Deiodinase Activity of Brown Adipose Tissue in Daurian Ground Squirrel During Hibernation and Arousal

http://www.paper.edu.cn Comparative Biochemistry and Physiology Part A 120 (1998) 745–752

Uncoupling protein mRNA, mitochondrial GTP-binding, and T4 5%-deiodinase activity of brown adipose tissue in Daurian ground squirrel during hibernation and arousal

Xiao-tuan Liu a, Qi-shui Lin b, Qing-fen Li a,*, Chen-xi Huang a, Ru-yong Sun a

a Laboratory of Animal Ecology, Department of Biology, Beijing Normal Uni6ersity, 19 Xinwai Street, Beijing 100875, People’s Republic of China b State key laboratory of Molecular Biology, Shanghai Institute of Biochemistry, Academia Sinica, 320 Yueyang Road, Shanghai 200031, People’s Republic of China

Received 21 November 1997; received in revised form 5 June 1998; accepted 11 June 1998

Abstract

The mRNA level of uncoupling protein (UCP) specific for brown adipose tissue (BAT) in Daurian ground squirrel, was detected by using a [32P]-labeled oligonucleotide probe. The UCP concentration in mitochondria was indirectly determined by 3 % titration with its specific ligand [ H]-labeled GTP. Type II T4 5 -deiodinase of BAT was assayed concomitantly. We found two species of mRNA for UCP with lengths of about 1.9 and 1.5 kb, respectively, both occurring in almost the same concentration. UCP mRNA content was elevated significantly during hibernation, but the UCP concentration did not change compared with that of nonhibernating controls kept at room temperature. When hibernating squirrels were aroused, the UCP mRNA remained at the elevated level as during hibernation, but the UCP concentration increased in comparison with that of nonhibernating controls or % during hibernating. Changes in T4 5 -deiodinase activity in BAT were similar to the variations of the UCP mRNA level. These % results suggest that the activation of T4 5 -deiodinase in BAT may be an important factor for the up-regulation and maintenance of UCP mRNA content needed for the synthesis of sufficient UCP to acquire the thermogenic capacity for arousal from hibernation. © 1998 Elsevier Science Inc. All rights reserved.

Keywords: Arousal; Brown adipose tissue; Daurian ground squirrel; GTP-binding to mitochondria; Hibernation; mRNA for % uncoupling protein; Nonshivering thermogenesis; T4 5 -deiodinase

1. Introduction [5]. The key element for the energy dissipation capacity of BAT is uncoupling protein (UCP or thermogenin), a Brown adipose tissue (BAT) is the main site of 32 kDa protein uniquely expressed in the inner mem- facultative thermogenesis in small rodents and probably brane of BAT mitochondria. UCP dissipates the proton most eutherian mammals during the early postnatal gradient created by the respiratory chain, thereby accel- period. Furthermore, this tissue is also traditionally erating respiration. The uncoupling protein has a high regarded as being of particular signiﬁcance in hibernating afﬁnity for purine nucleotides such as GDP and GTP species, principally in relation to arousal from hiberna- [19,27]. Recently, Huang and Klingenberg reported that tion and maintenance of the euthermic state in the cold GTP binds to mitochondria or UCP more tightly than GDP [10]. In addition, the binding capacity determined * Corresponding author. Dept. of Biology, Beijing Normal Univer- represents the overall nucleotide binding sites on the sity. Tel.: +86 10 62209029; fax: +86 10 62200567; e-mail: BAT mitochondria, which limits the thermogenic poten- [email protected] tial of BAT to respond to environmental stimuli.

746 X.-t. Liu et al. / Comparati6e Biochemistry and Physiology, Part A 120 (1998) 745–752 tion induced by cold exposure of laboratory rodents mine if they were hibernating. Each animal was allowed [32]. The optimal expression of UCP requires high to go through at least one spontaneous arousal before intracellular concentration of triiodothyronine (T3) being used. The hibernating squirrels were further di-

[1,2]. The high T3 concentration required is provided by vided into two small groups. One group, the deeply hibernating ones (n=4), were decapitated immediately. the sympathetic activation of type II thyroxine (T4) % The second group, the arousing squirrels (n=4), were 5 -deiodinase which converts T4 to T3 in BAT [1,3,29]. % allowed to partially arouse before being sacrificed. Therefore, type II T4 5 -deiodinase is a site of synergism between the sympathetic nervous system and thyroid Arousal was induced by gentle subsequent disturbing hormones in the thermogenic activation of BAT [6], the hibernating animal by insertion of a rectal thermis- and is regarded as another marker of brown adipose tor. Squirrels were sacrificed after about 45 min when the rectal temperature had risen by about 6°C. tissue [20]. Rodent hibernators, which frequently exhibit body temperatures as low as 4–6°C, undergo remarkable 2.2. Brown adipose tissue rewarming during periodic arousals, with body temperature increasing to 37°C within 2–3 h [18]. Recently, After sacrifice, the bilateral axillary brown fat pads Spermophilus duaricus, a typical hibernator inhabiting (the main BAT depot in ground squirrels) were rapidly North China, has been studied in our laboratory. It was dissected, cleaned of any adhering connective tissue, confirmed that BAT was the major source of nonshiver- weighed, quickly frozen in liquid N2, then stored at 70°C until use, except that T 5%-deiodinase was ing thermogenesis induced by cold acclimation in eu- − 4 immediately assayed with about 0.3 g fresh tissue as thermic squirrels [35]. But it remains unclear how BAT described below. functions during the bouts of hibernation and arousal in this species. In the present study, we further investigated the thermogenic capacity of brown adipose tissue 2.3. Isolation of BAT mitochondria and assay of total as measured by mitochondrial GTP binding, and the and mitochondrial protein relation between expression of UCP mRNA and type II T 5%-deiodinase during hibernation and partial arousal. BAT mitochondria were isolated as described previ- 4 ously [14]. BAT was homogenized and samples of homogenate were taken for the measurement of total 2. Material and methods tissue protein. The remainder of the homogenate was used to prepare mitochondria. Protein in homogenates and in isolated mitochondria was determined by the 2.1. Animals Lowry’s method [15] with bovine serum albumin as standard. Daurian ground squirrels were trapped on farmland on the outskirts of Yanggao county (40°51%N, 113°72%E), Shanxi province, China, between September 2.4. GTP-binding to BAT Mitochondria and October, 1996. They were transported to the animal facility of the Biological Sciences Building, Beijing The specific number of GTP binding sites was determined from a seven or eight-point Scatchard plot of Normal University, and four to five squirrels were 3 housed in one steel wire cage at a temperature of [ H]-GTP binding in single determination, as described by Lin [13]. Briefly, protein (0.3–0.5 mg) and 0.5 to 5 2292°C and kept on a 12 L:12 D light cycle (lights on vM[3H]-GTP (2.7 vCi ml−1) and [14C]-sucrose (0.27 at 06:00) with food (for laboratory mice and rat pro- vCi ml−1) were incubated for 15 min at 25°C in total duced by Beijing Institute of Laboratory Animal, Bei- volume of 150 vl buffer containing 20 mM MOPS, 0.16 jing) and water ad libitum. mM EDTA, 20 mM Na SO , pH 6.7. Incubation was Squirrels were divided into two groups in January 2 4 terminated by centrifugation for 10 min (16000×g), 1997. Both males and females, equally divided among the supernatant was carefully aspirated and the tube each group, were used. One was the nonhibernating wall was then washed quickly and gently by using 250 9 control (n=6) which were held at 22 2°C through- vl buffer without [3H]-GTP and [14C]-sucrose again by out, and the other was the experimental group (n=8). centrifugation for 5 min. This washing supernatant was The body mass of the latter group (n=8) was moni- discarded and the pellet was suspended in 50 vlof20% tored regularly. When squirrels had increased above Triton X-100. Radioactivity of both the dissolved pellet 250 g, they were placed into individual plastic cages and the supernatant after the first centrifugation was with sawdust and transferred to a room of 492°C in counted in a liquid scintillation counter. [14C]-Sucrose total darkness with neither food nor water. The animals was used to trace the amount of contaminating super- at 4°C were checked visually every morning to deter- natant in the pellet fraction. 中国科技论文在线 http://www.paper.edu.cn

X.-t. Liu et al. / Comparati6e Biochemistry and Physiology, Part A 120 (1998) 745–752 747

2.5. Assessment of uncoupling protein mRNA le6el buffer. The homogenized tissue was centrifuged at 800×g for 15 min at 4°C and the infranatant was 2.5.1. Oligonucleotide probe directly used to assay the activity of enzyme as previ- A 32-mer antisense oligonucleotide with a ho- ously described [12]. Protein content of the BAT in- mologous sequence speciﬁc for UCP mRNA from a franatant was quantitated as above [15]. The T4 (6.4 wide variety of species was selected [32]. The oligonu- nM) was used as substrate. Individual incubation ves- % cleotide with 5 -hydroxyl termini was synthesized in sels contained 0.1 ml infranatant and T4 in HEPES Shanghai Institute of Biochemisty (with DNA Synthe- buffer were incubated at 37°C for 60 min. At the end of sizer, Oligo 1000, Beckman) and labeled with [32P]- incubation, 2 vols. of ice-cold absolute ethanol were dATP and T4 polynucleotide kinase. added to each incubation mixture which was then vor- RNA extraction and Northern blotting analysis: texed and kept at −20°C overnight before centrifuga-

Frozen BAT in liquid N2 was pulverized immediately tion at 800×g for 20 min. The enzyme denatured by 2 and total RNAs were isolated using acidic guanidinium vols. of absolute ethanol before the addition of T4 was isothiocyanate [7]. The amount and the purity of RNA used as blank control. Ethanol extracts were then as- solution was investigated on Gene Quant. The ratio sayed for T3 by a radioimmunoassay using a triiodothy- 260/280 nm was routinely 1.8 or above. The RNA ronine RIA kit (produced by the China Institute of preparation was routinely tested for integrity on Atomic Energy, Beijing, Lot. 9707) except that the agarose minigels. Specific mRNA levels were assessed same volume of 67% ethanol, as used in the sample by Northern blotting. Total RNA (20 vg) was sepa- tubes, was added to tubes of the standards of the RIA rated by electrophoresis on 1.0% agarose and formalde- to minimize the effects of ethanol on the measurement hyde gel (in duplicate for each animal) and blotted onto of T3. charged nylon membrane in 20×SSC by capillary blotting for about 18 h. The membrane was cross-linked on 2.7. Analysis of data RNA side by exposure to UV-light. It was then incubated with prehybridization solution (5×Denhardt’s, Data were analyzed by a statistical program SPSS. 6×SSPE, 0.5% SDS) for 2–3 h and then several Paired t-tests were used for testing initial and final body milliliter hybridization solution overnight in bottles ro- mass data. One-way ANOVA was used for statistical tating in hybridization oven at 42°C. The hybridization analysis of other data. If treatment effects were signifi- solution contained the [32P]-dATP labeled oligonucle- cant, differences between group means were analyzed otide probe (2.1×108 cpm vg−1) for UCP mRNA. by the Duncan multiple range test. The significant Post-hybridization washes were: (i) 2×SSC/0.1% SDS, difference was indicated when PB0.05. for 5 min at room temperature (twice); (ii) 0.1×SSC/ 0.1% SDS for 15 min at 48°C (twice). Then the membrane was placed in a cassette with Kodak X-ray film 3. Results for autoradiography in a −70°C freezer. To correct for RNA loading and transfer by Northern blotting, a 3.1. Body mass and brown adipose tissue [32P]-dCTP labeled cDNA probe (4.4×108 cpm vg−1) specific for 18S ribosomal RNA was used (the cDNA The body weight, BAT pad mass and protein content probe was kindly provided by professor Zhang are shown in Table 1. The rectal temperature just Yonglian of State Key Laboratory of Molecular Biol- before death is also indicated in the same table. Within ogy In Shanghai Institute of Biochemistry). The posi- 2 weeks of being transferred to the cold, all animals had tion of the bands detected were monitored relative to experienced one bout of hibernation and arousal. At the 28S and 18S ribosomal bands. The detected bands the end of the experiment, the body mass of animals in of mRNA was quantitated by scanning densitometry cold was significantly decreased while there was no (with Dual-wavelength Thin-layer Chromato Scanner, marked difference between the initial and final body Model CS930, Shimadzu, Japan) mass of nonhibernating controls. One-way ANOVA indicated that the wet mass of BAT changed signifi- % 2.6. Assays of T4 5 -deiodinase in brown adipose tissue cantly in the three groups (F=6.85, P=0.012), which increased by 105% in the hibernating squirrels and by After the sacrifice, the BAT was removed, freed of 68% in the arousing animals, respectively, compared other adhering tissues, rinsed in ice-cold buffer contain- with the nonhibernating controls. When expressed as ing 0.32 M sucrose, 10 mM HEPES, 20 mM DTT, 1 grams of BAT per 100 g of final body weight, the mM EDTA, pH 7.0, blotted and weighed. The BAT differences in BAT weights were even more significant was then minced with scissors in a beaker immersed in (F=14.95, P=0.001). Similar to the result of the wet an ice-water bath and ground in a motor-driven Teflon mass, this is significantly higher both in hibernating and glass homogenizer in 10 volumes (w/v) of the HEPES in arousing animals than in nonhibernating controls. In 中国科技论文在线 http://www.paper.edu.cn

748 X.-t. Liu et al. / Comparati6e Biochemistry and Physiology, Part A 120 (1998) 745–752

Table 1 % Body and BAT weights, and protein content and T4 5 -deiodinase activity of BAT in nonhibernating, hibernating and arousing ground squirrels

Nonhibernating Hibernating Arousing

Body weight Initial (g) 268.2911.8 (6) 309.3918.6 (4) 320.7924.3 (4) Final (g) 270.3920.8 (6) 204.5914.3 (4)a 242.3936.0 (4)b Brown adipose tissue Wet mass (g) 1.5190.19 (6) 3.1090.23 (4)b 2.5390.54 (4)b g per 100 g body weight 0.5690.05 (6) 1.5290.19 (4)b 1.0390.17 (4)b,c BAT protein content Total (mg g−1 tissue) 269921.3 (6) 256916.5 (4) 265919.1 (4) Mitochondrial (mg g−1 tissue) 11.890.93 (6) 6.490.68 (4)b 7.890.62 (4)b % T4 5 -deiodinase Speciﬁc activity −1 −1 9 9 b 9 b,c (fmol T3 h mg protein) 31.9 1.98 (6) 93.7 7.69 (4) 51.6 2.88 (4) −1 −1 9 9 b 9 b,c (pmol T3 h g tissue) 7.2 0.49 (6) 19.2 1.34 (4) 11.8 0.84 (4) Total activity −1 9 9 b 9 b,c (pmol T3 h BAT pads) 10.7 1.94 (6) 60.5 2.07 (4) 29.0 2.85 (4) Body temperature (°C) 32.490.5 (6) 6.990.3 (4)b 12.390.3 (4)b,c

Values are means9S.E.M. The number of samples is shown in parentheses. a PB0.05 vs. initial body weight. b PB0.05 vs. nonhibernating. c PB0.05 vs. hibernating. addition, a markedly decreased BAT mass was exhib- total GTP binding, are calculated and also presented in ited in arousing animals compared with hibernating Table 2. Further analysis by one-way ANOVA showed ones. that the significant differences occurred to specific bind- One-way ANOVA showed that total protein contents ing (F=7.331, P=0.010) and to total binding (F= of BAT were not significantly different from each other 4.777, P=0.032). The specific binding in hibernating in the three groups (F=0.109, P=0.898). In contrast, animals was markedly reduced compared with the cor- the concentration of mitochondrial protein was responding values in arousing or in nonhibernating markedly decreased in the hibernating group compared animals. In addition, the values of specific binding in with that in nonhibernating controls. In arousing ani- the latter two groups are not statistically different. With mals, the concentration of mitochondrial protein was regard to total binding, an elevated value was shown in slightly increased compared with that of hibernating arousing animals compared with nonhibernating con- animals, but it was still lower than that of nonhibernat- trols. However, the total binding is not higher in arousing controls. ing animals than that in hibernating ones.

3.2. Mitochondrial GTP binding 3.3. mRNA of uncoupling protein

We used [3H]-GTP, a specific ligand to mitochondrial The mRNA of UCP as determined by Northern UCP, to determine the level of UCP indirectly and, blotting is shown in Fig. 2. Two mRNA species of thus, to assess the capacity of thermogenesis in brown about 1.5 and 1.9 kb were detected which are similar to fat. Scatchard analysis of GTP binding was performed the major UCP mRNA species detected in BAT from after incubation of mitochondria with a wide range of other rodents [4]. The UCP mRNA levels were not 9 GTP concentrations. The mean values of the concentra- significantly different between the hibernating (1.95 0.34) and the arousing animals (2.1790.33), although tion of GTP binding sites (Bmax) and the respective both were markedly elevated compared with that in dissociation constant (KD) are calculated from nonhibernating control animals (0.5190.10) (Fig. 3). Scatchard analysis (Table 2). It is obvious that the KDs in the three groups are not statistically different (F= % 0.230, P=0.799) (Fig. 1). However, the Bmax in the 3.4. T4 5 -deiodinase in BAT arousing animals is significantly greater than those in % the hibernating and nonhibernating animals, but no The activities of type II T4 5 -deiodinase in BAT were significant difference exists between hibernating and significantly higher in the arousing group and also in nonhibernating group. the hibernating group, when compared to the warm- The physiologically more relevant results in relation adapted, nonhibernating group (Table 1). Moreover, to thermogenic potential, i.e. specific GTP binding and both the specific and total activity in arousing animals 中国科技论文在线 http://www.paper.edu.cn

X.-t. Liu et al. / Comparati6e Biochemistry and Physiology, Part A 120 (1998) 745–752 749

Table 2 Parameters of GTP-binding to BAT mitochondria, speciﬁc and total GTP-binding of brown fat from nonhibernating, hibernating and arousing ground squirrels

Nonhibernating Hibernating Arousing

−1 9 9 9 a,b Bmax (nmol mg mitochondrial protein) 0.81 0.15 (6) 1.09 0.03 (4) 1.46 0.07 (4) v 9 9 9 KD ( M) 1.81 0.15 (6) 1.80 0.07 (4) 1.99 0.39 (4) Speciﬁc binding (nmol g−1 tissue) 10.890.59 (6) 7.090.95 (4)a 11.591.60 (4)b Total binding (nmol BAT−1 pads) 16.492.05 (6) 21.793.68 (4) 27.693.19 (4)a

Values are means9S.E.M. The number of samples is shown in parentheses. a PB0.05 vs. nonhibernating.b PB0.05 vs. hibernating. were significantly lower compared with their respective the increased metabolic cost for maintaining tempera- values in hibernating animals. ture homeostasis before entering hibernation and that spent during periodic arousals. However, brown fat deposits in the hibernating and arousing animals were 4. Discussion higher than in nonhibernating controls. Arousing animals also possessed slightly, but significantly, decreased Although BAT has long been recognized as being brown fat pads compared with hibernating animals. prominent in hibernating species and has been sug- These results suggest that the recruitment of BAT gested to have an important thermogenic function dur- induced by cold acclimation in hibernating animals was ing arousal from hibernation [16,17,22], few systematic utilized probably by increased lipid mobilization and studies on its adaptive thermogenesis during the bouts metabolic activity during arousal [18]. Such a signifi- of hibernation and arousal have been performed. We cant decrease of BAT in this study, however, may also observed a general increase in the potential for thermo- be related to the fact that the hibernating animals lost genesis of BAT in Daurian ground squirrels in arousing more their body mass than arousing ones. These differ- as well as hibernating animals. This increased ther- ence could be due to our small sample size (Table 1). mogenic capacity was largely held in check until Although brown fat mass is a convenient variable to arousal was initiated. measure, it is not a reliable indicator of the metabolic As presented in Table 1, hibernating and arousing potential of the tissue. The number of GTP binding animals significantly lost body mass. This may reflect sites to BAT mitochondria, which reflects its thermogenic capacity, was determined by Scatchard analy-

sis. Results showed that the maximum binding (Bmax) was markedly increased during arousal from hibernation, further conﬁrming that thermogenic function of BAT was ﬁnally activated in Daurian ground squirrels as suggested in other hibernators [9,16,17,22]. However,

Fig. 1. Scatchard plot of GTP binding to isolated BAT mitochondria of nonhibernating, hibernating and arousing ground squirrel. Regres- sion of bound/free values on bound values is significantly (PB0.001) Fig. 2. Northern hybridization of UCP mRNA. Total RNA from correlated for all plots. Regression coefficients (−KD) are not signifi- BAT in Daurian ground squirrels were added to each lane in the gel. cantly different. lane 1, 20 vg; lane 2, 10 vg; lane 3, 5 vg; lane 4, 1 vg; lane 5, 500 ng. 中国科技论文在线 http://www.paper.edu.cn

750 X.-t. Liu et al. / Comparati6e Biochemistry and Physiology, Part A 120 (1998) 745–752

a 3.8-fold increase in UCP mRNA level in hibernating

over that in nonhibernating squirrels. In contrast, Bmax is significantly higher (increased by 81%) in arousing animals than in the nonhibernating squirrels although the mRNA level in these animals was not markedly different from that in the hibernating animals. We assure that the increased GTP-binding capacity could still be due to an unmasking of nucleotide sites at the beginning of arousal [33]. Thus, further experiments are needed to clarify the mechanism in Daurian ground squirrel. From the investigation of UCP mRNA in BAT from Daurian ground squirrel, the following conclusion can be drawn. Firstly, the 32 base sequence probe cross-re- acts well with the UCP mRNA in Daurian ground squirrel (Fig. 2), further confirming its homology in small mammalian species [4,32]. Secondly, two mRNA Fig. 3. Level of UCP mRNA in BAT from nonhibernating, hibernat- species of about 1.9 and 1.5 kb, respectively, were ing and arousing ground squirrel. Total RNA (20 vg) from each detected, suggesting that the Duarian ground squirrel animal was applied to each lane in duplicate. Densitometric data are UCP gene contains two polyadenylation sites and both transformed, and expressed as a ratio of 18 S ribosomal mRNA sites are used. These results are consistent with that of content for each sample. Then the ratios are analyzed by one-way other species investigated, including mouse and rat [4], ANOVA followed by the Duncan multiple range test. Values are means9S.E.M. The number of samples is indicated at the bottom of Djungarian ground squirrel [34] and 13-line ground the bar. *, PB0.05 versus nonhibernating. squirrel [21]. However, in this species, the two mRNA species occurred almost in the same concentration whereas either 1.5 or 1.8 kb mRNA predominated in the apparent dissociation constant (KD) was almost the other rodent species. Moreover, Martins et al. [16] same between the nonhibernating and the hibernating observed only one mRNA species, corresponding to 1.5 and arousing groups, indicating that there was no kb, in garden dormouse. This evidence suggests that it change in the affinity for GTP binding during different is of great importance to explore the mechanism of phase of hibernating bouts. Therefore, the increase in UCP mRNA formation in the contest of evolutionary GTP binding sites during arousal did not result from a adaptation. change in affinity of UCP for ligand. Given that the As reported previously, expression of the UCP gene GTP-binding capacity determined at such equilibrium is vigorously stimulated within minutes of cold expo- binding assay might represent the total nucleotide bind- sure of rodents [11], and the full response of the UCP ing sites, i.e. UCP concentration in mitochondria as gene requires the synergism between norepinephrine

Huang and Klingenberg suggested [10], the increased (NE) and a high intracellular concentration of T3 pro- % binding capacity during arousal was likely due to a new vided by type II T4 5 -deiodinase [2]. In vitro studies synthesis of UCP. However it is doubtful that protein also demonstrated that the stimulation of UCP gene by synthesis should happen this quickly during such short either NE or T3 independently was modest, only 2- to period of early arousal. 4-fold, whereas if they were present together a 20- to Our data show that no significant change occurred in 30-fold increase in gene transcription was induced [1]. the level of UCP mRNA between the hibernating and Some evidence showed that the high UCP mRNA level arousing animals, whereas the UCP mRNA levels in could be maintained by a longer half-life or by reduc- these cold-acclimated animals were prominently ele- tion of its turnover in cold exposed rats [28]. Studies vated compared with that in the warm-acclimated, non- with freshly dispersed brown adipocytes further re- hibernating ones. This result suggests that it was during vealed that T3 can directly prolong the half-life of UCP the hibernating stage when the UCP mRNA needed mRNA by a mechanism which does not require ongo- had been induced to maintain sufficient UCP to finally ing transcription [3]. Recently, Obregon et al. [23] acquire the thermogenic capacity for arousal from hi- found that UCP mRNA levels in fetal BAT in labora- bernation. However, it is obvious that the UCP levels tory rats correlated well with its T3 concentration and % determined by GTP binding to mitochondria were not confirmed that BAT type II T4 5 -deiodinase played an entirely consistent with their corresponding UCP important role in the production and maintenance of T3 mRNA levels as expected. The UCP concentration as concentration. With our results, it is presumed that the expressed by Bmax in hibernating animals was not sig- T3 availability in BAT is modulated by type II T4 nificantly higher than in nonhibernating animals despite 5%-deiodinase, and this enzyme was significantly acti- 中国科技论文在线 http://www.paper.edu.cn

X.-t. Liu et al. / Comparati6e Biochemistry and Physiology, Part A 120 (1998) 745–752 751

vated during hibernation and arousal, and thus caused References the increase of UCP mRNA in Duarian ground squirrels. However, We cannot speculate on the relative role [1] Bianco AC, Sheng X, Silva JE. Triiodothyronine amplifies nore- of gene transcription or mRNA turnover for the in- pinephrine stimulation of uncoupling protein gene transcription by a mechanism not requiring protein synthesis. J Biol Chem crease of UCP mRNA. Further studies are needed to 1988;263:18168–75. reveal the mechanism. [2] Bianco AC, Silva JE. Optimal response of key enzymes and However, two puzzling questions were raised. The uncoupling protein to cold in brown adipose tissue depends on % first is how an increased activity of T4 5 -deiodinase is local T3 generated. Am J Physiol 1987;253:255–63. maintained in the squirrels during hibernation. In an- [3] Bianco AC, Kieffer JD, Silva JE. Adenosine 3%,5%-monophos- other obligative hibernator, the 13-line ground squirrel, phate and thyroid hormone control of uncoupling protein mes- % senger ribonucleic acid in freshly dispersed brown adipocytes. high level of T4 5 -deiodinase activity in brown fat Endocrinology 1992;130:2625–33. during hibernation has also been reported [31]. It is well [4] Brander F, Keith JS, Trayhurn P. A 27-mer oligonucleotide % known that the T4 5 -deiodinase in BAT is rapidly and probe for the detection and measurement of the mRNA for h uncoupling protein in brown adipose tissue of different species. markedly stimulated mainly via 1-adrenergic receptors [29]. Furthermore, in the golden hamster, a facultative Comp Biochem Physiol 1993;104B:125–31. [5] Cannon B, Bengtsson T, Bronnikov G, Svoboda P, Zhao J, hibernator, during cold acclimation, it was reported h Nedergaard J. Nonshivering thermogenesis in a hibernator, the that there was an increased concentration of 1-adren- golden hamster: mediation exclusively through beta 3-adrenergic i ergic receptors [26] and desensitized 3-adrenergic re- receptors which are desensitized due to cold acclimation. In: ceptors, the main receptor subtype mediating Geiser F, Hulbert AJ, Nicol SC, editors. Adaptation to Cold. nonshivering thermogenesis of BAT [5]. Therefore, University of New England Press: Armidale, 1996:271–9. Himms-Hagen suggested that there is an enhanced h - [6] Carvalho SD, Kimura ET, Bianco AC, Silva JE. Central role of 1 brown adipose tissue thyroxine 5%-deiodinase on thyroid hor- adrenergic responsiveness, which results in increased mone dependent thermogenic response to cold. Endocrinology % i synthesis of T4 5 -deiodinase, and reduced -adrenergic 1991;128:2149–59. responsiveness [8]. This leads to the need for only a [7] Chomczynski P, Sacchi N. Single-step method of RNA isolation slight increase in activity of sympathetic neurons to by acid guanidinium thiocynate-phenol-chloroform extraction. keep the BAT with a high thermogenic potential but Anal Biochem 1987;162:156–9. [8] Himms-Hagen J. Brown adipose tissue thermogenesis: role in actually in a quiescent state. From a physiological point thermoregulation, energy regulation and obesity. In: Schonbaum of view, it seems that this strategy would be an advan- E, Lomax P, editors. Thermoregulation: Physiology and Bio- tage for hibernators, which allows them to preserve chemistry. New York: Pergamon, 1990:327–414. their BAT ready only for use during arousal from [9] Horwitz BA, Hamilton JS, Kott KS. GDP-binding to hamster hibernation when the thermogenic function of BAT brown fat mitochondria is reduced during hibernation. Am J Physiol 1985;249:R689–93. reaches its maximum [9,18]. % [10] Huang SG, Klingenberg M. Nature of the masking of nucle- The second is why the activity of T4 5 -deiodinase is otide-binding sites in brown adipose tissue mitochondria: in- markedly decreased, while the thermogenic function of volvement of endogenous adenosine triphosphate. Eur J BAT was finally activated as indicated by the increased Biochem 1995;229:718–25. GTP-binding capacity of mitochondria during arousal [11] Jacobsson A, Muhleisen M, Cannon B, Nedergaard J. The (Table 2). Such dissociation of T 5%-deiodinase activity uncoupling protein thermogenin during acclimation: indications 4 of pretranslational control. Am J Physiol 1994;267:R999–R1007. from the recruitment of BAT and maintenance of high [12] Leonard JL, Mellen SA, Larsen PR. Thyroxine 5%-deiodinase in level of UCP or UCP mRNA was also seen in labora- brown adipose tissue. Endocrinology 1983;112:1153–5. tory rat and mice [25]. Thus, there needs to be more [13] Lin CS, Klingenberg M. Characteristics of the isolated purine evidence to reveal how this enzyme activity is controlled nucleotide binding protein from brown fat mitochondria. Bio- by neuroendocrine [29,30] during arousal from chemistry 1982;21:2950–6. [14] Liu XT, Li QF, Huang CX, Sun RY. Effects of thyroid status on hibernation. cold-adaptive thermogenesis in Brandt’s vole, Microtus brandti. Physiol Zool 1997;70:352–61. [15] Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin phenol reagent. J Biol Chem Acknowledgements 1951;193:265–75. [16] Martins R, Atgier C, Ginester L, Nibbelink M, Ambid L, Ricquier D. Increased GDP-binding and thermogenic activity in We wish to thank Dr Xiao-hong Fu, Die Wang and brown adipose tissue mitochondria during arousal of hibernating Rui Zhou of the Shanghai Institute of Biochemistry, garden dormouse (Elimys quercinus L.). Comp Biochem Physiol Academy Sinica for their hospitality and cooperation. 1991;98A:311–6. The project was supported by State Key Laboratory of [17] Milner RE, Wang LCH, Trayhurn P. Brown fat thermogenesis Molecular Biology, Shanghai Institute of Biochemistry, during hibernation and arousal in Richardson ground squirrel. Am J Physiol 1989;256:R42–8. Academy Sinica, National Natural Science Foundation [18] Nedergaard J, Cannon B. Preferential utilization of brown of China (grant No.39670112) and Doctoral Pro- adipose tissue lipids during arousal from hibernation in ham- gramme Foundation of Institute of Higher Education. sters. Am J Physiol 1984;247:R506–12. 中国科技论文在线 http://www.paper.edu.cn 752 X.-t. Liu et al. / Comparati6e Biochemistry and Physiology, Part A 120 (1998) 745–752

[19] Nicholls DG. Hamster brown-adipose-tissue mitochondria. brane particles from brown fat mitochondria. Eur J Biochem Purine nucleotide control of ion conductance of inner mem- 1994;223:971–80. brane, the nature of nucleotide-binding site. Eur J Biochem [28] Rehnmark S, Bianco AC, Kieffer JD, Silva JE. Transcriptional 1976;62:223–8. and post transcriptional mechanisms in uncoupling protein [20] Nicol F, Lefranc H, Arthur JR, Trayhurn P. Characterization mRNA response to cold. Am J Physiol 1992;262:E58–67. and postnatal development of 5%-deiodinase activity in goat [29] Silva JE, Larsen PR. Adrenergic activation of triiodothyronine perirenal fat. Am J Physiol 1994;267:R144–9. production in brown adipose tissue. Nature 1983;305:712–3. [21] Nizielski SE, Billington CJ, Levine AS. BAT thermogenic activ- [30] Silva JE, Larsen PR. Hormonal regulation of iodothyronine ity and capacity are reduced during lactation in ground squirrels. 5%-deiodinase in rat brown adipose tissue. Am J Physiol Am J Physiol 1993;264:R16–21. 1986;251:E639–43. [22] Nizielski SE, Billington CJ, Levine AS. Brown fat GDP-binding [31] Tomasi TE, Stribling AM. Thyroid function in 13-line ground and circulating metabolites during hibernation and arousal. Am squirrel. In: Geiser F, Hulbert AJ, Nicol SC, editors. Adaptation J Physiol 1989;257:R536–41. to Cold. Armidale: University of New England Press, 1996:263– [23] Obregon MJ, Calvo R, Hernandez A, Rey FE, Escobar GM. 9. Regulation of uncoupling protein messenger ribonucleic acid and [32] Trayhurn P, Duncan JS. Rapid chemiluminescent detection of 5%-deiodinase activity by thyroid hormones in fetal brown the mRNA for uncoupling protein in brown adipose tissue by adipose tissue. Endocrinology 1996;137:4721–9. Northern hybridization with a 32-mer oligonucleotide end-la- [24] Otsen M, Den Bieman M, Van Zutphen LFM. Linkage of the beled with digoxigenin. Int J Obes 1994;18:449–52. gene for uncoupling protein to esterase-1,2 and haptoglobin in [33] Trayhurn P, Milner RE. A commentary on the interpretation of the rat. J Heredity 1993;84:149–51. in vitro biochemical measures of brown adipose tissue thermoge-

[25] Park IRA, Mount DB, Himms-Hagen J. Role of T3 in ther- nesis. Can J Physiol Pharmacol 1989;67:811–9. mogenic and trophic responses of brown adipose tissue to cold. [34] Weisinger H, Klaus S, Heldmaier G. Increased nonshivering Am J Physiol 1989;257:E81–7. thermogenesis, brown fat cytochrome-C oxidase activity, GDP- [26] Raasmaja A, Mohell N, Nedergaard J. Increased alpha1-adren- binding and uncoupling protein mRNA levels after short daily ergic receptor density in brown adipose tissue of cold-acclimated cold exposure. Can J Physiol Pharmacol 1990;68:195–200. rats and hamsters. Eur J Pharmacol 1984;106:489–98. [35] Zhang SZ, Li QF, Huang CX. Cold-adaptive thermogenesis of

[27] Rafeal J, Pampel I, Wang X. Effect of pH and MgCl2 on the euthermic ground squirrels, Spermophilus dauricus. Acta Theriol binding of purine nucleotides to the uncoupling protein in mem- Sinica 1996;17:67–72.