of both genders and age of 65 ± 5 days. Values are expressed as means ± S.D.2-a p < 0.001 compared with wild-type mitochondria. Open table in a new tab The experiments were performed as described for Fig. 2 for both genotypes in the presence of 1 mm GDP. De-energization is defined as the ratio between the reduction of Δp and the maximum of Δp. Concentrations of FFA are given as nominal concentrations in the presence of 0.1% BSA. n ≥ 9, animals of both genders and age of 65 ± 5 days. Values are expressed as means ± S.D. An alternative strategy for showing the importance of UCP1 for fatty acid-induced de-energization of isolated BAT mitochondria is based on the fact that long-term cold exposure increases UCP1 in BAT mitochondria (25Ricquier D. Mory G. Bouillaud F. Thibault J. Weissenbach J. FEBS Lett. 1984; 178: 240-244Crossref PubMed Scopus (64) Google Scholar, 26Jacobsson A. Stadler U. Glotzer M.A. Kozak L.P. J. Biol. Chem. 1985; 260: 16250-16254Abstract Full Text PDF PubMed Google Scholar). Accordingly, we exposed normal mice to 4 °C for 1 week and performed the same de-energization experiments as described above. BAT mitochondria from cold-exposed mice showed 39 ± 3% de-energization after the addition of 66 μm lauric acid versus 25 ± 6% in control mice at 27 °C, indicating a positive correlation between higher content of UCP1 and fatty acid-induced de-energization. The key step in heat production of BAT is the activation of energy expenditure by uncoupling the respiration from ATP synthesis by UCP1. Respiratory studies were carried out to determine whether impaired de-energization in UCP1-deficient mitochondria leads to impaired activation of respiration. Fig. 4 (A andB) shows representative traces for oxygen consumption in BAT mitochondria. The BAT mitochondria from control mice reduced oxygen consumption after the addition of 1 mm GDP, whereas respiration of UCP1-deficient mitochondria was not affected by GDP. After the addition of 20 μm linoleic acid, the coupled mitochondria from control mice showed a 20% increase in oxygen consumption (Fig. 4 A) compared with UCP1-deficient mitochondria, whose respiration remained unaffected upon the addition of the same amount of fatty acid (Fig. 4 B). Maximum respiration was achieved by the addition of 2 μm FCCP. To show quantitative effects on fatty acid-stimulated respiration, we performed dose-response experiments. We coupled BAT mitochondria with 1 mm GDP and monitored oxygen consumption after the addition of varying amounts of lauric acid. These results are summarized in Fig.4 C. The nominal concentration of fatty acid that caused an increase in respiration was between 25 and 50 μm for wild-type mitochondria and between 75 and 100 μm for UCP1-deficient mitochondria. Our finding that fatty acid-induced de-energization of isolated BAT mitochondria is affected by UCP1 deficiency is contrary to the findings of Matthias et al.(1Matthias A. Jacobsson A. Cannon B. Nedergaard J. J. Biol. Chem. 1999; 274: 28150-28160Abstract Full Text Full Text PDF PubMed Scopus (102) Google Scholar). The only known variable that might account for the differences between our results and those of Matthias et al. is the genetic background of these mice. The cold resistance of UCP1-deficient mice on a hybrid background compared with cold sensitivity on inbred backgrounds suggests the existence of an alternative thermogenic mechanism induced in hybrid mice. To determine if this thermogenic pathway is mediated by the UCP1 homologs UCP2 and UCP3, we performed expression analysis at the mRNA level. Consistent with previous reports (1Matthias A. Jacobsson A. Cannon B. Nedergaard J. J. Biol. Chem. 1999; 274: 28150-28160Abstract Full Text Full Text PDF PubMed Scopus (102) Google Scholar, 10Enerback S. Jacobsson A. Simpson E.M. Guerra C. Yamashita H. Harper M.-E. Kozak L.P. Nature. 1997; 387: 90-94Crossref PubMed Scopus (1082) Google Scholar), the Ucp2 mRNA levels in BAT from UCP1-deficient mice were elevated over the controls (Fig.5 A). However, the degree ofUcp2 overexpression varied between 250% for the C57BL/6J genetic background and 600% for the 129/SvImJ genetic background.Ucp2 mRNA levels in cold-resistant F1 progeny were intermediate with respect to the parental strains, suggesting that variation in Ucp2 mRNA levels may be co-dominantly inherited. There were no differences in Ucp3 expression levels compared with control mice. Knowing that Ucp3 is mainly expressed in skeletal muscle and has a thermogenic potential (27Clapham J.C. Arch J.R. Chapman H. Haynes A. Lister C. Moore G.B. Piercy V. Carter S.A. Lehner I. Smith S.A. Beeley L.J. Godden R.J. Herrity N. Skehel M. Changani K.K. Hockings P.D. Reid D.G. Squires S.M. Hatcher J. Trail B. Latcham J. Rastan S. Harper A.J. Cadenas S. Buckingham J.A. Brand M.D. Nature. 2000; 406: 415-418Crossref PubMed Scopus (518) Google Scholar), we also analyzed skeletal muscle. Differences in Ucp2and Ucp3 mRNA levels in skeletal muscle cannot account for cold resistance of F1 hybrid Ucp1knockout mice (Fig. 5 B). Determining the interrelationships between thermogenesis, energy expenditure, and body weight regulation, particularly with respect to energy reserves, has both fundamental and practical importance in mammalian biology. However, progress in understanding how thermogenesis can be manipulated to affect energy expenditure and body weight has been slow since the number of potential mechanisms that can produce heat is very large. UCP1-dependent non-shivering thermogenesis in brown adipose tissue is a well defined mechanism that can be up-regulated pharmacologically, and several genetically altered mouse models provide experimental systems to evaluate its role in the regulation of body weight. Transgenic mice in which Ucp1 is overexpressed have consistently provided strong evidence that elevated UCP1 can reduce excess adiposity. However, paradoxically, there is no evidence to suggest that the absence of UCP1 will produce the opposite phenotype, i.e. increased adiposity. Other phenotypes of mice carrying the targeted inactivation of Ucp1 need explanation. In our original publication (10Enerback S. Jacobsson A. Simpson E.M. Guerra C. Yamashita H. Harper M.-E. Kozak L.P. Nature. 1997; 387: 90-94Crossref PubMed Scopus (1082) Google Scholar), we also reported that 15% of the homozygous targeted mice were resistant to cold exposure. How is it that occasionally mice homozygous for the inactive allele are resistant to cold? Subsequently, Matthias et al.(1Matthias A. Jacobsson A. Cannon B. Nedergaard J. J. Biol. Chem. 1999; 274: 28150-28160Abstract Full Text Full Text PDF PubMed Scopus (102) Google Scholar) obtained data showing that mitochondria isolated from targeted Ucp1 mice are uncoupled by FFAs to the same degree as mitochondria from wild-type mice, suggesting that FFAs do not directly activate UCP1-dependent proton flux. Since these effects of the mutation on body weight gain and thermosensitivity and the effects of FFAs were contrary to our expectations, we established the Ucp1 mutation on congenic backgrounds to eliminate uncontrolled genetic variability as a confounding factor. Experiments on thermosensitivity and FFA activation of UCP1 have been described here; the effects of the targeted Ucp1 mutation on body weight gain is under investigation. Cold sensitivity is a complex phenotype that was linked to metabolic disorders some 40 years ago with the measurement of body temperature in genetically obese (ob/ob) mice during cold exposure (28Davies T.R.A. Mayer J. Am. J. Physiol. 1954; 177: 222-226Crossref PubMed Scopus (63) Google Scholar, 29Trayhurn P. Thurlby P.L. James W.P.T. Nature. 1977; 266: 60-62Crossref PubMed Scopus (160) Google Scholar). Since then, mouse models lacking key proteins for non-shivering thermogenesis have been established, but not all of them proved to be cold-sensitive. Mice lacking the β3-adrenergic receptor, which has been postulated to be essential for the differentiation of the brown adipocyte (30Zhao J. Cannon B. Nedergaard J. Am. J. Physiol. 1998; 275: R2002-R2011PubMed Google Scholar), can be exposed to 4 °C without symptoms of hypothermia, apparently through compensation by β1-adrenergic receptors (31Susulic V.S. Frederich R.C. Lawitts J. Tozzo E. Kahn B.B. Harper M.-E. Himms-Hagen J. Flier J.S. Lowell B.B. J. Biol. Chem. 1995; 270: 29483-29492Abstract Full Text Full Text PDF PubMed Scopus (417) Google Scholar). More relevant with respect to the UCP1-deficient mice are the dopamine β-hydroxylase-deficient mice that are unable to synthesize norepinephrine (12Thomas S.A. Palmiter R.D. Nature. 1997; 387: 94-97Crossref PubMed Scopus (221) Google Scholar). Even though these mice are on an undefined genetic background, they are even more sensitive to cold compared with UCP1-deficient mice, and there is no evidence that any of the mutant animals are resistant to cold. The fact that these mice have low levels of Ucp1 mRNA suggests that the extreme sensitivity to cold is based on the impairment of mechanisms of thermogenesis in brown adipocytes besides UCP1 levels. Certainly the ability of these dopamine β-hydroxylase-deficient mice to generate fatty acids for β-oxidation from the triglyceride pools has been strongly reduced, as evidenced by the accumulation of lipid in the cytoplasm of the brown adipocytes. Other processes, including the stimulation of blood flow through the tissue, must be reduced in the absence of norepinephrine. We have shown that UCP1-deficient mice are cold-sensitive on both 129/SvImJ and C57BL/6J backgrounds, but not on the (129/SvImJ × C57BL/6J) F1 hybrid background. The resistance to cold in the hybrid animal cannot be explained by the characteristics of its mitochondrial physiology, at least within the limits of the experiments on membrane potentials and respiration and activation by FFAs. The results with inbred and hybrid UCP1-deficient mice are indistinguishable. In addition, there are no differences in expression levels ofUcp2and Ucp3in BAT or skeletal muscle that would suggest that hybrid mice are resistant to cold due to elevated expression of these genes. In addition to the resistance to cold found in the hybrid mice, UCP1-deficient mice on an inbred background can be adapted to tolerate cold by gradually reducing the ambient temperature; however, the acquisition of this property is variable. 2W. E. Hofmann, X. Liu, C. M. Bearden, M.-E. Harper, and L. P. Kozak, unpublished observation. We interpret these results to suggest that the cold sensitivity phenotype is very sensitive to the environment and that alternative thermogenic mechanisms can be induced to compensate for UCP1-based thermogenesis. The adaptation of C57BL/6J and 129/SvImJ mice to cold exposure or the constitutive resistance in hybrid mice can provide a biological system to identify alternative mechanisms of thermogenesis. The cold resistance of UCP1-deficient mice on a hybrid background is a dramatic example of heterosis or hybrid vigor. Heterosis has been extensively studied in plants with respect to growth; it has also been studied in mammals for complex phenotypes like behavioral traits (32Comings D.E. MacMurray J.P. Mol. Genet. Metab. 2000; 71: 19-31Crossref PubMed Scopus (284) Google Scholar). This might also account for cold resistance of UCP1-deficient mice on a F1 hybrid background as well. One of the more relevant examples of heterosis comes from a study of hybrids of four parental lines of chicken in which positive heterosis was observed for norepinephrine levels in the hypothalamus in a line-specific manner (33Barbato G.F. Kruzelock R.P. Behav. Genet. 1992; 22: 381-398Crossref PubMed Scopus (5) Google Scholar). In these same hybrids, heterosis was found for dihydroxyphenylacetic acid in the optic tectum only. These results suggest that heterosis can affect the level of hormones that are relevant to thermogenesis. At the gene level, mice heterozygous for the inactive allele for α-calcium/calmodulin kinase II gene show much more defensive aggression than either the homozygous wild-type or mutant mice (34Chen C. Rainnie D.G. Greene R.W. Tonegawa S. Science. 1994; 266: 291-294Crossref PubMed Scopus (248) Google Scholar). An explanation of the mechanisms for these effects is not known, but it is clear that heterosis is not an uncommon phenomenon. Despite the fact that heterosis is poorly understood, hybrid effects may explain the cold resistance of some UCP1-deficient mice on a mixed undefined background in our original study (10Enerback S. Jacobsson A. Simpson E.M. Guerra C. Yamashita H. Harper M.-E. Kozak L.P. Nature. 1997; 387: 90-94Crossref PubMed Scopus (1082) Google Scholar). Using UCP1-deficient mice on 129/SvImJ, C57BL/6J, and F1 hybrid backgrounds, we observed a significant reduction of fatty acid-induced de-energization of BAT mitochondria compared with control mice. This indicates that FFAs directly activate the uncoupling activity of UCP1 in BAT mitochondria. However, UCP1-deficient BAT mitochondria are still more prone to fatty acid-induced de-energization compared with liver mitochondria. This indicates that despite similar protonmotive force, UCP1-deficient BAT mitochondria and liver mitochondria are not similar, as already shown by the lack of ATP synthase activity in the former (1Matthias A. Jacobsson A. Cannon B. Nedergaard J. J. Biol. Chem. 1999; 274: 28150-28160Abstract Full Text Full Text PDF PubMed Scopus (102) Google Scholar). It also suggests that UCP1-deficient mitochondria retain a significant capacity to generate a membrane potential that can be uncoupled by fatty acids. It is uncertain whether this residual uncoupling contributes to heat production. Our finding that fatty acids have a reduced ability to de-energize BAT mitochondria in the absence of UCP1 irrespective of the genetic background of the mice does not agree with the findings of Matthiaset al. (1Matthias A. Jacobsson A. Cannon B. Nedergaard J. J. Biol. Chem. 1999; 274: 28150-28160Abstract Full Text Full Text PDF PubMed Scopus (102) Google Scholar). They showed no impaired fatty acid-induced de-energization of UCP1-deficient BAT mitochondria compared with control mice. However, inspection of their data generated from UCP1-deficient mice on an undefined genetic background shows a trend toward UCP1 dependence. The uncoupling activity for seven fatty acids (as shown in Table III of Ref. 1Matthias A. Jacobsson A. Cannon B. Nedergaard J. J. Biol. Chem. 1999; 274: 28150-28160Abstract Full Text Full Text PDF PubMed Scopus (102) Google Scholar) indicates that the uncoupling activity for each fatty acid is slightly decreased in UCP1-deficient BAT mitochondria. Considering all seven fatty acids together, an average value of 1.03 mV/μm can be calculated for wild-type mice versus 0.67 mV/μm FFA in UCP1-deficient mice. It should be emphasized that their data were expressed as millivolts/nominal FFA concentration in the presence of 0.1% BSA. Due to the exponential relationship between nominal and free (unbound) concentrations of fatty acids in the presence of BSA, the differences, expressed as millivolts/unbound FFA, would be more remarkable. However, using mice on an undefined background, their analysis did not show statistically significant differences. The reason why the impairment of fatty acid-induced de-energization by UCP1 deficiency is more significant in a mouse on a congenic background (our mice) is presently unknown. We conclude that, as previously postulated from biochemical studies, FFAs are vital physiological activators of UCP1-mediated uncoupling of respiration by the stimulation of proton flux. There are two models describing how FFAs interact with UCP1 to uncouple respiration. According to Winkler and Klingenberg (4Winkler E. Klingenberg M. J. Biol. Chem. 1994; 269: 2508-2515Abstract Full Text PDF PubMed Google Scholar), FFAs provide carboxyl groups for transporting protons together with amino acid carboxyl groups in the transmembrane channel of UCP1. In the second model, the function of UCP1 is indirect (2Jezek P. Drahota Z. Ring K. J. Lipid Mediators. 1990; 2: 85-94PubMed Google Scholar, 3Skulachev V.P. FEBS Lett. 1991; 294: 158-162Crossref PubMed Scopus (393) Google Scholar). This model proposes that neutral FFAs can pass the inner mitochondrial membrane transporting protons. After deprotonation in the matrix, UCP1 transports the carboxylate anions back across the membrane, which completes the fatty acid cycle of proton shuttling. Although our experiments do not provide any insight as to which model is correct, they do show that BAT mitochondria from UCP1-deficient mice have an impaired fatty acid-induced de-energization. We suggest that the interaction of FFAs and UCP1 enables changes in the intercellular concentration of FFA to regulate the uncoupling activity of UCP1. It is also evident from our results on heterosis that the absence of a UCP1-based thermogenesis does not preclude the emergence of alternative mechanisms of thermogenesis to protect mice from cold exposure. We thank Robert Koza and Martin Rossmeisl for critical reading of the manuscript. brown adipose tissue free fatty acids FCCP, carbonyl cyanide p-trifluoromethoxyphenylhydrazone
Referência(s)