Evidence for a Compensated Thermogenic Defect in Transgenic Mice ...

6 downloads 62 Views 264KB Size Report
J. ENRIQUE SILVA. Division of Endocrinology ...... Anal Biochem 72:248–254. 35. de Lange P, Lanni A, Beneduce L, Moreno M, Lombardi A, Silvestri E, Goglia.
0013-7227/03/$15.00/0 Printed in U.S.A.

Endocrinology 144(12):5469 –5479 Copyright © 2003 by The Endocrine Society doi: 10.1210/en.2003-0687

Evidence for a Compensated Thermogenic Defect in Transgenic Mice Lacking the Mitochondrial Glycerol-3Phosphate Dehydrogenase Gene ROSANGELA A. DOSSANTOS, ASSIM ALFADDA, KAZUHIRO ETO, TAKASHI KADOWAKI, J. ENRIQUE SILVA

AND

Division of Endocrinology (R.A.D., A.A., J.E.S.), Lady Davis Institute for Medical Research, Jewish General Hospital, McGill University, Montre´al, Que´bec, Canada H3T 1E2; Department of Metabolic Diseases (K.E., T.K.), Graduate School of Medicine, University of Tokyo, Tokyo 113-8655, Japan; and Core Research for Evolutional Science and Technology of Japan Science and Technology Corp. (K.E., T.K.), Saitama 332-0012, Japan A role for mitochondrial glycerol-3-phosphate dehydrogenase (mGPD) in thermogenesis was investigated in transgenic mice lacking the mGPD gene (mGPDⴚ/ⴚ). Reared and studied at 22 C, these mice have a small, but significant, reduction (7–10%) in energy expenditure, as evidenced by oxygen consumption (QO2) and food intake, and show signs of increased brown adipose tissue (BAT) stimulation, higher plasma T4 and T3 concentrations, as well as increased uncoupling protein 3 (UCP3) expression in muscle. When acclimated at thermoneutrality temperature (32 C), QO2 decreased in both genotypes, but the difference between them widened to 16%, whereas BAT underwent atrophy, and plasma T4 and T3 levels and UCP3 mRNA decreased, yet T3 and UCP3 persisted at significantly higher levels in mGPDⴚ/ⴚ mice. Such differences disappeared when the mice were rendered hypothyroid. A compensatory role for the observed changes in BAT, thyroid

T

HERMOGENESIS, the generation of heat by the body, is crucial to maintain body temperature in homeothermic species. Thermogenesis can be the obligatory consequence of the energy transformations inherent to life (obligatory thermogenesis) or can be produced as an adaptive response to maintain body temperature in homeothermic species exposed to cold (facultative or adaptive thermogenesis). Obligatory thermogenesis is greater in homeothermic than in poikilothermic species, and such a difference results in part from a more active metabolism (more energy transactions and ATP turnover) and in part from a lower thermodynamic efficiency of the homeothermic machine (1, 2). Both clinical and experimental evidence indicate that thyroid hormone is necessary to realize the higher thermogenic potential of homeothermic species (reviewed in Ref. 3). Thus, thyroid hormone increases ATP turnover and reduces the thermodynamic efficiency of both muscle contraction (4) and muscle ATP synthesis (5). Several lines of evidence suggest a role for mitochondrial glycerol-3-phosphate dehydrogenase (mGPD; EC 1.1.99.5) in thermogenesis. This enzyme is a flavin adenine dinucleotideAbbreviations: BAT, Brown adipose tissue; DHAP, dihydroxyacetone-3-phosphate; G3P, glycerol-3-phosphate; mGPD, mitochondrial glycerol-3-phosphate dehydrogenase; NE, norepinephrine; PTU, propylthiouracil; QO2, oxygen consumption; UCP3, uncoupling protein 3; WT, wild-type.

hormone levels, and UCP3 was investigated with a 2-h cold challenge of 12 C in euthyroid and hypothyroid mice. No hypothermia ensued if the mice had been acclimated at 22 C, but when acclimated at 32 C, euthyroid mGPDⴚ/ⴚ mice became significantly more hypothermic than the wild-type controls. When rendered hypothyroid, this difference was accentuated, and the mGPDⴚ/ⴚ mice developed profound hypothermia (⬃28 vs. 34 C in wild-type mice; P < 0.001). Thus, mGPD-deficient mice have, despite increased plasma T4 and T3, a small, but distinct, reduction in obligatory thermogenesis, which is compensated by increased BAT facultative thermogenesis and by thyroid hormone-dependent mechanisms using other proteins, possibly UCP3. The results support a role for mGPD in thyroid hormone thermogenesis. (Endocrinology 144: 5469 –5479, 2003)

linked dehydrogenase, located on the outer surface of the inner mitochondrial membrane that directly transfers reducing equivalents (H⫹, e⫺) into the complex III of the respiratory chain (6, 7). The interconversion between dihydroxyacetone-3-phosphate (DHAP) and glycerol-3-phosphate (G3P) plays an important role in intermediary metabolism. DHAP serves as a transient acceptor of reducing equivalents generated in the cytoplasm and transferred from NADH [and NADPH (8)] to DAHP, in a reaction catalyzed by the cytoplasmic G3P dehydrogenase (EC 1.1.1.8). The resulting G3P is reoxidized to DHAP by the mGPD. The two G3P dehydrogenases constitute the G3P shuttle, which, along with the malate-aspartate shuttle, transfers reducing equivalents from cytoplasm to mitochondria (9). The G3P shuttle thus allows the rapid generation of ATP in aerobic conditions. Because the reducing equivalents from G3P enter the respiratory chain in complex III, only two, instead of three, ATPs are generated per atom of oxygen or pair of electrons (7), that is the efficiency of ATP generation by this pathway is lower than when the reducing equivalents enter the respiratory chain in complex I. Accordingly, the expression of mGPD is particularly high in tissues that require rapid generation of ATP, such as the flight muscle of insects, sperm cells, and pancreatic ␤-cells, but also in tissues producing heat, such as brown adipose tissue (BAT). The mGPD is also expressed in several other tissues. In the mouse, skeletal muscle is next to

5469

5470

Endocrinology, December 2003, 144(12):5469 –5479

BAT as the tissue most rich in mGPD, followed by brain, kidney, and liver (10). In skeletal muscle the G3P shuttle is much more important than the aspartate-malate shuttle (9). The preferential use of this shuttle in muscle allows the rapid generation of ATP from NADH produced by glycolysis, limiting lactate accumulation and producing more heat than tissues where the malate-aspartate shuttle is more important (e.g. liver). The importance of the G3P shuttle in muscle has been demonstrated in cytoplasmic G3P dehydrogenasedeficient mice (11, 12) and is further demonstrated here. An additional reason to consider a contribution of mGPD to obligatory thermogenesis is its stimulation by thyroid hormone in several tissues (13). Actually, there is a good correlation between the stimulation of oxygen consumption (14) and that of mGPD activity or mRNA by thyroid hormone in rat tissues (13, 15). Moreover, the hormone does not stimulate mGPD in species where it does not stimulate oxygen consumption, despite the presence in these species of thyroid hormone receptors and the control by thyroid hormone of development and other aspects of physiology (16). Although these observations do not establish a cause-effect relationship between thyroid hormone thermogenesis and mGPD, they suggest that the enzyme is somehow necessary for the thermogenic effect of thyroid hormone. To investigate a role for mGPD in thermogenesis and temperature homeostasis, we have performed the studies we report here in transgenic C57BL/6J mice with targeted disruption of the mGPD gene (17). When the present studies were in progress, Brown et al. (18) described the phenotype of mice with targeted disruption of the mGPD gene. Their mGPD⫺/⫺ mice had reduced viability, lower weight, and glycemia, among other differences, with the cognate wildtype genotype, but they did not seem to have a gross thermogenic defect and responded to thyroid hormone as the wild-type phenotype. However, as many biochemical mechanisms contribute to thermogenesis (reviewed in Ref. 19), the possibility of compensations should be considered. The body maintains its core temperature adjusting the level of facultative thermogenesis to the level of obligatory thermogenesis and can increase or decrease heat loss (3, 20). Thus, the lack of hypothermia or of reduced total energy expenditure in transgenic animals lacking a given gene, e.g. mGPD or uncoupling protein 3 (UCP3) (21), does not exclude the participation of that gene in thermogenesis. To define the role of each putative thermogenic gene in temperature homeostasis requires consideration of all factors contributing to thermal balance as well as compensations and responses to challenges. We report here that mGPD-deficient mice have a thermogenic deficiency that activates compensatory mechanisms. Materials and Methods

DosSantos et al. • Glycerol-3-Phosphate Shuttle and Thermogenesis

genome of these cells is 50% C57BL/6J and 50% CBA (22). Targeted cells where then injected in C57BL/6J blastocysts to obtain chimera mice. The mice were crossed with C57BL/6J mice to obtain heterozygous mice, which were backcrossed into the C57BL/6J background. Heterozygous breeding couples with a genetic background greater than 96.9% C57BL/6J (five rounds of backcrossing with this strain) were shipped from Japan to our laboratory to create our own colony by inbreeding. Wild-type (WT; mGPD⫹/⫹) and homozygous mGPD null mice (mGPD⫺/⫺) were obtained by crossing heterozygous (mGPD⫹/⫺) males either with mGPD⫹/⫹ or mGPD⫺/⫺ females. Mice were housed at a maximum of five per cage. They were fed ad libitum and were reared and kept at 22 ⫾ 1 C on a 12-h light, 12-h dark cycle, unless noted otherwise. Genotype was determined by PCR in DNA obtained from the tail tip around the time of weaning (17). Mice were subsequently segregated by genotype and gender for future breeding and/or experiments. The lack of mGPD activity in the mGPD⫺/⫺ genotype was confirmed by direct assay of the enzyme in isolated mitochondria as described previously (23). All experiments described here are comparisons between males homozygous mGPD⫹/⫹ or mGPD⫺/⫺ and were performed in mice 3– 4 months old. Mice selected for an experiment were matched by age within 15 d. As Brown et al. (18), we find that the weight of mGPD⫺/⫺ males is lower than that of WT, although the difference in our mice was less marked. Thus, between 3 and 5 months of age, the difference is 5–10%, so age matching did not result in significant weight differences within experiments. Where indicated below, in experiments testing cold tolerance, mice, in addition, were carefully matched by weight. Hypothyroidism was induced with an iodine-deficient diet containing 0.15% propylthiouracil (PTU), obtained from Harlan Teklad Co. (Indianapolis, IN). Controls were given the same diet lacking PTU and supplemented with potassium iodate. T3 or T4 dissolved in mildly alkaline saline containing 0.1% BSA, or this vehicle, was injected ip or sc in the dose regimens described with individual experiments.

Energy expenditure This was measured by indirect calorimetry in an open-circuit system (Oxymax System, Columbus Instruments, Inc., Columbus, OH). Airflow through airtight cages housing individual mice was kept constant with a computer-operated pump at 0.65 liter/min. O2 and CO2 concentrations were detected with appropriate electrodes in the incoming and outgoing air. The timing and duration of sampling were programmed in a computer. The concentrations of O2 or CO2, along with the weight of the mice and the measuring parameters were fed into the computer that calculated oxygen consumption (QO2) and carbon dioxide production rates per unit of time and weight, as well as the respiratory quotient. An exponent of 0.7 was used for expression of gases exchange by weight, and unless otherwise indicated, QO2 and carbon dioxide production were expressed in milliliters per hour ⫻ 100 g⫺0.7 (20). Animals were allowed to adjust to the chamber for 1 h, and measurements were performed for 2– 6 h to accumulate at least 10 readings/mouse, between 1000 and 1600 h. As animals are not restrained and move freely in the plastic chamber, physical activity, usually consisting of displacements and grooming, was an additional source of variability in the measurements (20). To minimize the effect of activity, usually occurring in spurts, we measured the total amount of O2 consumed over the whole period of observation and calculated the QO2 rate from the slope of the accumulation over time, corrected by weight as indicated above. This approach minimizes the effect of bursts of activity, because these will only shift up the curve, but will not substantially affect the slope unless those spurts occur repeatedly and in sequence. As shown in Fig. 1, A–C, the cumulative oxygen consumption was indeed linear with time. Measurements were made at a constant set temperature ⫾0.1 C, as indicated for each experiment, for which with the system was installed in a walk-in environmental chamber (SureTemp, Raleigh, NC).

Animals All studies reported here were approved by the McGill University animal research committee. Mice with targeted disruption of the mGPD gene were those described by Eto et al. (17). A neomycin resistance gene was substituted for exon 5 in a genomic clone from the BALB/c mouse that contained exons 4, 5, and 6 of the mGPD gene. The resulting DNA, cloned in the appropriate vector, was then inserted by homologous recombination in the genome of TT2 mouse embryonic stem cells. The

Body and BAT temperature Core body temperature was measured with a flexible rectal probe (YSI 423, Yellow Springs Instrument Co., Dayton, OH), which was connected to a high precision thermometer (YSI Precision 4000A thermometer). Animals were not anesthetized or sedated and were restrained for not longer than 30 sec for the measurement. A small rubber ring placed at 2.5 cm of the tip of the probe ensured uniformity in the depth

DosSantos et al. • Glycerol-3-Phosphate Shuttle and Thermogenesis

Endocrinology, December 2003, 144(12):5469 –5479 5471

sured as described previously (26). For histological exam, tissues were placed in buffered 3.7% formaldehyde immediately after harvesting.

RNA isolation and PCR analysis Tissue RNA was extracted with acid guanidinium-phenol-chloroform (27), quantified by absorbance at 260 nm, and stored in diethylpyrocarbonate-treated water at ⫺80 C until further analysis. The integrity of RNA was routinely verified by agarose gel electrophoresis. mRNAs of interest were quantified by competitive RT-PCR, using primers and competitors described below. This method uses a competitor DNA with identical ends to the target cDNA, but with different internal sequence. Such internal sequence is from v-erbB (PCR MIMIC kit, Clontech Laboratories, Inc., Palo Alto, CA) and spans a segment of DNA within ⫾30% the size of the test cDNA product. The design of competitor mimics has been described previously (28 –30). RT was performed on 1 ␮g RNA using random primers. An aliquot of the RT reaction was then submitted to PCR amplification with the appropriate primers and competitors. The primers, sizes of the PCR products, and accession numbers of the target sequences are presented in Table 1. Three pooled RNA aliquots were routinely sham reverse transcribed (i.e. reverse transcriptase omitted) to ensure the absence of products other than those from the reverse transcribed mRNAs. PCR products were resolved in 1% or 2% agarose gel electrophoresis, stained with ethidium bromide, and quantified by video densitometry. The amount of specific mRNAs was calculated from the signal intensity ratio of the test mRNA over that of competitor, corrected by the difference in size between them and by the size of the aliquots, and expressed in attomoles of mRNA per microgram of total RNA. PCR cycles were adjusted to be in the linear part of PCR product accumulation. ␤-Actin mRNA was measured to ensure the uniformity of RNA aliquots (for the validation of these assays, see Refs. 29 and 30).

Western blot analysis of BAT and skeletal muscle mitochondria

FIG. 1. QO2 quantification and variability. The accumulated oxygen consumed was computed from readings made every 17.5 min. A, This panel shows the intermeasurement variability in a same animal. Cumulative oxygen consumed by one WT and one mGPD⫺/⫺ mouse measured on separate consecutive days. B, Combined regression lines from the data in A. The combined slopes of the WT and the mGPD⫺/⫺ mouse differ significantly. Neither the individual measurements nor the combined regression line deviate from linearity. C, Results of an experiment (experiment 3 in Table 2) with the combined measurements of seven mice in each group. Slopes have been expressed per hour, as in Table 2. the probe was introduced into the rectum. The effects of constant jugular infusion of norepinephrine (1.25 nmol/min or 1.14 nmol/␮l at 1.1 ␮l/ min) on BAT temperature was measured as described previously (24) along with rectal temperatures.

Blood and tissue sampling Mice were killed by exsanguination under light isoflurane anesthesia immediately after ending in vivo experiments. Blood was allowed to coagulate at room temperature, and cleanly collected serum was stored at ⫺20 C until analyzed. Total T4 and T3 levels were measured by RIA using commercial kits (Diagnostic Products, Los Angeles, CA) with the modifications described previously to adapt it for use with mouse serum (25). The serum free fatty acid concentration was measured with a commercial kit (NEFA C, Wako Chemicals, Richmond, VA). Tissues were snap-frozen in liquid nitrogen and kept at ⫺80 C until analyzed for mRNAs, DNA, protein, G3P, or enzyme activities. G3P was mea-

Mitochondria were isolated by standard differential centrifugation. All procedures were carried out at 0 – 4 C. BAT was homogenized in 0.32 m sucrose and 3 mm MgCl2 (31), and muscle (tibialis anterior) was homogenized in 180 mm KCl, 10 mm EDTA, 5 mm MgCl2, 1 mm ATP, 50 mm Tris-HCl (pH 7.4), and 0.06% protease inhibitor cocktail (32). Mitochondria were obtained by centrifugation of the 1,000 ⫻ g homogenate at 12,000 ⫻ g for 10 min. The pellet was washed once and submitted to analysis. Protein was assayed with a Bio-Rad protein assay kit (Richmond, CA) using BSA as standard. Twenty (BAT) or 40 ␮g mitochondrial protein was resolved in 12% polyacrylamide SDS-PAGE and electrotransferred to a nitrocellulose membrane. Antirat UCP1 primary antibody has been described (31) and was used at a dilution of 1:2,000. Antihuman UCP3 primary antibody was commercially available (Chemicon, Temecula, CA) and used at a dilution of 1:250. The signals were detected with goat antirabbit IgG and a standard ECL kit (Amersham Pharmacia Biotech, Arlington Heights, IL).

Other tissue measurements DNA and tissue protein content were measured by a standard methods (33, 34). Histological analysis was performed in freshly fixed tissue samples stained with hematoxylin-eosin or Sudan Black (for fat).

Statistical analysis Results are expressed as the mean ⫾ sem. Experiments involving several treatments (i.e. times and doses) were analyzed by ANOVA, followed by post hoc tests for multiple comparisons (Newman-Keuls). Two-way ANOVA was used to compare the effects of treatments on two experimental groups (e.g. two genotypes or thyroid status). Individual means were then compared by Bonferroni’s test if across experimental groups or by Newman-Keuls test if within one experimental group.

Results Measures of energy balance and thermogenesis at 22 C

Mice were reared and housed at the usual temperature of the vivarium (21–23 C), and all measurements were per-

5472

Endocrinology, December 2003, 144(12):5469 –5479

DosSantos et al. • Glycerol-3-Phosphate Shuttle and Thermogenesis

TABLE 1. RT-PCR analysis of various mRNAs: primers and size of PCR products Gene

UCP1 Sense Antisense UCP2 Sense Antisense UCP3 Sense Antisense mGPD Sense Antisense

PCR product size (bp)

Primers

mRNA

Competitor

5⬘-AAGGCCAGGCTTCCAGTACTATTAGGT 5⬘-GGTTTGATCCCATGCAGATGGCTCTG

318

454

5⬘-TTCGACAGTGCTCTGGTATCTCCGACC 5⬘-GCTAGTGCGCACCGCAGCCAGCGCCCA

336

462

5⬘-TACGCCTGGGAACTGGAGGAGAGAGGA 5⬘-TCCTGAGCCACCATCTTCAGCATACAG

354

516

5⬘-GTGCGCAAAATGGATGATA 5⬘-CATCCCCTCTTCTCACTTCAA

291

326

Complete sequences are online at http://ncbi.nlm.nih.gov under the following accession numbers: UCP1, BC012701; UCP2, U69135; UCP3, AF032902; and mGPD, BC021359. TABLE 2. Measures of energy balance and thermogenesis in C57BL/6J mice with targeted disruption of the mGPD gene (mGPD⫺/⫺) compared to WT genotype Genotype WT

Body weight (g) Food intake (g/d䡠100 g⫺0.7) QO2 (ml/h䡠100 g⫺0.7) Experiment 1 Experiment 2 Experiment 3 Rectal temperature (C)

mGPD⫺/⫺

25.6 ⫾ 0.33 (34) 11.08 ⫾ 0.3 (12)

24.3 ⫾ 0.35 (27) 9.95 ⫾ 0.39 (11)

199.8 ⫾ 3.98 (12) 207.1 ⫾ 1.51 (10) 193.37 ⫾ 3.69 (7) 37.96 ⫾ 0.08 (16)

187.2 ⫾ 4.74 (14) 193.4 ⫾ 1.65 (9) 178.79 ⫾ 5.58 (7) 38.02 ⫾ 0.12 (15)

% Diff

⫺5.1 ⫺10.2 ⫺6.3 ⫺6.6 ⫺7.5 ⫹0.15

P

0.0099 0.03 0.057 0.0036 0.0296 0.64

All mice were 3.5 month ⫾ 1 wk old. Animals were reared and maintained at the vivarium temperature (21–23 C), and measurements were made at 22 C. Values are the mean ⫾ SEM, with the number of animals in parentheses. % Diff, Percent difference between WT controls and mGPD-deficient mice.

formed at 22 ⫾ 0.1 C. The thermoneutrality temperature for mice is about 30 C (20); hence, 22 C represents a significant, albeit not extreme, cold stress. This was considered a favorable situation to make evident the recruitment of adaptive mechanisms in the event the mGPD⫺/⫺genotype was associated with a thermogenic defect. Most relevant results are summarized in Table 2. As reported previously (17, 18), body weight was significantly lower in mGPD⫺/⫺ mice, by about 5% (P ⬍ 0.01), which, as reported recently (18), was due not to different body composition but to smaller body size. Food intake was also reduced, even if corrected by weight, by about 10% (P ⫽ 0.03), at variance with Brown et al. (18). The results of measurements of QO2, performed weeks apart, in three separate groups of mice show a small, but significant, reduction in total energy expenditure of about 7% in mGPD⫺/⫺ mice, with remarkable good agreement of the values. Such reduction in energy expenditure is consistent with and not significantly different from the reduction in food intake associated with the lack of mGPD (⫺10.2%). Core body temperature was well maintained in these mice kept in a 22 C environment, confirming the findings presented by Brown et al. (18). This small difference in energy expenditure could result from reduced basal metabolic rate, reduced thermogenesis, or physical activity. Nonexercise physical activity (e.g. grooming and displacement in the cage) has been reported not to be affected by the lack of mGPD in another study (18), so that the difference is more likely to result from reduced basal metabolic rate (i.e. reduced obligatory thermogenesis),

reduced facultative thermogenesis, or both. If a reduction in obligatory thermogenesis is the primary consequence of the lack of mGPD, this is expected to activate facultative thermogenesis (20) and possibly other mechanisms involved in obligatory thermogenesis. In addition, the experimental prevention of such putative compensatory mechanisms could uncover the real impact of the lack of mGPD on thermogenesis and temperature homeostasis. The following experiments and measurements were aimed to investigate these possibilities. Ambient temperature-dependent changes in mGPD⫺/⫺ mice

Thermoneutrality temperature is defined as the ambient temperature at which body temperature can be maintained only with obligatory thermogenesis (20), without the participation of facultative thermogenesis or heat-saving or heatdissipating mechanisms. As indicated above, this temperature is 30 C for mice (20). Consequently, we studied mGPD⫺/⫺ mice after acclimation to 32 C for 3 wk. This temperature, 2 C over the thermoneutrality point, was chosen in the event that modestly reduced thermogenesis in mGPD⫺/⫺ mice could have moved up their thermoneutrality point. Acclimation to 32 C was associated with a marked reduction in QO2, as illustrated in Fig. 2. The results of the three experiments performed at 22 C (Table 2) are shown as a reference. The combined means for QO2 at 22 C were 200.8 ⫾ 3.25 and 187.1 ⫾ 4.3 for the WT and mGPD⫺/⫺ genotypes,

DosSantos et al. • Glycerol-3-Phosphate Shuttle and Thermogenesis

respectively. The mean reduction in QO2 in mGPD⫺/⫺ mice in the three experiments performed at 22 C was 6.9% and remained statistically different despite being three separate experiments (P ⬍ 0.02). Acclimation at 32 C was associated with 26% and 35% decreases (P ⬍ 0.001 for both genotypes) in QO2 in WT and mGPD⫺/⫺ mice, respectively. The values at 32 C were 147.5 ⫾ 7.4 and 122.7 ⫾ 8.1. Not only did the difference between the two genotypes remain significant, but it also widened to ⫺16.9%. Table 3 shows the serum concentrations of T4 and T3 in both genotypes. At 22 C, both hormone concentrations were significantly higher (⬃20%) in mGPD⫺/⫺ than in WT mice. With acclimation to 32 C, both hormone concentrations decreased, T3 by as much as 35% (P ⬍ 0.001), and T4 by less and not significantly. At 32 C, there was no difference between the two genotypes in T4 concentrations, but T3 continued to be significantly higher in mGPD⫺/⫺ mice. To prove that the differences in serum thyroid hormone concentrations were not due to changes in either plasma binding or clearance, hypothyroid mice (3 wk on a low iodine-PTU diet) were given two daily sc injections of T4 totaling 15 ng/g䡠d for 1 wk. They were killed approximately 16 h after the last injection. As shown in the last row of Table 3, the levels of T4 and T3 were in the euthyroid range and were not affected by genotype.

FIG. 2. Effect of ambient temperature on QO2 in WT and mGPD⫺/⫺ mice. The mean (thick horizontal line) and SEM are shown. Measurements in mice acclimated at 22 C are those shown in Table 2 with their combined mean and SEM (200.8 ⫾ 3.25 and 187.1 ⫾ 4.3 for the WT and mGPD⫺/⫺ genotypes, respectively). Measurements at 32 C were made after a 3-wk acclimation to this ambient temperature. Results are the mean of two consecutive measurements in seven mice per genotype. The values at 32 C were 147.5 ⫾ 7.4 (n ⫽ 14) and 122.7 ⫾ 8.1 (n ⫽ 14). Data were analyzed by ANOVA, followed by Newman-Keuls test for multiple comparisons.

Endocrinology, December 2003, 144(12):5469 –5479 5473

Figure 3 shows the effects of genotype and ambient temperature on muscle UCP3 expression. UCP3 mRNA levels were significantly increased (⬃2-fold) in skeletal muscles of mGPD⫺/⫺ mice. The differences were highly significant (P ⬍ 0.0001). UCP3 mRNA decreased significantly with acclimation at 32 C, but the differences between genotypes persisted (P ⬍ 0.01). UCP3 (protein) was also approximately 2-fold elevated in the tibialis anterior muscle of mGPD⫺/⫺ mice when expressed as concentration, i.e. per mitochondrial protein or per whole muscle. Others have also found a good correlation between UCP3 mRNA and protein, including their responses to thyroid hormone (35). As it has been reported that UCP3 mRNA levels can be increased by free fatty acids, we measured their concentrations in serum samples from these mice. They were virtually identical [WT, 788.2 ⫾ 28.48 (n ⫽ 5); mGPD⫺/⫺, 804.4 ⫾ 32.82 (n ⫽ 7) ␮Eq/liter]. BAT is the main site of facultative nonshivering thermogenesis in small rodents and is expected to be active in mice acclimated at 22 C, 8 C below thermoneutrality (36, 37). Furthermore, the degree of stimulation of BAT will depend on the need of facultative thermogenesis, being inversely related to obligatory thermogenesis. For example, in conditions where obligatory thermogenesis is reduced or increased, such as hypothyroidism or hyperthyroidism, the tissue will show, respectively, signs of enhanced (38) or reduced stimulation (39). The histological appearance of BAT in WT and mGPD⫺/⫺ mice is shown in Fig. 4. There is a striking difference between mice maintained at 22 C and those acclimated at 32 C for 3 wk. At 32 C, reflecting the quiescence of BAT, fat is accumulated in large droplets, much as it is found in white adipose tissue. At 22 C, the tissue shows evidence of being active, with its typical multilocular fat accumulation, more cytoplasm, and bigger nuclei. At this temperature there are differences between the two genotypes: BAT of mGPD⫺/⫺ mice has lower fat content, smaller and more numerous fat droplets per cell (confirmed with Sudan Black staining; data not shown), as well as more cytoplasm and bigger nuclei. The mass of interscapular BAT normalized to 100 g body weight was not different between genotypes, but the DNA and protein contents per gram of tissue were significantly higher in the mGPD⫺/⫺ mice (Table 4), consistent with the greater cell density and lower fat content per cell in BAT of these mice, as shown in Fig. 4. UCP1 was increased by 22% (P ⫽ 0.041). These changes are indicative of chronic increased sympathetic stimulation (40).

TABLE 3. Effect of genotype and ambient temperature on serum T4 and T3 concentrations in C57BL/6J mice Ambient temperature (C)

22 32 Temperature effect %⌬ (P value) 22, Hypothyroid ⫹ T4a

Serum T4 (␮g/dl) WT

GPD

⫺/⫺

% Diff. (P value)

Serum T3 (ng/dl) WT

GPD⫺/⫺

% Diff. (P value)

2.67 ⫾ 0.11 (19) 2.57 ⫾ 0.25 (7) ⫺3.5 (NS)

3.21 ⫾ 0.18 (20) 2.74 ⫾ 0.41 (5) ⫺14.6 (NS)

20.1 (0.018) 5.3 (NS)

70.6 ⫾ 1.95 (9) 47.7 ⫾ 3.15 (12) ⫺32.4 (⬍0.001)

82.9 ⫾ 2.97 (7) 53.2 ⫾ 3.75 (12) ⫺35.8 (⬍0.001)

17.5 (0.002) 11.7 (0.033)

3.09 ⫾ 0.18 (5)

2.84 ⫾ 0.21 (5)

⫺8 (NS)

74.2 ⫾ 4.46 (5)

62.6 ⫾ 3.88 (5)

⫺15.6 (NS)

Values are the mean ⫾ SEM, with the number of animals in parentheses. % Diff., Percent difference between WT controls and mGPD-deficient mice. a Mice made hypothyroid on a low iodine diet plus PTU (see Materials and Methods) were injected sc with T4 (15 ng/g䡠d), divided into two doses for 1 wk, and killed approximately 16 h after the last injection.

5474

Endocrinology, December 2003, 144(12):5469 –5479

DosSantos et al. • Glycerol-3-Phosphate Shuttle and Thermogenesis

FIG. 3. Upper panel, UCP3 mRNA levels in skeletal muscle (tibialis) of euthyroid WT and mGPD⫺/⫺ mice acclimated at 22 or 32 C. Data were analyzed by ANOVA, followed by Newman-Keuls test. Lower panel, UCP3 protein levels in the same muscle of mice acclimated at 22 C. Data were analyzed by unpaired t test. UCP3 mRNA is expressed as attomoles (attmol) per microgram of reverse transcribed total RNA; the corresponding protein is shown as arbitrary densitometric units (A.U.) present in either 40 ␮g mitochondrial protein (left) or the whole muscle (right).

FIG. 4. Histological appearance of BAT of WT and mGPD⫺/⫺ mice reared and maintained at 22 or acclimated at 32 C for 3 wk.

Effect of ambient temperature and thyroid status on acute cold adaptation

Together, the previous observations suggest that mGPD⫺/⫺ mice have lower obligatory thermogenesis, so that when acclimated at 22 C they need more BAT facultative thermogenesis to maintain their body temperature. Furthermore, we observed an increase in serum T4 and T3 concentrations in mGPD⫺/⫺ mice as well as an increase in UCP3, suggesting the recruitment of other adaptive mechanisms. The histological aspect of BAT in mGPD⫺/⫺ suggests that the tissue is under a sustained higher adrenergic stimulation than in WT mice. The functional reflection of this is the preparedness of BAT to respond to adrenergic stimulation, what has being called recruitment by Nedergaard’s group (41). The acute thermogenic response of BAT was then tested by measuring the increase in temperature after a short infusion with norepinephrine (NE). Groups of mGPD⫺/⫺ and WT mice (n ⫽ 5 each) were infused with NE through a

jugular catheter at 1.25 nmol/min for 30 min under anesthesia, whereas temperature was recorded with a BAT and a rectal probe (24). The weight of the animals was nearly identical (WT, 29.9 ⫾ 1.9; mGPD⫺/⫺, 30.5 ⫾ 1.3 g). Results are shown in Fig. 5. Both BAT and rectal temperature increased linearly with the time of infusion, and the increments were higher in BAT than in the rectum. In both anatomical locations, the increase in temperature was higher in mGPD⫺/⫺ than in WT mice (P ⫽ 0.01 and P ⫽ 0.03, respectively). Note that the temperature increments are higher in BAT than in the rectum at all times. The same rate of saline infusion did not affect either BAT or rectal temperature (data not shown). As shown above (Fig. 4) and previously by us (42) and others (43), the acclimation of animals to thermoneutrality results in a marked reduction of BAT sympathetic stimulation. The expression of the UCP1 gene is promptly stopped after placing rats at this temperature (42), and UCP1 decreases markedly over the course of several days (43). Thus, acclimation of animals to thermoneutrality provides an opportunity to test temperature homeostasis without the participation of BAT. When small rodents so acclimated are acutely exposed to cold, their body temperature will be defended only by shivering thermogenesis for a day or more while BAT thermogenic capacity is being restored (44). As depicted in Fig. 6A, when mice were acclimated at 22 C, i.e. when BAT was kept active, and they were acutely placed at 12 C, they maintained their core body temperature for 2 h, and the curves of both genotypes were superimposable. In contrast, when acclimated at 32 C for 3 wk, the same cold challenge induced hypothermia in both groups, but this was prompt and more marked (by ⱖ1 C) in mGPD⫺/⫺ mice (P ⬍ 0.001; Fig. 6B). These results indicate that in the presence of quiescent BAT, mGPD⫺/⫺ mice have a greater drop in tem-

DosSantos et al. • Glycerol-3-Phosphate Shuttle and Thermogenesis

Endocrinology, December 2003, 144(12):5469 –5479 5475

TABLE 4. Measures of chronic BAT stimulation in WT and mGPD⫺/⫺ mice acclimated at 22 C Genotype

Weight (mg/100 g body weight) DNA (mg/g BAT) Protein (mg/g BAT) UCP1 (AU)a

WT

mGPD⫺/⫺

384 ⫾ 20 (6) 2.91 ⫾ 0.12 (5) 151 ⫾ 3.4 (6) 689 ⫾ 54 (5)

392 ⫾ 10 (6) 3.97 ⫾ 0.33 (6) 175 ⫾ 5.9 (5) 843 ⫾ 31 (5)

% Change

P

2 36 16 22

0.75 0.018 0.005 0.041

Values are the mean ⫾ SEM, with the number of animals in parentheses. a UCP1 protein as assessed by Western blot. AU, arbitrary densitometric units in 20 ␮g mitochondrial protein.

FIG. 5. Effect of NE infusion on BAT and rectal temperatures of WT and mGPD⫺/⫺ mice. Animals were reared and maintained at 22 C at all times. NE was infused through an indwelling catheter placed in a jugular vein and were infused at a rate of 1.25 nmol/min. Animals were anesthetized with ketamine and xylazine, and kept over a heating pad set at 30 C during the procedure to prevent hypothermia and avoid shivering. See Materials and Methods and Ref. 24 for details. Data were analyzed by two-way ANOVA. There was a nearly linear, time-dependent increase in temperature in BAT and rectum (P ⬍ 0.0001). The effect of genotype on BAT and rectal temperature increments was significant at P ⫽ 0.01 and P ⫽ 0.03, respectively. n ⫽ 5 for each genotype.

perature than WT mice, supporting a role for BAT-increased stimulation in the maintenance of core temperature at 22 C and also suggesting that shivering is less efficient than in WT mice as a means of maintaining core temperature. As mGPD⫺/⫺ mice acclimated at 32 C for 3 wk maintain higher serum T3 levels and have more UCP3 mRNA in muscle than the WT genotype, we repeated the experiment in hypothyroid animals acclimated at 32 C (Fig. 6C). Core temperature decreased more and faster in these mice than in euthyroid animals (note the difference in the y-axis), but the descent was faster and greater in mGPD⫺/⫺ mice. After 90 min at 12 C, the core temperature was significantly lower in these mice; they became severely hypothermic at the end of the experiment, with a core temperature of approximately 28 C (vs. ⬃34 C in WT mice; P ⬍ 0.001). Effect of the thyroid status and T3 on energy expenditure and UCP expression

Our results show that mGPD⫺/⫺ mice have modestly reduced energy expenditure despite a 15–20% increase in plasma thyroid hormone levels, suggesting that mGPD is necessary for thyroid hormone thermogenesis. On the other hand, Brown et al. (18) reported that mGPD-deficient mice rendered hyperthyroid increased their QO2 similarly to the WT genotype. To test whether this was the case in our model, we performed the following experiment. Ten WT and 10

mGPD⫺/⫺ mice were rendered hypothyroid for 4 wk as described in Materials and Methods, and half of them were treated with daily ip injections of 20 ng/g body weight T3, equivalent to approximately 5 times the daily production rate for 20 d, to induce moderate thyrotoxicosis. Controls were injected simultaneously with the corresponding volume of vehicle. Mice were kept at 22 C at all times. QO2 was measured at various times during the treatment. Results are shown in Fig. 7. Data obtained in euthyroid WT and mGPD⫺/⫺ mice (Table 2 and Fig. 2) are depicted for reference. Hypothyroidism was associated with the expected reduction in QO2, whereas T3 treatment increased QO2 within 3 d to a plateau approximately 40% higher than the euthyroid level. There was however no significant influence of genotype on the response of QO2 to T3. Thus, our mGPD⫺/⫺ model responded as that of Brown et al. (18) to moderate hyperthyroidism, indicating that that mGPD is not essential under these circumstances and that other proteins could compensate for the lack of the enzyme. Hypothyroidism was associated with a decrease in UCP3 mRNA in both tibialis anterior and soleus muscles. In tibialis, values were 124 ⫾ 25 and 131 ⫾ 17 attomoles/␮g RNA for WT and mGPD⫺/⫺ mice, whereas in soleus the corresponding values were 81 ⫾ 26 and 76 ⫾ 12 (Fig. 8), compared with values of over 300 attomoles/␮g RNA in muscle of euthyroid mice kept at 22 C, as shown in Fig. 3. The differences between

5476

Endocrinology, December 2003, 144(12):5469 –5479

DosSantos et al. • Glycerol-3-Phosphate Shuttle and Thermogenesis

FIG. 7. Effect of thyroid status and T3 treatment on oxygen consumption of WT and mGPD⫺/⫺ mice maintained and measured at 22 C. Hypothyroidism was induced with an iodine-deficient, PTU-containing diet given for approximately 25 d, as described in Materials and Methods. **, Euthyroid values are the mean of the three experiments shown in Table 2, as depicted in Fig. 2, and are shown for reference (P ⫽ 0.02). The dose of T3 was 20 ng/g body weight䡠d for the indicated times on the abscissa. Genotype had no significant effect on the QO2 of hypothyroid mice or their responses to T3.

Discussion

FIG. 6. Influence of temperature of acclimation and thyroid status on the core temperature of WT and mGPD⫺/⫺ submitted to an acute 2-h exposure to 12 C. In all cases, core temperature was measured with a flexible rectal probe introduced 2.5 cm in the rectum, as described in Materials and Methods. Mice were matched by weight. A, Mice reared at 22 C. B, The same mice were acclimated for 3 wk to 32 C and then subjected to the cold challenge. C, Mice were made hypothyroid with a low iodine diet containing 0.15% propylthiouracil (see Materials and Methods) and acclimated at 32 C for 3 wk. Statistical significance was determined by two-way ANOVA, followed by Bonferroni’s test to compare the means of both genotypes at each time point. A, Neither time at 12 C nor genotype had a significant effect on core temperature. Both variables had a significant effect in B and C (P ⬍ 0.001). For individual time points, **, P ⬍ 0.01; ***, P ⬍ 0.001.

the two genotypes in either tibialis or soleus were not significant. In both genotypes and both muscles, UCP3 mRNA responded significantly to T3 treatment, but the response was significantly greater in mGPD⫺/⫺ than in WT mice (Fig. 8, A and B). Regarding UCP2, T3 did not stimulate expression in WT mice, but it did in mGPD⫺/⫺ mice and to a higher level than in WT mice (Fig. 8, C and D). For mGPD mRNA, WT mice showed the expected changes with thyroid status (15), and this was not affected by other experimental manipulations (data not shown).

We have shown here that mice lacking mGPD have a modest, but significant, reduction in total energy expenditure, as can be judged by both food intake and indirect calorimetry. Reared at 22 C, mGPD⫺/⫺ mice have signs of chronically increased stimulation of BAT, increased BAT responsiveness to NE, elevation of serum T4 and T3, and a 2-fold increase in the level of UCP3 mRNA in muscle, findings altogether suggestive that these animals are under some degree of cold stress. The acclimation of the animals to a temperature slightly over the thermoneutrality temperature for mice (20) reduced or obliterated these differences. Furthermore, as these potential compensatory mechanisms were progressively eliminated by acclimating mice at 32 C and then rendering them hypothyroid, the thermogenic defect of the mGPD⫺/⫺ mice became increasingly evident as revealed by a 2-h exposure to 12 C cold challenge. A central finding in the present report is that despite a significant increase in plasma thyroid hormone concentrations, mGPD⫺/⫺ mice had no increase, but a modest reduction, in energy expenditure. This is a strong evidence for a role for mGPD in thyroid hormone thermogenesis. This does not contradict the finding by Brown et al. (18) that moderate thyrotoxicosis stimulates QO2 to the same level in both mGPD-deficient and WT mice. We have here confirmed that mGPD is indeed not essential for such a level of thyroid function to stimulate the metabolic rate. This is probably an indication that supraphysiological levels of thyroid hormone can recruit other thermogenic mechanisms, as will be discussed further. Such results also lend support to the idea that multiple mechanisms participate in thermogenesis, and that they may be recruited in a dose-dependent manner. The participation of several mechanisms in heat production is in agreement with the finding that the deletion of several potentially thermogenic genes is not associated with gross thermogenic defects (21, 45, 46). It could be argued that the presence of compensations

DosSantos et al. • Glycerol-3-Phosphate Shuttle and Thermogenesis

Endocrinology, December 2003, 144(12):5469 –5479 5477

FIG. 8. Effect of hypothyroidism and T3 (20 ng/g䡠d) for the indicated time on UCP3 (A and B) and UCP2 mRNA (C and D) in two skeletal muscles, tibialis anterior and soleus, of WT and mGPD⫺/⫺ mice. Hypothyroidism induction and duration and T3 treatments were exactly as described in Fig. 7. Data were analyzed by ANOVA, followed by Newman-Keuls test for multiple comparisons. Note the low levels of UCP3 mRNA in the hypothyroid mice compared with the levels in euthyroid mice in Fig. 3.

derives from the smaller size of the mGPD⫺/⫺ mice. It is known that smaller animals have higher thermogenesis per unit of body mass because the surface to volume ratio is higher (20). However, if this were the case, mGPD⫺/⫺ mice should have higher, not lower, energy expenditure. Besides, in the majority of the experiments the difference in weight between WT and mGPD⫺/⫺ mice was not significant. Anticipating that body size could be a factor in the hypothermia induced by the cold challenge, we were particularly careful to match the animals by weight, which was done without creating a significant difference in age between the two genotypes. Therefore, the magnitude of the cold stress or the resulting hypothermia cannot be attributed to bias in weight or age in these experiments. BAT is probably the most important site of nonshivering facultative thermogenesis in small mammals. The tissue is highly efficient and is expected to make a significant contribution to overall thermogenesis in small rodents such as mice reared at 22 C, 8 C colder than their thermoneutrality (20, 37). A measure of BAT contribution to energy expenditure at 22 C was the 26% and 35% reductions in QO2 in WT and mGPD⫺/⫺ mice, respectively, observed when ambient temperature was raised to 32 C (P ⬍ 0.001 for both genotypes; Fig. 2), along with the atrophy of the tissue. The greater contribution of BAT to thermogenesis at 22 C in mGPD⫺/⫺ than in WT mice is supported by two observations. First, there was a more vigorous response of BAT to acute NE infusion. This observation, along with histological appearance, density of cells and protein, and a modest, but signif-

icant, elevation of UCP, indicate that BAT activity is greater in mGPD⫺/⫺ mice. Second, these mice suffered more profound hypothermia when BAT was rendered quiescent by acclimating them at 32 C and then they were acutely exposed to 12 C for 2 h (Fig. 6). As mentioned, animals acclimated to thermoneutrality defend their body temperature with shivering when acutely exposed to cold (44). The need for shivering relates inversely to obligatory thermogenesis and the capacity of the body to decrease heat dissipation (44). A failure to curtail heat losses can be excluded as the sole explanation because if that were the case then energy expenditure would be increased, not decreased. Altogether, the reduced energy expenditure of mGPD⫺/⫺ mice at both 22 and 32 C indicates that obligatory thermogenesis is reduced, and the hypothermia observed during acute cold exposure suggests further that shivering is not as efficient in these animals as in the WT mice. Shivering is a form of intense muscle activity, which is expected to generate large amounts of reducing equivalents, the oxidation of which via mGPD produces heat and rapidly generates ATP to support continued shivering. It is likely that of all tissues, muscle thermogenesis is more affected by the lack of mGPD. Not only does muscle contain abundant mGPD, but the malate-aspartate shuttle is of minor importance (9). We have found that G3P is 2- to 3-fold elevated in muscle, but not in liver, of mGPD⫺/⫺ mice (Alfadda, A., R. A. DosSantos, K. Eto, T. Kadowaki, and J. E. Silva, manuscript in preparation), further supporting the idea of an important role for mGPD in muscle more than in liver. On the other hand, we

5478

Endocrinology, December 2003, 144(12):5469 –5479

have confirmed that the malate-aspartate shuttle is of minor importance in muscle compared with liver, as aminooxyacetate, an inhibitor of the malate-aspartate shuttle (47), clearly reduces respiration in liver, but not in muscle, with succinate or G3P as substrates (St. Laurent, S., H. Marrif, R. A. DosSantos, and J. E. Silva, manuscript in preparation). Another important observation in mGPD⫺/⫺ mice was the approximately 20% elevation of serum T4 and T3 levels over WT levels when the animals were acclimated to 22 C. The hypothalamic-pituitary-thyroid axis in rodents has been long recognized as sensitive to cold stimulation (48, 49). Thus, the increase in T4 and T3 concentrations seen in mGPD⫺/⫺ mice acclimated at 22 C is also an indication of cold stress. The fact that both T4 and T3 are elevated suggests increased thyroidal secretion, but there may be a simultaneous increase in T4to T3 conversion as well. The concentrations of both iodothyronines were temperature dependent, in agreement with being caused by central thyroid stimulation. Although the difference in serum T4 between the two genotypes ceased to be statistically significant at 32 C, the difference in T3 persisted, although less pronounced than at 22 C (17% vs. 11% at 32 C), indicating that even at this temperature some degree of thermal adaptation is necessary. The contribution of the higher T3 level to temperature homeostasis is supported by the more profound hypothermia the mGPD⫺/⫺ than in WT mice when rendered hypothyroid. One thermogenic mechanism that could possibly compensate for the lack of mGPD is UCP3. Deletion of the UCP3 gene by Gong et al. (21) was not associated with a gross thermogenic defect, and the temperature and QO2 responses to T3 were preserved. However, this group apparently did not investigate the possibility of compensatory responses to the lack of UCP3 (21), so that this protein, for reasons explained above, cannot be excluded as participating in thyroid hormone thermogenesis. The positive responses of the gene to cold, ␤-adrenergic agonists, and thyroid hormone favor a role for UCP3 in thermogenesis (50); hence, the increase in its mRNA in mGPD⫺/⫺ mice should be considered another possible compensation. Other investigators, notably de Lange et al. (35) and Lanni et al. (51), have found a close correlation between UCP3 and thyroid hormone stimulation of QO2. Not only UCP3 mRNA, but also the protein, were elevated about 2-fold in mGPD⫺/⫺ mice in the studies here reported. Moreover, the level of UCP3 mRNA is temperature dependent, and the increase in this mRNA in mGPD⫺/⫺ mice is thyroid hormone dependent. Hypothyroidism caused a reduction in UCP3 mRNA to an equally low level in both genotypes, whereas the treatment with T3 was associated with a vigorous response of this mRNA, which was significantly greater in mGPD⫺/⫺ mice. Such responses suggest that thyroid hormone is essential for the expression of the gene and that its effect may be potentiated by another factor, presumably the sympathetic nervous system (50, 52). Regarding UCP2, the same reasoning applies. Even though no gross thermogenic defect has been reported in UCP2-deficient mice (45, 46), it cannot be excluded on these bases that UCP2 and its greater stimulation by T3 in mGPD⫺/⫺ mice are playing a role in reducing the impact of mGPD deficiency. Although free fatty acids can stimulate the expression of both UCP2 and UCP3 (53, 54), such a mechanism is unlikely to

DosSantos et al. • Glycerol-3-Phosphate Shuttle and Thermogenesis

play any role in the increased responsiveness to T3 of UCP2 and UCP3 or in the baseline elevation of UCP3 mRNA and protein. Thus, FFA serum levels were not significantly affected by the genotype, nor did they increase more in response to T3 in mGPD⫺/⫺ mice (data not shown). Sympathetic activity and responsiveness to catecholamines, on the other hand, are known to be enhanced in the muscles of animals undergoing cold stress (55, 56) as may be the case in mGPD⫺/⫺ mice acclimated at 22 C. To establish the mechanism of the elevation of UCP3 and the increased response to thyroid hormone in mGPD⫺/⫺ mice certainly requires further systematic investigation. In summary, the experimental results presented here suggest that mGPD plays a significant role in temperature homeostasis. Such a role is evidenced by a slight reduction in energy expenditure and the recruitment of several potentially compensatory mechanisms, such as chronic sympathetic stimulation of BAT, increased concentrations of T4 and T3, as well as increased UCP3 gene expression in skeletal muscle. The reduced energy expenditure despite an elevation of plasma thyroid hormone concentrations supports a role for the enzyme in the thermogenesis induced by physiological levels of thyroid hormones. The functional obliteration of BAT, as in acclimation to thermoneutrality, is associated with a reduced capacity of mGPD⫺/⫺ mice to tolerate moderate cold for 2 h, strongly supporting the idea that there is a compensated thermogenic defect. Acknowledgments We thank Erika Vignault and Hai Ling Zhang for technical help. Received June 2, 2003. Accepted August 20, 2003. Address all correspondence and requests for reprints to: J. Enrique Silva, M.D., Jewish General Hospital, 3755 Chemin de la Coˆ te-Ste-Catherine, Room E104, Montre´ al, Que´ bec, Canada H3T 1E2. E-mail: [email protected]. This work was supported by a grant from the Canadian Institutes of Health Research (MOP 44071, to J.E.S.).

References 1. Else PL, Hulbert AJ 1981 Comparison of the “mammal machine” and the “reptile machine:” energy production. Am J Physiol 240:R3–R9 2. Woledge RC 1989 Energy transformations in living muscle. In: Wieser W, Gnaiger E, eds. Energy transformations in cells and organisms. Stuttgart, New York: Verlag; 36 – 45 3. Silva JE 1995 Thyroid hormone control of thermogenesis and energy balance. Thyroid 5:481– 492 4. Leijendekker WJ, van Hardeveld C, Elzinga G 1987 Heat production during contraction in skeletal muscle of hypothyroid mice. Am J Physiol 253:E214 – E220 5. Lebon V, Dufour S, Petersen KF, Ren J, Jucker BM, Slezak LA, Cline GW, Rothman DL, Shulman GI 2001 Effect of triiodothyronine on mitochondrial energy coupling in human skeletal muscle. J Clin Invest 108:733–737 6. Klingenberg M 1970 Localization of the glycerol-phosphate dehydrogenase in the outer phase of the mitochondrial inner membrane. Eur J Biochem 13:247– 252 7. Dawson AG 1979 Oxidation of cytosolic NADH formed during aerobic metabolism in mammalian cells. Trends Biochem Sci 4:171–176 8. Fahien LA, Laboy JI, Din ZZ, Prabhakar P, Budker T, Chobanian M 1999 Ability of cytosolic malate dehydrogenase and lactate dehydrogenase to increase the ratio of NADPH to NADH oxidation by cytosolic glycerol-3-phosphate dehydrogenase. Arch Biochem Biophys 364:185–194 9. Lehninger AL, Nelson DL, Cox MM 1993 Principles of biochemistry. 2nd ed. New York: Worth 10. Koza RA, Kozak UC, Brown LJ, Leiter EH, MacDonald MJ, Kozak LP 1996 Sequence and tissue-dependent RNA expression of mouse FAD-linked glycerol-3-phosphate dehydrogenase. Arch Biochem Biophys 336:97–104 11. MacDonald MJ, Marshall LK 2000 Mouse lacking NAD⫹-linked glycerol

DosSantos et al. • Glycerol-3-Phosphate Shuttle and Thermogenesis

12.

13. 14. 15.

16. 17.

18.

19. 20. 21.

22. 23. 24.

25. 26. 27. 28.

29.

30.

31. 32.

33. 34.

phosphate dehydrogenase has normal pancreatic ␤ cell function but abnormal metabolite pattern in skeletal muscle. Arch Biochem Biophys 384:143–153 MacDonald MJ, Marshall LK 2001 Survey of normal appearing mouse strain which lacks malic enzyme and Nad⫹-linked glycerol phosphate dehydrogenase: normal pancreatic ␤ cell function, but abnormal metabolite pattern in skeletal muscle. Mol Cell Biochem 220:117–125 Lee YP, Lardy HA 1965 Influence of thyroid hormones on l-␣-glycerophosphate dehydrogenases and other dehydrogenases in various organs of the rat. J Biol Chem 240:1427–1436 Barker SB, Klitgaard HM 1952 Metabolism of tissues excised from thyroxineinjected rats. J Physiol 170:81– 86 Gong DW, Bi S, Weintraub BD, Reitman M 1998 Rat mitochondrial glycerol3-phosphate dehydrogenase gene: multiple promoters, high levels in brown adipose tissue, and tissue-specific regulation by thyroid hormone. DNA Cell Biol 17:301–309 Weirich RT, Schwartz HL, Oppenheimer JH 1987 An analysis of the interrelationship of nuclear and plasma triiodothyronine in the sea lamprey, lake trout, and rat: evolutionary considerations. Endocrinology 120:664 – 677 Eto K, Tsubamoto Y, Terauchi Y, Sugiyama T, Kishimoto T, Takahashi N, Yamauchi N, Kubota N, Murayama S, Aizawa T, Akanuma Y, Aizawa S, Kasai H, Yazaki Y, Kadowaki T 1999 Role of NADH shuttle system in glucoseinduced activation of mitochondrial metabolism and insulin secretion. Science 283:981–985 Brown LJ, Koza RA, Everett C, Reitman ML, Marshall L, Fahien LA, Kozak LP, MacDonald MJ 2002 Normal thyroid thermogenesis but reduced viability and adiposity in mice lacking the mitochondrial glycerol phosphate dehydrogenase. J Biol Chem 277:32892–32898 Silva JE 2001 The multiple contributions of thyroid hormone to heat production. J Clin Invest 108:35–37 Gordon CJ 1993 Temperature regulation in laboratory rodents. New York: Cambridge University Press Gong DW, Monemdjou S, Gavrilova O, Leon LR, Marcus-Samuels B, Chou CJ, Everett C, Kozak LP, Li C, Deng C, Harper ME, Reitman ML 2000 Lack of obesity and normal response to fasting and thyroid hormone in mice lacking uncoupling protein-3. J Biol Chem 275:16251–16257 Yagi T, Tokunaga T, Furuta Y, Nada S, Yoshida M, Tsukada T, Saga Y, Takeda N, Ikawa Y, Aizawa S 1993 A novel ES cell line, TT2, with high germline-differentiating potency. Anal Biochem 214:70 –76 Gardner RS 1974 A sensitive colorimetric assay for mitochondrial ␣-glycerophosphate dehydrogenase. Anal Biochem 59:272–276 Ribeiro MO, Carvalho SD, Schultz JJ, Chiellini G, Scanlan TS, Bianco AC, Brent GA 2001 Thyroid hormone-sympathetic interaction and adaptive thermogenesis are thyroid hormone receptor isoform-specific. J Clin Invest 108: 97–105 Weiss RE, Murata Y, Cua K, Hayashi Y, Seo H, Refetoff S 1998 Thyroid hormone action on liver, heart, and energy expenditure in thyroid hormone receptor ␤-deficient mice. Endocrinology 139:4945– 4952 Lang G 1985 l(-)-Glycerol 3-phosphate. In: Bergmeyer HU, Bergmeyer J, Grasse M, eds. Metabolites I: carbohydrates. Basel: VCH Press; vol 6:525–531 Chomczynski P, Sacchi N 1987 Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162: 156 –159 Lo´pez-Solache I, Marie V, Vignault E, Camirand A, Silva JE 2002 Regulation of uncoupling protein-2 mRNA in L6 myotubules. I. Thiazolidinediones stimulate uncoupling-2 gene expression by a mechanism requiring ongoing protein synthesis and an active MAP kinase. Endocrine 19:197–208 Boivin M, Camirand A, Carli F, Hoffer LJ, Silva JE 2000 Uncoupling protein 2 and 3 messenger RNAs in adipose tissue and skeletal muscle of healthy males: variability, factors affecting expression and relation to measures of metabolic rate. J Clin Endocrinol Metab 85:1975–1983 Soumano K, Desbiens S, Rabelo R, Bakopanos E, Camirand A, Silva JE 2000 Glucocorticoids inhibit the transcriptional response of the uncoupling protein-1 gene to adrenergic stimulation in a brown adipose cell line. Mol Cell Endocrinol 165:7–15 Bianco AC, Silva JE 1987 Intracellular conversion of thyroxine to triiodothyronine is required for the optimal thermogenic function of brown adipose tissue. J Clin Invest 79:295–300 Jimenez M, Yvon C, Lehr L, Leger B, Keller P, Russell A, Kuhne F, Flandin P, Giacobino JP, Muzzin P 2002 Expression of uncoupling protein-3 in subsarcolemmal and intermyofibrillar mitochondria of various mouse muscle types and its modulation by fasting. Eur J Biochem 269:2878 –2884 Giles KW, Myers A 1965 An improved diphenylamine method for the estimation of deoxyribonucleic acid. Nature 206:93 Bradford MM 1976 A rapid and sensitive method for the quantitation of

Endocrinology, December 2003, 144(12):5469 –5479 5479

35. 36. 37. 38. 39. 40. 41. 42. 43.

44.

45.

46.

47. 48. 49. 50. 51.

52.

53.

54.

55. 56.

microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72:248 –254 de Lange P, Lanni A, Beneduce L, Moreno M, Lombardi A, Silvestri E, Goglia F 2001 Uncoupling protein-3 is a molecular determinant for the regulation of resting metabolic rate by thyroid hormone. Endocrinology 142:3414 –3420 Foster DO, Frydman ML 1978 Brown adipose tissue: the dominant site of nonshivering thermogenesis in the rat. Experientia 32(Suppl):147–151 Cannon B, Nedergaard J 1985 The biochemistry of an inefficient tissue: brown adipose tissue. Essays Biochem 2:110 –164 Mory G, Ricquier D, Pesquies P, Hemon P 1981 Effects of hypothyroidism on the brown adipose tissue of adult rats: comparison with the effects of adaptation to the cold. J Endocrinol 91:515–524 Sundin U 1981 GDP binding to rat brown fat mitochondria: effects of thyroxine at different ambient temperature. Am J Physiol 241:C134 –C139 Girardier L, Seydoux J 1986 Neural control of brown adipose tissue. In: Trayhurn P, Nicholls DG, eds. Brown adipose tissue. London: Arnold; 122–151 Dicker A, Cannon B, Nedergaard J 1995 Cold acclimation-recruited nonshivering thermogenesis: the Syrian hamster is not an exception. Am J Physiol 269:R767–R774 Rehnmark S, Bianco AC, Kieffer JD, Silva JE 1992 Transcriptional and posttranscriptional mechanisms in the uncoupling protein mRNA response to cold. Am J Physiol 262:E58 –E67 Desautels M, Himms-Hagen J 1980 Parallel regression of cold-induced changes in ultrastructure, composition and properties of brown adipose tissue mitochondria during recovery of rats from acclimation to cold. Can J Biochem 58:1057–1068 Nedergaard J, Golozoubova V, Matthias A, Shabalina I, Ohba K, Ohlson K, Jacobsson A, Cannon B 2001 Life without UCP1: mitochondrial, cellular and organismal characteristics of the UCP1-ablated mice. Biochem Soc Trans 29: 756 –763 Arsenijevic D, Onuma H, Pecqueur C, Raimbault S, Manning BS, Miroux B, Couplan E, Alves-Guerra MC, Goubern M, Surwit R, Bouillaud F, Richard D, Collins S, Ricquier D 2000 Disruption of the uncoupling protein-2 gene in mice reveals a role in immunity and reactive oxygen species production. Nat Genet 26:435– 439 Zhang CY, Baffy G, Perret P, Krauss S, Peroni O, Grujic D, Hagen T, Vidal-Puig AJ, Boss O, Kim YB, Zheng XX, Wheeler MB, Shulman GI, Chan CB, Lowell BB 2001 Uncoupling protein-2 negatively regulates insulin secretion and is a major link between obesity, ␤ cell dysfunction, and type 2 diabetes. Cell 105:745–755 MacDonald MJ 1982 Evidence for the malate aspartate shuttle in pancreatic islets. Arch Biochem Biophys 213:643– 649 Yamada T, Kajihara A, Onaya T, Kobayashi I, Takemura Y, Shichijo K 1965 Studies on acute stimulatory effect of cold on thyroid activity and its mechanism in the guinea pig. Endocrinology 77:968 –976 Montoya E, Seibel MJ, Wilber JF 1975 Thyrotropin-releasing hormone secretory physiology: studies by radioimmunoassay and affinity chromatography. Endocrinology 96:1413–1418 Gong DW, He YF, Karas M, Reitman M 1997 Uncoupling protein-3 is a mediator of thermogenesis regulated by thyroid hormone, ␤3-adrenergic agonists, and leptin. J Biol Chem 272:24129 –24132 Lanni A, Beneduce L, Lombardi A, Moreno M, Boss O, Muzzin P, Giacobino JP, Goglia F 1999 Expression of uncoupling protein-3 and mitochondrial activity in the transition from hypothyroid to hyperthyroid state in rat skeletal muscle. FEBS Lett 444:250 –254 Larkin S, Mull E, Miao W, Pittner R, Albrandt K, Moore C, Young A, Denaro M, Beaumont K 1997 Regulation of the third member of the uncoupling protein family, UCP3, by cold and thyroid hormone. Biochem Biophys Res Commun 240:222–227 Son C, Hosoda K, Matsuda J, Fujikura J, Yonemitsu S, Iwakura H, Masuzaki H, Ogawa Y, Hayashi T, Itoh H, Nishimura H, Inoue G, Yoshimasa Y, Yamori Y, Nakao K 2001 Up-regulation of uncoupling protein 3 gene expression by fatty acids and agonists for PPARs in L6 myotubes. Endocrinology 142:4189 – 4194 Weigle DS, Selfridge LE, Schwartz MW, Seeley RJ, Cummings DE, Havel PJ, Kuijper JL, Beltrandelrio H 1998 Elevated free fatty acids induce uncoupling protein 3 expression in muscle: a potential explanation for the effect of fasting. Diabetes 47:298 –302 Dulloo AG, Young JB, Landsberg L 1988 Sympathetic nervous system responses to cold exposure and diet in rat skeletal muscle. Am J Physiol 255: E180 –E188 Shiota M, Masumi S 1988 Effect of norepinephrine on consumption of oxygen in perfused skeletal muscle from cold-exposed rats. Am J Physiol 254:E482– E489