AR Signaling Required for Diet-Induced Thermogenesis and Obesity Resistance Eric S. Bachman, Harveen Dhillon, Chen-Yu Zhang, Saverio Cinti, Antonio C. Bianco, Brian K. Kobilka, and Bradford B. Lowell

Animal Care-Care of the mice was within institutional IACUC guidelines. Mice were housed in groups of 2-4 at 22-24°C using a 14-h light 10-h dark cycle with chow food (Teklad F6 Rodent Diet 8664, 4.05 kcal/g, 3.3 kcal/gram metabolizable energy, 12.5 % kcal from fat, Harlan Teklad, Madison, WI. www.harlan.com) and water provided ad libitum. For food intake studies, male mice were housed individually. Large, intact pellets of food were provided every 4 days in order to reduce spillage, and cages were changed every time that food weight was measured. Mice were killed by CO₂ narcosis. Mice were fed chow (above) or high fat diet (Research Diets, New Brunswick, NJ, high sucrose diet D12331, 5.56 kcal/g, 5.56 kcal/g metabolizable energy, 58 % kcal from fat; www.researchdiets.com).

Derivation of experimental mouse lines -The parental strains were: mice homozygous for disruption of both 1 and 2 AR genes (provided by one of the authors (B.K.) (S1), and mice homozygous for 3AR disruption (S2). Genotypes of parental strains in Jackson Labs nomenclature are Ardb1,2tm1Bkk (beta 1,2 knockout) and Ardb3tm1Lowl (beta 3 knockout) . Two lines of wt and -less mice were derived and matings within these individual lines of mice produced the experimental animals (fig. S1A).

Body composition-6-10 week old mice were anesthetized using ketamine and whole animals were analyzed using DEXA analysis (Lunar Pixi, Janesville, WI).

Oxygen consumption was measured by indirect calorimetry. Mice were placed at room temperature (22-24° C) in 1.0 L chambers in an OXYMAX system 4.93 (Columbus Instruments, Columbus, OH) with a settling time of 100 s and a measuring time of 50 s and the reference as room air. Food and water provided ad libitum. In vivo response to 紡gonists ((-)isoproterenol, Sigma-Aldrich) was performed in 6-8 week old mice that were acclimated to thermoneutrality for 3 days prior to experiments.

Serum chemistries

Insulin was measured by ELISA (Chrystal Chem., Inc., Chicago Illinois) Leptin was measured by radioimmunoassay (Amersham Biosciences, Piscataway N.J.) Serum thyroid hormone measurements were performed using radioimmunoassay (Linco, Inc., Racine, WI).

Type II deiodinase (D2) activity was measured by one of the authors (A.B.) as described (S3). Dissection of interscapular BAT was performed after euthanasia by trimming all visible WAT and weighing tissue immediately. Brown adipose tissue histology was analyzed by rapidly euthanizing mice in CO2, carefully dissecting brown adipose tissue and placing immediately into phosphate-buffered formalin for 24 hours. Sections were then made and stained with H and E at this facility. Immunohistochemistry was performed as described (S4).

Response to cold exposure-Mice were placed at 4°C in filtered cages with food and water provided ad libitum. Rectal temperatures were taken every thirty minutes using YSI model 43 telethermometer with series 500 probe (Yellow Springs Instrument Co. Inc, Ohio). The experiment was stopped when rectal temperatures reached 25°C, which was usually after 4 hours in 僕ess mice.

Western blotting/BAT biochemistry-BAT was dissected and homogenized in Tris/EDTA/Aprotinin/Leupeptin/PMSF buffer as described. Protein content was determined using the Bradford assay. Western blots using 40 ug of total BAT protein were performed and approximately equal protein loading was determined by Ponceau staining. Blots were incubated using goat anti-mouse UCP-1 antibody at 1:1000 (Santa Cruz Biotechnology, Santa Cruz, CA) and developed using goat anti-rabbit secondary antibody and the ECL+ Chemiluminescent Kit (Amersham Biosciences, Piscataway N.J.).

In vitro BAT studies-Isolated brown adipocytes were harvested as previously described (S5). Briefly, 6-8 week old male mice that had been housed at room temperature were euthanized and BAT was trimmed of white adipose tissue. BAT from wt or 僕ess mice was pooled, minced, and placed in 2 mL/mouse DMEMF-12 (Invitrogen, Carlsbad CA)/4% BSA (Calbiochem, La Jolla CA) with 2 mg/mL type I collagenase (Worthington Biochemicals, Lakewood, NJ) for 10-15 Food intake during this 5-day period was similar between groups (3.23+/-.227 g/day in 僕ess mice versus 3.29 +/-.074 in wild type). minutes until cells were dissociated. Cells were washed twice in medium/BSA alone and allowed to settle. Cell viability was assessed by trypan blue exclusion.

Activity/core body temperature- Mice were anesthetized using ketamine, and a coupled sensor/transmitter (TA11PA-40, Data Sciences International, St. Paul, Minnesota) was implanted subcutaneously in the abdominal cavity 10 days prior to data collection and signal processing, which was done using DataQuest III (Data Sciences International).

High fat feeding

3 month-old male mice that had been fed chow diet were switched to a high-fat sucrose-containing diet containing 58 % kcal from fat (see above). Food intake was measured every 3 days for 14 days total during the period of weight gain. Cages were changed every 3 days at the time that food intake was measured.

In another series of experiments, mice were acclimated to oxygen consumption containers for 24 hours, fed a chow diet for 24 hours to achieve baseline values, and then high-fat diet was provided. Daily, individual weight and food intake were assessed, and data were derived for 23/24 hours for 6 continuous days. Fresh food was provided daily.

Statistical analysis Data were analyzed using unpaired Student痴 t-test or one-way ANOVA. All results are presented as the mean +/- se. Differences were accepted as significant when p<0.05.

Supporting Text

Body temperatures tended to be lower in -less mice versus wt mice (38.03 +/-0.17 °C versus 38.5 +/-0.3 in wt mice), however this failed to reach statistical significance (p=0.51, n=12 in each group).

Thyroid hormone levels in 僕ess mice were similar to wt mice (T4=5.25 +/-0.31 g/dL versus 5.2 +/-0.71 in wt mice, T3=111 +/-4.9 ng/dL vs 109 +/-10.2, in wt mice). Physical activity, measured over 7 days using implanted activity transmitters, showed a trend toward increased activity in -less (3.97+/-0.58 movements/min) compared with wt mice (3.04 +/-0.22, p=0.09). Thus, the decreased energy expenditure in 僕ess mice was not due to reduced physical activity.

Supporting Figures

Supplemental Figure 1. Derivation of wt and 僕ess mice. A. Breeding scheme for generation of 僕ess and wt control mice. Lines of wild-type (wt) control and 僕ess mice were derived from the same parents over 2 generations of breeding. For the parental strains, description of genotype is listed in the first row within each box, and the second row lists genetic background of original strains (1, 2). The first breeding produced a generation of mice that were heterozygous for disruption of all 3 ARs (mixed), and these were subsequently bred to select 2 lines each of mice that were either homozygous wild type for all 3 ARs (wt), homozygous null for 1 and 3-adrenergic receptor genes but homozygous wt for 2AR (-1,3-less, not shown) or homozygous null for 1, 2 and 3-adrenergic receptor genes (-less).

Medium version | Full size version

Supplemental Figure 2. Brown adipose histology from line 2. BAT from 8-week old male mice representing the genotypes shown was fixed, sectioned and stained. Seen is Hematoxylin and Eosin staining at 10X power with brown adipose tissue and adjacent white adipose tissue (WAT) seen in the upper corners of each panel.

Medium version | Full size version

Supporting Tables

Supplemental Table 1. Body composition of wt and 僕ess mice at 6 and 15 weeks of age as measured by DEXA (see Materials and Methods above). All data give as mean +/- SEM.
Weeks of age/genotype (n)	Body Weight g	Body Fat %	Fat Mass g	Lean Mass g
6/Wt (5)	22.3+/-0.74	16.53+/-1.9	3.75+/-0.35	19.34+/-1.0
6/Beta-less (5)	22.7+/-0.58	20.58+/-0.75*	4.45+/-0.36*	17.65+/-0.09*
15/Wt (5)	27.3 +/-0.4	16.2 +/- 1.9	4.44 +/-0.5	21.4 +/-0.6
15/ Beta-less (5)	32.2 +/- 1.2*	22.2 +/-0.9*	7.18 +/-0.6*	22.4 +/-0.7

* p<0.05 compared to age-matched wt control mice

Supplemental Table 2. Body weight gain on chow and high-fat diet. Weight gain in 僕ess mice is similar to weight gain seen in leptin-deficient ob/ob mice (S6). Caloric (food) intake, expressed as kcal/day, increased equivalently in both 僕ess and wt mice on a high-fat diet. This was calculated from available food energy (Materials and Methods above) given daily food intake (chow diet fed wt mice consumed 5.05 g/day versus 5.04 g/day in beta-less mice, and high fat fed wt mice consumed 3.26 g/day and beta-less mice ate 3.26 g/day).
Strain/diet	Wt gain (g) Over 7 weeks	Food intake (kcal/day)
Wt chow (n=8)	2.5+/-0.38	16.2+/-0.21
-less chow (8)	4.2+/-0.26	16.2+/-0.33
Wt high-fat (8)	6.7+/-0.34	18.1+/-0.33
-less high fat (8)	20.1+/-2.46	18.1+/-0.35
ob/ob (ref S6)	20.2

Supporting References and Notes

S1. D. K. Rohrer, A. Chruscinski, E. H. Schauble, D. Bernstein, B. K. Kobilka, J. Biol. Chem. 274, 16701 (1999).

S2. V. S. Susulic, et al. J. Biol. Chem. 270, 29483 (1995).

S3. Curcio et al. J. Biol. Chem. 276, 30183 (2001).

S4. Cinti et al. Endocrinology 138, 797 (1997).

S5. Carrvalho, S.D. et al. Endocrinology 137, 5519 (1996).

S6. C. Y. Zhang, et al., Cell 105, 745-55. (2001).