Adrenergically stimulated blood flow in brown ...

Am J Physiol Endocrinol Metab 308: E822–E829, 2015. First published March 3, 2015; doi:10.1152/ajpendo.00494.2014.

Adrenergically stimulated blood flow in brown adipose tissue is not dependent on thermogenesis Gustavo Abreu-Vieira,1 Carolina E. Hagberg,2 Kirsty L. Spalding,2 Barbara Cannon,1 and Jan Nedergaard1 1

Department of Molecular Biosciences, The Wenner-Gren Institute, Stockholm University, Stockholm, Sweden; and 2Department of Cell and Molecular Biology, Karolinska Institute, Stockholm, Sweden Submitted 30 October 2014; accepted in final form 26 February 2015

Abreu-Vieira G, Hagberg CE, Spalding KL, Cannon B, Nedergaard J. Adrenergically stimulated blood flow in brown adipose tissue is not dependent on thermogenesis. Am J Physiol Endocrinol Metab 308: E822–E829, 2015. First published March 3, 2015; doi:10.1152/ajpendo.00494.2014.—Brown adipose tissue (BAT) thermogenesis relies on blood flow to be supplied with nutrients and oxygen and for the distribution of the generated heat to the rest of the body. Therefore, it is fundamental to understand the mechanisms by which blood flow is regulated and its relation to thermogenesis. Here, we present high-resolution laser-Doppler imaging (HR-LDR) as a novel method for noninvasive in vivo measurement of BAT blood flow in mice. Using HR-LDR, we found that norepinephrine stimulation increases BAT blood flow in a dose-dependent manner and that this response is profoundly modulated by environmental temperature acclimation. Surprisingly, we found that mice lacking uncoupling protein 1 (UCP1) have fully preserved BAT blood flow response to norepinephrine despite failing to perform thermogenesis. BAT blood flow was not directly correlated to systemic glycemia, but glucose injections could transiently increase tissue perfusion. Inguinal white adipose tissue, also known as a brite/beige adipose tissue, was also sensitive to cold acclimation and similarly increased blood flow in response to norepinephrine. In conclusion, using a novel noninvasive method to detect BAT perfusion, we demonstrate that adrenergically stimulated BAT blood flow is qualitatively and quantitatively fully independent of thermogenesis, and therefore, it is not a reliable parameter for the estimation of BAT activation and heat generation. blood flow; brown adipose tissue; high-resolution laser-Doppler imaging; thermogenesis; uncoupling protein 1. THE SEARCH FOR THERAPIES AGAINST OBESITY has focused on ways to alter the balance between caloric intake and energy expenditure (19). Caloric intake is a function of appetite, food composition, digestion, and absorption, whereas expenditure incorporates a wide range of metabolic processes that require energy and produce heat. From a therapeutic perspective, increasing energy expenditure has traditionally been associated with physical activity. However, since the identification of brown adipose tissue (BAT) in adult humans (16, 22), stimulation of nonshivering thermogenesis has become a promising strategy for the control of obesity. In consequence, means to determine BAT activity become relevant, and blood flow could be considered an interesting measure. Indeed, the identification of BAT as the main site for nonshivering thermogenesis in mammals was based on measurements of BAT blood flow variations during adrenergic

Address for reprint requests and other correspondence: J. Nedergaard, Dept. of Molecular Biosciences, The Wenner-Gren Institute, The Arrhenius Laboratories F3n, Stockholm University, SE-106 91, Stockholm, Sweden (e-mail: [email protected]). E822

stimulation (12, 17). Local blood flow rates and arterial-venous differences of O2 and CO2 correlate well with the capacity for whole body, adrenergically stimulated thermogenesis (11). Therefore, decreased O2 tension in BAT caused by nonshivering thermogenesis could be thought to be a direct triggering factor for the measured changes in tissue perfusion. The implication of this is that blood flow has generally been assumed to be a reliable parameter to estimate BAT thermogenesis. In an attempt to further elucidate questions regarding the relation between BAT blood flow and thermogenesis, we have applied high-resolution laser-Doppler imaging (HR-LDI) to assess noninvasively the adrenergically stimulated blood flow in BAT. Here, we evaluate the effect of environmental temperature acclimation on the blood flow of BAT and inguinal white adipose tissue (iWAT), the effect of blood glycaemia on BAT blood flow, and the effect of norepinephrine (NE) on BAT blood flow in mice lacking uncoupling protein 1 (UCP1). Through this, we demonstrate that adrenergic control of BAT blood flow can occur fully independently of nonshivering thermogenesis, making blood flow rates not a reliable indicator of BAT thermogenic activity. MATERIALS AND METHODS

Animal husbandry. Adult NMRI mice of both sexes aged !10 wk were single-caged and housed either at standard room temperature (21°C), within their thermoneutral zone (30°C), or in the cold (4°C) for !3 wk. Before the cold acclimation, mice passed through a 2-wk preacclimation period at 18°C (14). UCP1-knockout (KO) mice (7) back-crossed into a C57Bl/6 background for at least 10 generations were housed at 21°C. The following conditions were applied for all animals: light-dark cycles were 12:12 h, lights on at 8 AM, and chow diet (R70; Labfor, Kimstad, Sweden) and water were offered ad libitum. All mice used in the experiments were born and raised in our facilities and kept single-caged during the experimental period. With the exception of the dose-response experiment (Fig. 1H), every mouse was used for a single experiment. Animal protocols were approved by the Animal Ethics Committee of the North Stockholm Region. Histological staining of vessels. Interscapular BAT (iBAT) from mice housed at 21°C was dissected and postfixed in 4% paraformaldehyde for 48 h, subsequently processed for paraffin embedding using standard procedures, and thereafter immunostained for podocalyxin to visualize the vascular network. Briefly, 17-!m-thick sections were incubated at 4°C overnight with goat anti-podocalyxin primary antibody (no. AF1556; R & D Systems), whereafter horseradish peroxidase-conjugated donkey anti-goat secondary antibody (Jackson ImmunoResearch) was applied for 3 h at room temperature. Peroxidase activity was detected using a 3,3=-diaminobenzidine peroxidase substrate kit (Vector Laboratories); no signal was detected on sections stained with secondary antibody only (Fig. 1D). Staining was

0193-1849/15 Copyright © 2015 the American Physiological Society

http://www.ajpendo.org

E823

REGULATION OF BROWN ADIPOSE TISSUE BLOOD FLOW

A

Baseline

B

Adrenergically stimulated 10.0 7.5 5.0 2.5 0.0 HR-LDI signal (a.u.)

C

H 6

2

-20

-10 0 10 20 30 Time from injection (min)

300

200

100

40

0 Saline

10 100 1000 Norepinephrine (µg/kg)

I

F

10

HR-LDI signal (a.u.)

E

NE

4

0

D

HR-LDI signal (% of baseline)


G

8 NE

6

4 Cold-acclimated Room temperature Thermoneutrality

2 0

-20


40

Fig. 1. Measurement of interscapular brown adipose tissue (iBAT) blood flow and its relation to ambient temperature and body composition. A: basal blood flow in the interscapular region (1.6 cm2) visualized by high-resolution laser-Doppler imaging (HR-LDI) in a cold-acclimated, anesthetized mouse. The BAT region of interest is outlined (dashed trapezoid). B: HR-LDI image 5 min post-norepinephrine (NE) injection. C: histological visualization of a transverse section of interscapular adipose tissue, with both lobes of BAT in the center and white adipose tissue (WAT) in the laterals. D: microscopy of BAT without secondary antibody staining. E and F: podocalyxin staining of the vascular network in adjacent BAT (E) and WAT (F). Scale bars in D–F, 200 !m. G: example of quantification of HR-LDI signal in the iBAT of a mouse acclimated to room temperature at baseline and post-NE injection. H: dose-response curve of NE on iBAT blood flow. I: quantification of HR-LDI data from mice acclimated to thermoneutrality, room temperature, or cold before and after NE injection. In H and I, data points represent means # SE of 4 –7 mice. AU, arbitrary units.

photographed using the tile scan function of an LSM700 microscope at "5 or "20 magnification. Adrenergic stimulation. NE bitartrate (A9512-1G; Sigma-Aldrich, St. Louis, MO) was injected ip routinely at a dose of 1 mg/kg, the dose necessary for reaching maximal vascular response in iBAT of NMRI mice acclimated to 21°C (as presented in Fig. 1H). Blood flow measurements. For the analysis of BAT blood flow, all mice were placed at 21°C for 1 h prior to experimentation to avoid the carryover effects of different acclimation temperatures. After the mice had been anesthetized (75 mg/kg ip pentobarbital sodium), the fur over the interscapular or inguinal region was carefully removed with the use of a veterinary cordless trimmer (Aesculap Isis GT420; B. Braun Vet Care), and the animals were placed over a heated pad in the prone [for interscapular BAT (iBAT) studies] or supine (for iWAT) position. Pentobarbital sodium was the drug of choice for anesthesia because it does not impair BAT function (24). In certain experiments,

body temperature was measured simultaneously with blood flow measurements with the use of a type T thermocouple rectal probe coupled to a digital thermometer (BAT-12; Physitemp Instruments, Clifton, NJ). At the start of the experiment, the heated pad set point was defined so that mice would maintain a stable body temperature of 36.0 # 0.5°C. This relatively low setting was selected to minimize hyperthermia caused by NE-induced thermogenesis. Blood flow was measured by HR-LDI (Moor Instruments). Routinely, measurements were done at the speed of 4 ms/pixel over an area of 1.6 cm2. At this setting, the time for image acquisition was $3.5 min, and scan resolution was 10 pixels/mm2. Body weight was measured before blood flow studies, and body composition was accessed immediately afterward by MRI (EchoMRI, Houston, TX). Image analysis. Acquired images were analyzed with MoorLDI Image Processing software (research version 5.3). The instrument was calibrated with the two reference standards (one being zero) provided

AJP-Endocrinol Metab • doi:10.1152/ajpendo.00494.2014 • www.ajpendo.org

E824


by the manufacturers. Because of the standards calibration, values are comparable between different experiments within the same settings. A single region of interest (ROI) was set to comprise both lobes of iBAT and used for all mice in a given experimental series. Another ROI was defined for iWAT based on the anatomic position of the adipose tissue deposit along the inguinal ligament contralateral to the ip injection site. Quantified values represent the mean flux (a function of blood cell concentration and movement speed) inside the ROI. Baseline perfusion level was defined as the average value between the last three measured points (15 min) before stimulation (e.g., NE injection). Peak blood flow was quantified at the time point where the mean signal inside the ROI was the highest during the 40 min following stimulation. Because there is no scale for the conversion of laser-Doppler values to absolute flow (e.g., ml/min), results are presented as arbitrary units or percent variation from baseline. The images were not treated (e.g., deconvolved) before quantification. In the images presented in the figures, a mild smoothening function was applied for clarity purposes. Indirect calorimetry. For the determination of whole body oxygen consumption, mice were anesthetized as described above and placed inside calorimeters (INCA System; Somedic, Hörby, Sweden) that kept a stable temperature of 30°C (14). After a baseline had been recorded for 20 min, NE was injected ip, and measurements were continued for 40 min. Data are presented as oxygen consumption in ml·min%1·mouse%1. Blood glucose. Mice were injected ip with the standardly used dose for glucose tolerance tests (2 g/kg dissolved in saline) or with an isovolumetric injection of the vehicle. Blood glucose was measured with a standard glucometer (Accu-check Aviva; Roche) from a droplet collected from a small incision made at the tip of the tail. Statistical analysis. Comparison between groups at a single time point was performed by Student’s t-test. For the analysis of interdependence of body weight, composition, and blood flow, as well as blood glucose levels, each data set was initially tested for normality by Shapiro-Wilk test and then analyzed by Pearson (for data sets with a normal distribution) or Spearman correlation (on nonnormally distributed data sets). Statistical significance was assumed for P values &0.05. Points in line graphics represent means # SE, and sample number is indicated in the figure legends. RESULTS AND DISCUSSION

HR-LDI allows for noninvasive visualization of subcutaneous adipose tissue blood flow in, e.g., iBAT with flexible temporal resolution and without the need to inject contrastenhancing agents. Using this technique, we found that blood flow variations in the interscapular adipose tissue could be easily analyzed over the course of 1 h, and the response kinetics between different groups could be clearly delineated. The anatomic location of the scanned area of the mouse, as well as representative images of basal unstimulated iBAT blood flow, is shown in Fig. 1A. In response to NE stimulation, the HR-LDI signal increased significantly throughout the iBAT region, and the lobes of iBAT became visible (Fig. 1B). Histological analysis and podocalyxin staining of the interscapular vascular network displayed the vast capillary bed that permeates iBAT (Fig. 1, C and E). Because of the high vascular density in the tissue (as seen by comparing the unstained iBAT in Fig. 1D with podocalyxin staining in Fig. 1E), all blood flow data were quantified as mean perfusion in the region of interest that comprises BAT. Of note is that we found a remarkable distinction between the amount of vascularization in adjacent brown (Fig. 1E) and white (Fig. 1F) regions of the interscapular adipose tissue [as observed previously by Fawcett (9)].

Adrenergic stimulation increases BAT blood flow. Images of the area comprising the iBAT of a single mouse were taken at 5-min intervals. The mean signal in the tissue was then quantified and presented in a curve for visualization of the kinetics of adrenergically stimulated blood flow (Fig. 1G). During the first measurements, prior to adrenergic stimulation, the signal was very stable. Following NE injection, an immediate and steady increase was detected during the following 10 min. From that point on, a plateau was reached, indicating a stably increased perfusion rate in the tissue. Because the HR-LDI method bases its measurements on several parameters, including blood cell concentration, this method is more appropriate for comparisons between groups of treatments than for the determination of absolute BAT blood flow. Therefore, the blood flow unit is referred to here as the HR-LDI signal, and apparent fold changes should not be interpreted as absolute blood flow variation. Rather, it should be assumed that the fold changes in blood flow are substantially higher than what is apparent here due to an unquantifiable basal signal. To examine the modulating effect of NE on blood flow, multiple experiments similar to that depicted in Fig. 1E were performed with different doses of NE (Fig. 1H). The maximal effect was measured as an increase from the unstimulated baseline and plotted as a dose-response curve. Saline injection in these anesthetized animals did not cause significant changes in the measured blood flow. Notably, the dose capable of increasing blood flow to its maximum (1 mg/kg) was the same as that classically used for experimental measurement of whole body thermogenic capacity (14). Therefore, this dose was taken as the standard for all following experiments. Temperature acclimation modulates NE-induced iBAT blood flow. BAT has a high level of plasticity when an organism undergoes cold acclimation. This is made evident by increased UCP1 protein levels and an increased thermogenic response to NE as well as angiogenesis (4, 34). Therefore, we measured the dynamic responses of blood flow in iBAT of mice previously acclimated to different environmental temperatures. In mice housed at 21°C, NE injection caused the iBAT blood flow to increase markedly over the unstimulated nominal baseline (Fig. 1I). In mice acclimated to thermoneutrality, baseline values were only marginally lower than those of the animals acclimated to 21°C, and the peak response was somewhat lower than that seen in 21°C-acclimated mice. In coldacclimated mice, baseline values were twice as high as those of mice housed at 21 or 30°C. This higher basal blood flow is likely a reflection of the increased angiogenesis in BAT. In the same mice, NE induced a very marked increase over the unstimulated blood flow baseline. Based on the higher baseline level, in cold-acclimated mice, angiogenesis may be the most important factor allowing a greater blood flow to BAT both at rest and during NE stimulation. The observed changes are in agreement with expected adaptions in blood supply to a tissue that is thermogenically adapted to the different environmental temperature demands. Body composition does not affect HR-LDI measurements. Laser penetration can vary depending on the composition and optical properties of the visualized tissue. The possibility of blood flow measurements being affected by body composition could create a technical drawback in the application of this


E825


method to the analysis of BAT blood flow in mice of different fatness. Therefore, we analyzed the correlation between the unstimulated blood flow levels measured by HR-LDI in iBAT of mice and the total body weight or specific elements of body composition (lean and fat mass) of those mice. Baseline blood flow data of NMRI mice used in Figs. 1I and 4B was compared with their body composition, which was measured by MRI immediately after the HR-LDI measurements. Mice housed at 21 and 30°C were treated as a single group, as no differences were found in blood flow in the unstimulated state between both groups, whereas mice housed at 4°C were analyzed separately because of their higher basal blood flow (as seen in Fig. 1I). We found no significant effect of either body weight (Fig. 2A) or fat mass (Fig. 2B) on the basal perfusion rates of iBAT. Lean tissue mass correlated weakly with the HR-LDI signal in mice housed at 21 and 30°C (r2 ' 0.208, P ' 0.049) but not in mice housed in the cold. Previous studies using a variety of methods have investigated the effects of body adiposity on iBAT blood flow, with a focus on the biological implications of obesity for BAT function. Blood flow in BAT has been suggested to be decreased in obese humans (25), whereas in obese mice contradictory effects of obesity have been reported (6, 31). In our analysis, unstimulated iBAT blood flow was unaltered in fatter animals. From a methodological perspective, this supports the suitability of our system for the analysis of mice with variable degrees of adiposity. NE-stimulated BAT blood flow is independent of thermogenesis. In a cold environment, activated BAT is able to combust large amounts of lipids and glucose by dissipating the mitochondrial proton gradient through UCP1, thus generating heat (4). During this process, the tissue has to be supplied with a sufficient amount of blood delivering oxygen and macronutrients while removing carbon dioxide and warm blood into the systemic circulation. Therefore, blood flow may be a limiting factor for brown fat-dependent thermogenesis (21, 34). It has been assumed that the increased blood flow during thermogenesis is triggered by the metabolic demands of the tissue (i.e., O2 requirements) (10). To test this hypothesis, we studied BAT blood flow in wild-type and UCP1-KO mice. Due to the absence of UCP1, brown adipocytes from UCP1-KO mice are unable to produce heat (20), and thus these mice lack NEinduced adaptive nonshivering thermogenesis (14). Correspondingly, BAT in wild-type mice experience local hypoxia

B

5



A

4 3

2 Cold Room temperature Thermoneutrality

1 0

20

30

40 50 Total body weight (g)

60

due to high oxygen consumption, whereas in UCP1-KO mice this is not the case (34). However, the absence of thermogenesis does not seem to impair neoangiogenesis, as indicated by an identical vascularity of UCP1-KO and wild-type mice (34). We first confirmed the impairment in NE-stimulated thermogenesis in the UCP1-KO mice that were acclimated to 21°C. For this, whole body oxygen consumption in response to NE was measured by indirect calorimetry (Fig. 3A), body temperature variations were analyzed with a rectal probe (Fig. 3B), and the ventilatory rate was visually assessed (Fig. 3C). Taken together, the impaired oxygen consumption of the UCP1-KO mice, the lower rise in body temperature, and the decreased ventilatory rate in response to NE clearly demonstrated the absence of nonshivering thermogenesis in these mice. Subsequently, HR-LDI was employed to measure BAT blood flow with or without adrenergic stimulation (Fig. 3D). In wild-type mice, the iBAT region was localized and the average HR-LDI signal measured in the tissue. Following NE injection, a clear rise in signal intensity was detected in the area representing the lobes (Fig. 3D, top). UCP1-KO mice were analyzed in the same manner. After localization of the tissue, NE was injected and the signal change representing blood flow measured (Fig. 3D, bottom). Surprisingly, indistinguishable responses were recorded in wild-type and UCP1-KO mice. Quantification of the average tissue flow (Fig. 3E) indicated no significant differences in basal perfusion of UCP1-KO or wild-type mice. Peak blood flow in response to NE injection was equally unaltered, as was the shape of the response curves. Similar comparisons of wild-type and UCP1-KO mice acclimated to thermoneutrality, which is believed to represent a condition that is more representative of a modern human environment, provided qualitatively identical results (Fig. 4). Taken together, these results indicate that although UCP1 is essential for nonshivering thermogenesis, NE-induced blood flow in iBAT is completely preserved in the absence of UCP1-mediated thermogenesis and oxygen consumption. Using contrast-enhanced ultrasound, Baron et al. (3) reported that NE is able to increase blood flow in UCP1-ablated mice. However, in their hands, KO animals demonstrated a diminished blood flow response compared with wild-type mice. This discrepancy with our results can possibly be explained by a difference in protocols. Whereas our mice were breathing spontaneously during the experiment, Baron et al. (3) intubated and mechanically ventilated their mice under hyper-

5 4 3 2 1

0

r2 = 0.208 p = 0.049

0

5

10 Fat weight (g)

15

20

25 30 Lean weight (g)

35

Fig. 2. Body composition and laser-Doppler signal. Correlation between mouse total body weight (A) or body composition (B), with the signal measured as BAT blood flow by HR-LDI. Only the solid line represents significant correlation. Fat weight is represented by triangles and lean weight by squares. AJP-Endocrinol Metab • doi:10.1152/ajpendo.00494.2014 • www.ajpendo.org

E826


O2 consumption (ml/min)

Wildtype UCP1-KO

B ***

40

39

1.0 Tb (°C)

NE

0.5

C

Wild type UCP1 KO

38

NE

37

*

*** ** **

36

0.0

35

-20


D

Baseline

40

-20

-10

0 10 20 30 Time from injection (min)

Ventilatory rate (vpm)

A 1.5

200

HR -LDI signal (a.u.)

Wildtype

3 2

UCP1-KO

1

50 0

E 4

*

100

40

20 min post NE

NE

150

* ** *

Wild type UCP1KO -20


40

4 NE

3 2 1 0

0

*

Wildtype UCP1-KO -20



40

Fig. 3. Blood flow and thermogenesis in uncoupling protein 1 (UCP1)-ablated mice acclimated to 21°C. A–C: measurements of whole body oxygen consumption (A), body temperature (B), and ventilatory rate (C) in wild-type and UCP1-knockout (UCP1-KO) mice. D: HR-LDI visualization of blood flow in iBAT of wild-type and UCP1-KO mice before and after NE injection; the defined region of interest used for image quantification is shown in the dashed trapezoids. E: quantification of blood flow in wild-type and UCP1-KO mice. Data points represent means # SE of 4 –5 mice. Where relevant, statistical significance between groups is indicated as follows: *P & 0.05; **P & 0.01; ***P & 0.001.

A 1.5

B

O 2 consumption (ml/min)

Wildtype UCP1-KO

1.0

Tb (°C)

NE

37

NE *

36

-20

-10

0 10 20 30 40 Time from injection (min)

50

35

C

Wildtype UCP1-KO

38

0.5

0.0

39

esis. It seems plausible to assume that under normoxic conditions, the results of Baron et al. (3) would be in agreement with ours. Previous experiments indicated that UCP1 was essential for selective (3-adrenoceptor-induced blood flow in BAT (15). The reasons for this apparent discrepancy are unclear, but the outcome may be related to an inability of such agents to stimulate, e.g., the endothelial cells in the tissue, in contrast to what would be seen when a more physiological stimulation (i.e., with NE) is used. In summary, our data demonstrate that acute adrenergic regulation of iBAT blood flow is independent of the metabolic

-20

**

**


*** ***

40


oxic conditions (100% O2). However, hyperoxia generally causes vasoconstriction of blood vessels, thus decreasing blood flow (2, 18, 26, 32). Moreover, adrenergically induced blood flow in BAT is inversely modulated by the O2 levels in the blood supplying the tissue, with hyperoxic levels decreasing BAT blood flow (10). Thus, in the report by Baron et al (3), wild-type mice could have counteracted the vasoconstrictor effect of hyperoxia by using O2 for thermogenesis. The same would not happen in UCP1-KO mice, where vasoconstriction would remain even following adrenergic stimulation. The resulting differences between genotypes would appear as a dependence of BAT blood flow on UCP1-dependent thermogen-

4

Wildtype UCP1-KO

3

NE

2 1 0

-20


40

Fig. 4. Blood flow and thermogenesis in UCP1-ablated mice acclimated to thermoneutrality. Measurements of whole body oxygen consumption (A), body temperature (B), and quantification of blood flow (C) in wild-type and UCP1-KO mice. NE injection is indicated by the arrows, as described in MATERIALS AND METHODS. Data points represent means # SE of 6 –7 mice. Statistical significance is indicated as follows: *P & 0.05; **P & 0.01; ***P & 0.001. AJP-Endocrinol Metab • doi:10.1152/ajpendo.00494.2014 • www.ajpendo.org


requirements of the tissue, and therefore, increased blood flow in BAT following adrenergic stimulation is not necessarily a reliable indicator of thermogenesis (but of course, there can be no enhanced thermogenesis without increased blood flow). Blood glycemia affects iBAT blood flow but not thermogenesis. In WAT, postprandial elevation of blood flow is considered to be regulated mostly by central sympathetic drive acting on local adrenoceptors, causing the production and release of nitric oxide (NO) from both parenchymal and vascular cells (1, 30). This effect seems to be dependent on the glucose present in meals but not on the lipid content of the meals (8, 30). In accord with this, in BAT blood flow also increases postprandially following a high-carbohydrate meal compared with a high-fat meal (13). Therefore, we investigated whether an abrupt elevation in blood glucose levels (as caused by a glucose tolerance test) would be able to increase BAT perfusion. To this end, we measured iBAT blood flow in anesthetized mice before and after glucose injection. Blood glucose levels were measured simultaneously every 10 min, showing the expected excursion following injection (Fig. 5A). Using HRLDI, the blood flow was monitored and compared with salineinjected control mice (Fig. 5B). A significant increase in iBAT blood flow was observed 5 min after glucose injection, and the blood flow returned to baseline values 15 min after the injection. Interestingly, we did not find any detectable variations in body temperature in response to glucose, which would have been expected in the case of BAT activation (Fig. 5C). To evaluate the correlation between blood glycemia and iBAT blood flow, all correspondent data points were plotted together. However, no correlation was found between circulating glucose levels and iBAT blood flow (Fig. 5D). An explanation for the absence of increased body temperature is that although short-term thermogenesis could have been triggered by glucose injection, it did not last long enough to cause measurable changes in body temperature. Alternatively,


25 20

Glucose

15 10 5 0

C

B

30

-20


39 Tb (°C)

50%

D

40

38

Glucose

37 36 35

100%

0%

40


40

Glucose Saline -20

-10

0 10 20 30 Time from injection (min)

40

4 3

R² = 0.0147

Fig. 5. Effects of systemic glucose levels on iBAT blood flow. A: blood glucose levels after glucose injection. B: measurement of iBAT blood flow following glucose injection compared with saline-injected group. C: body temperature (Tb). D: correlation between iBAT blood flow and systemic glycemia values; no statistically significant correlation was found between the parameters. Data points represent means # SE of 4 –7 animals. In B, significance is indicated by *P & 0.05.

2 1

0 -20

it can be speculated that because of the dually innervated nature of iBAT, signals from the brain could trigger specifically the NPY-containing nerves surrounding the vasculature without affecting thermogenesis in the parenchymal cells (5). Nevertheless, our results confirm that, regarding BAT, blood flow responds to glucose similarly to WAT (30). Although a transient increase in iBAT blood flow can be measured, it does not seem to be accompanied by classical activation of thermogenesis, supporting the interpretation that these phenomena are physiologically dissociated from each other. Cold acclimation increases NE-induced blood flow in inguinal WAT. During the last few years, brown adipocyte-like cells in WAT (brite/beige adipocytes) have been identified (27), suggesting that the beneficial functions of BAT could be further expanded to “browned” WAT. Following cold acclimation or pharmacological stimulation, iWAT has been shown to present brown-like features such as multilocular adipocytes (33) and functional UCP1 (29). Therefore, we examined whether iWAT had the same characteristics of adrenergic and adaptive blood flow regulation as BAT. We measured blood flow in iWAT of mice housed at thermoneutrality or room temperature or acclimated to cold, as described above for the iBAT experiments. Baseline blood flow in the inguinal area was similar in all groups (Fig. 6A). After NE injection, a distinct area over the inguinal ligament (at the proximal end of the more clearly identifiable femoral artery) became visible by HR-LDI, as indicated by the subtracted images (Fig. 6A, right). The response is localized in iWAT, which was verified by postmortem analysis of the same region (Fig. 6B). Image quantification demonstrated an increased vascular response to NE following acclimation to lower environmental temperatures, with the highest changes in blood flow being measured in mice acclimated to cold (Fig. 6C). This provides evidence that blood flow capacity in iWAT is sensitive to temperature acclimation and responds to adrenergic stimulation similarly to iBAT, in accord with the previously described angiogenesis in

*

150%

BAT blood flow (a.u.)

Blood Glucose (mM)

A

E827

0

5

10 15 20 25 Blood glucose (mM)

30


E828


Baseline

Norepinephrine stimulation

Norepinephrine effect

B 10.0

Thermoneutrality

A

7.5 5.0 2.5

Room temperature

0.0 HR-LDI signal (a.u.) Peritoneum Inguinal WAT Femoral artery


Cold -acclimated

C

10 8

NE

6

4 Cold-acclimated Room temperature Thermoneutrality

2 0

-20


40

Fig. 6. Blood flow in inguinal WAT (iWAT) following NE administration. Blood flow variations were measured in iWAT of anesthetized mice acclimated to thermoneutrality, standard room temperature (21°C), or cold. A: HR-LDI images of the inguinal region during baseline and 10 min post-NE injection and )perfusion images. The region of interest delineates iWAT anatomic location. B: postmortem visualization of the imaged region with a focus on iWAT and the femoral artery. C: quantification of the effects of ambient temperature acclimation on adrenergically stimulated blood flow in iWAT. Data points represent means # SE of 5– 6 mice.

iWAT following cold acclimation (34). Because iWAT, particularly at thermoneutrality, has a relatively low metabolic rate (29), adrenergically stimulated blood flow most likely functions to supply the systemic circulation with fatty acids released through lipolysis (28). These results provide additional support that blood flow is not secondary to hypoxia/ thermogenesis, similar to what was demonstrated above for iBAT. Considerations regarding BAT blood flow. There are different ways to conceptualize adrenergic regulation of BAT blood flow while keeping its dissociation from thermogenesis. Under physiological conditions (e.g., cold exposure and increased sympathetic drive), blood flow could be stimulated by NPYcontaining sympathetic nerves, which directly target the vasculature of BAT but not the adipocytes (5). It is also possible that the effects on blood flow are caused by a metabolite released by brown adipocytes during adrenergic stimulation, provided that the metabolite production is not secondary to O2 consumption. It is likely that NO may play a role in linking adrenergic activation and increased blood flow. NO is one of the main molecules involved in smooth muscle relaxation and modulation of tissue blood flow (1, 30). Indeed, inhibition of NO synthesis totally abolishes adrenergically stimulated blood flow (21) in BAT, demonstrating that alternative mediators of vasodilation would not be involved. It is not presently possible

to conclude anything concerning the cellular origin of the NO. NO is produced by brown adipocytes in response to adrenergic stimulation (23). Because the brown adipocyte production of NO is caused by increased intracellular cAMP levels (23), NO release occurs in parallel to UCP1 activation. Thus, vasodilation can be caused by NO with a brown fat cell origin despite the absence of thermogenesis. Alternatively, the NO could be released from the specific NPY-containing sympathetic fibers innervating the vasculature (5) and thus occur totally independent of the adrenergic signals to the brown fat cells. It is clear from our data that a lowered O2 level is not a necessity for increased blood flow, but our results do not exclude the possibility of O2 effects on BAT perfusion rates. However, we demonstrate clearly that enhanced blood flow is not a valid parameter to demonstrate that thermogenesis actually takes place in the tissue. Therefore, future studies need to exercise caution when interpreting blood flow data as a measure of the thermogenic activity of BAT. ACKNOWLEDGMENTS We thank Moor Instruments (UK) for access to the HR-LDI apparatus and Brian Lock for providing excellent technical support during the establishment of the technique. GRANTS This work was funded by grants from the Swedish Research Council [Naturvetenskap och Teknik (Natural Sciences and Technology) and Medicin


REGULATION OF BROWN ADIPOSE TISSUE BLOOD FLOW och Hälsa (Medicine And Health)], the DIABAT consortium within the European Union Seventh Framework Programme, the Wallenberg Foundation’s program on brown adipose tissue, the Novo Nordisk Foundation, and the European Research Council (Young Investigator Grant to K. L. Spalding). C. E. Hagberg was supported by grants from the Wilhelm och Else Stockmanns Stiftelse and the Swedish Society for Medical Research. DISCLOSURES J. Nedergaard is a member of the scientific advisory board and a shareholder in Ember Therapeutics. B. Cannon is a member of the Scientific Advisory Board of Metabolic Solutions Development Company. AUTHOR CONTRIBUTIONS G.A.-V., C.E.H., K.L.S., B.C., and J.N. conception and design of research; G.A.-V. and C.E.H. performed experiments; G.A.-V. and C.E.H. analyzed data; G.A.-V., C.E.H., K.L.S., B.C., and J.N. interpreted results of experiments; G.A.-V. and C.E.H. prepared figures; G.A.-V. and J.N. drafted manuscript; G.A.-V., C.E.H., K.L.S., B.C., and J.N. edited and revised manuscript; G.A.-V., C.E.H., K.L.S., B.C., and J.N. approved final version of manuscript. REFERENCES 1. Ardilouze JL, Sotornik R, Dennis LA, Fielding BA, Frayn KN, Karpe F. Failure to increase postprandial blood flow in subcutaneous adipose tissue is associated with tissue resistance to adrenergic stimulation. Diabetes Metab 38: 27–33, 2012. 2. Ariyaratnam P, Loubani M, Bennett R, Griffin S, Chaudhry MA, Cowen ME, Guvendik L, Cale AR, Morice AH. Hyperoxic vasoconstriction of human pulmonary arteries: a novel insight into acute ventricular septal defects. ISRN Cardiol 2013: 685735, 2013. 3. Baron DM, Clerte M, Brouckaert P, Raher MJ, Flynn AW, Zhang H, Carter EA, Picard MH, Bloch KD, Buys ES, Scherrer-Crosbie M. In vivo noninvasive characterization of brown adipose tissue blood flow by contrast ultrasound in mice. Circ Cardiovasc Imaging 5: 652–659, 2012. 4. Cannon B, Nedergaard J. Brown adipose tissue: function and physiological significance. Physiol Rev 84: 277–359, 2004. 5. Cannon B, Nedergaard J, Lundberg JM, Hökfelt T, Terenius L, Goldstein M. “Neuropeptide tyrosine” (NPY) is co-stored with noradrenaline in vascular but not in parenchymal sympathetic nerves of brown adipose tissue. Exp Cell Res 164: 546 –550, 1986. 6. Clerte M, Baron DM, Brouckaert P, Ernande L, Raher MJ, Flynn AW, Picard MH, Bloch KD, Buys ES, Scherrer-Crosbie M. Brown adipose tissue blood flow and mass in obesity: a contrast ultrasound study in mice. J Am Soc Echocardiogr 26: 1465–1473, 2013. 7. Enerback S, Jacobsson A, Simpson EM, Guerra C, Yamashita H, Harper ME, Kozak LP. Mice lacking mitochondrial uncoupling protein are cold-sensitive but not obese. Nature 387: 90 –94, 1997. 8. Evans K, Clark ML, Frayn KN. Effects of an oral and intravenous fat load on adipose tissue and forearm lipid metabolism. Am J Physiol Endocrinol Metab 276: E241–E248, 1999. 9. Fawcett DW. A comparison of the histological organization and cytochemical reactions of brown and white adipose tissues. J Morphol 90: 363–405, 1952. 10. Foster DO, Depocas F. Evidence against noradrenergic regulation of vasodilation in rat brown adipose tissue. Can J Physiol Pharmacol 58: 1418 –1425, 1980. 11. Foster DO, Depocas F, Frydman ML. Noradrenaline-induced calorigenesis in warm- and cold-acclimated rats: relations between concentration of noradrenaline in arterial plasma, blood flow to differently located masses of brown adipose tissue, and calorigenic response. Can J Physiol Pharmacol 58: 915–924, 1980. 12. Foster DO, Frydman ML. Nonshivering thermogenesis in the rat. II. Measurements of blood flow with microspheres point to brown adipose tissue as the dominant site of the calorigenesis induced by noradrenaline. Can J Physiol Pharmacol 56: 110 –122, 1978. 13. Glick Z, Wickler SJ, Stern JS, Horwitz BA. Blood flow into brown fat of rats is greater after a high carbohydrate than after a high fat test meal. J Nutr 114: 1934 –1939, 1984.

E829

14. Golozoubova V, Cannon B, Nedergaard J. UCP1 is essential for adaptive adrenergic nonshivering thermogenesis. Am J Physiol Endocrinol Metab 291: E350 –E357, 2006. 15. Granneman JG, Burnazi M, Zhu Z, Schwamb LA. White adipose tissue contributes to UCP1-independent thermogenesis. Am J Physiol Endocrinol Metab 285: E1230 –E1236, 2003. 16. Hany TF, Gharehpapagh E, Kamel EM, Buck A, Himms-Hagen J, von Schulthess GK. Brown adipose tissue: a factor to consider in symmetrical tracer uptake in the neck and upper chest region. Eur J Nucl Med Mol Imaging 29: 1393–1398, 2002. 17. Heim T, Hull D. The blood flow and oxygen consumption of brown adipose tissue in the new-born rabbit. J Physiol 186: 42–55, 1966. 18. Iscoe S, Beasley R, Fisher JA. Supplementary oxygen for nonhypoxemic patients: O2 much of a good thing? Crit Care 15: 305, 2011. 19. Kushner RF. Weight loss strategies for treatment of obesity. Prog Cardiovasc Dis 56: 465–472, 2014. 20. Matthias A, Ohlson KB, Fredriksson JM, Jacobsson A, Nedergaard J, Cannon B. Thermogenic responses in brown fat cells are fully UCP1dependent. UCP2 or UCP3 do not substitute for UCP1 in adrenergically or fatty scid-induced thermogenesis. J Biol Chem 275: 25073–25081, 2000. 21. Nagashima T, Ohinata H, Kuroshima A. Involvement of nitric oxide in noradrenaline-induced increase in blood flow through brown adipose tissue. Life Sci 54: 17–25, 1994. 22. Nedergaard J, Bengtsson T, Cannon B. Unexpected evidence for active brown adipose tissue in adult humans. Am J Physiol Endocrinol Metab 293: E444 –E452, 2007. 23. Nisoli E, Tonello C, Briscini L, Carruba MO. Inducible nitric oxide synthase in rat brown adipocytes: implications for blood flow to brown adipose tissue. Endocrinology 138: 676 –682, 1997. 24. Ohlson KB, Lindahl SG, Cannon B, Nedergaard J. Thermogenesis inhibition in brown adipocytes is a specific property of volatile anesthetics. Anesthesiology 98: 437–448, 2003. 25. Orava J, Nuutila P, Noponen T, Parkkola R, Viljanen T, Enerback S, Rissanen A, Pietilainen KH, Virtanen KA. Blunted metabolic responses to cold and insulin stimulation in brown adipose tissue of obese humans. Obesity (Silver Spring) 21: 2279 –2287, 2013. 26. Pasgaard T, Stankevicius E, Jørgensen MM, Ostergaard L, Simonsen U, Frøbert O. Hyperoxia reduces basal release of nitric oxide and contracts porcine coronary arteries. Acta Physiol (Oxf) 191: 285–296, 2007. 27. Petrovic N, Walden TB, Shabalina IG, Timmons JA, Cannon B, Nedergaard J. Chronic peroxisome proliferator-activated receptor gamma (PPARgamma) activation of epididymally derived white adipocyte cultures reveals a population of thermogenically competent, UCP1-containing adipocytes molecularly distinct from classic brown adipocytes. J Biol Chem 285: 7153–7164, 2010. 28. Rochon L, Bukowiecki LJ. Alterations in adipocyte response to lipolytic hormones during cold acclimation. Am J Physiol Cell Physiol 258: C835–C840, 1990. 29. Shabalina IG, Petrovic N, de Jong JM, Kalinovich AV, Cannon B, Nedergaard J. UCP1 in brite/beige adipose tissue mitochondria is functionally thermogenic. Cell Rep 5: 1196 –1203, 2013. 30. Sotornik R, Brassard P, Martin E, Yale P, Carpentier AC, Ardilouze JL. Update on adipose tissue blood flow regulation. Am J Physiol Endocrinol Metab 302: E1157–E1170, 2012. 31. Thurlby PL, Trayhurn P. Regional blood flow in genetically obese (ob/ob) mice. The importance of brown adipose tissue to the reduced energy expenditure on non-shivering thermogenesis. Pflugers Arch 385: 193–201, 1980. 32. Werkmeister RM, Palkovits S, Told R, Groschl M, Leitgeb RA, Garhofer G, Schmetterer L. Response of retinal blood flow to systemic hyperoxia as measured with dual-beam bidirectional Doppler Fourierdomain optical coherence tomography. PLoS One 7: e45876, 2012. 33. Vitali A, Murano I, Zingaretti MC, Frontini A, Ricquier D, Cinti S. The adipose organ of obesity-prone C57BL/6J mice is composed of mixed white and brown adipocytes. J Lipid Res 53: 619 –629, 2012. 34. Xue Y, Petrovic N, Cao R, Larsson O, Lim S, Chen S, Feldmann HM, Liang Z, Zhu Z, Nedergaard J, Cannon B, Cao Y. Hypoxia-independent angiogenesis in adipose tissues during cold acclimation. Cell Metab 9: 99 –109, 2009.
