Mitochondrial Patch Clamp of Beige Adipocytes Reveals ... - Cell Press

0 downloads 0 Views 8MB Size Report
Apr 4, 2017 - Beige adipocytes of abdominal fat are mostly UCP1 negative but possess thermogenic capacity associated with a futile creatine cycle. Bertholet ...
Article

Mitochondrial Patch Clamp of Beige Adipocytes Reveals UCP1-Positive and UCP1-Negative Cells Both Exhibiting Futile Creatine Cycling Graphical Abstract

Authors Ambre M. Bertholet, Lawrence Kazak, Edward T. Chouchani, ..., Shingo Kajimura, Bruce M. Spiegelman, Yuriy Kirichok

Correspondence [email protected]

In Brief Using patch-clamping and bioenergetics analyses of isolated mitochondria, Bertholet et al. investigated the thermogenic mechanisms of beige fat and identified two distinct types of beige adipocytes: UCP1 positive and UCP1 negative. Beige adipocytes of abdominal fat are mostly UCP1 negative but possess thermogenic capacity associated with a futile creatine cycle.

Highlights d

Patch clamp is used to directly measure mitochondrial H+ leak in beige adipocytes

d

Two types of beige adipocytes are identified: UCP1 positive and UCP1 negative

d

Both types of beige adipocytes display a robust futile creatine cycle

d

Abdominal beige fat, while mostly UCP1 negative, has clear thermogenic capacity

Bertholet et al., 2017, Cell Metabolism 25, 811–822 April 4, 2017 ª 2017 Elsevier Inc. http://dx.doi.org/10.1016/j.cmet.2017.03.002

Cell Metabolism

Article Mitochondrial Patch Clamp of Beige Adipocytes Reveals UCP1-Positive and UCP1-Negative Cells Both Exhibiting Futile Creatine Cycling  ska,1 Ishan Paranjpe,1 Ambre M. Bertholet,1 Lawrence Kazak,2,3 Edward T. Chouchani,2,3 Marta G. Bogaczyn Gabrielle L. Wainwright,1 Alexandre Be´tourne´,1 Shingo Kajimura,4,5,6 Bruce M. Spiegelman,2,3 and Yuriy Kirichok1,7,* 1Department

of Physiology, University of California, San Francisco, San Francisco, CA 94143, USA of Cancer Biology, Dana-Farber Cancer Institute, Boston, MA 02115, USA 3Department of Cell Biology, Harvard Medical School, Boston, MA 02115, USA 4Diabetes Center 5Department of Cell and Tissue Biology 6Eli and Edythe Broad Center of Regeneration Medicine and Stem Cell Research University of California, San Francisco, San Francisco, CA 94143, USA 7Lead Contact *Correspondence: [email protected] http://dx.doi.org/10.1016/j.cmet.2017.03.002 2Department

SUMMARY

Cold and other environmental factors induce ‘‘browning’’ of white fat depots—development of beige adipocytes with morphological and functional resemblance to brown fat. Similar to brown fat, beige adipocytes are assumed to express mitochondrial uncoupling protein 1 (UCP1) and are thermogenic due to the UCP1-mediated H+ leak across the inner mitochondrial membrane. However, this assumption has never been tested directly. Herein we patch clamped the inner mitochondrial membrane of beige and brown fat to provide a direct comparison of their thermogenic H+ leak (IH). All inguinal beige adipocytes had robust UCP1-dependent IH comparable to brown fat, but it was about three times less sensitive to purine nucleotide inhibition. Strikingly, only 15% of epididymal beige adipocytes had IH, while in the rest UCP1-dependent IH was undetectable. Despite the absence of UCP1 in the majority of epididymal beige adipocytes, these cells employ prominent creatine cycling as a UCP1-independent thermogenic mechanism.

INTRODUCTION Brown fat thermogenesis has emerged as a promising target to €ck, 2015; combat the metabolic syndrome (Betz and Enerba Cypess et al., 2009, 2015; van Marken Lichtenbelt et al., 2009; Virtanen et al., 2009). Heat production in brown adipocytes depends on the H+ leak across the inner mitochondrial membrane, mediated by uncoupling protein 1 (UCP1). White fat also contains diffuse islands of UCP1-expressing adipocytes, and the number of these cells is significantly increased by cold adaptation or injection of b3-adrenergic agonists (Cousin et al., 1992; Granneman et al., 2005; Guerra et al., 1998; Himms-Hagen

et al., 2000; Kozak et al., 2010; Loncar et al., 1988; Wu et al., 2012, 2013). The overall morphology of these UCP1-expressing cells is comparable to brown adipocytes: they have numerous mitochondria and are multilocular. This is in contrast to the low mitochondrial abundance and unilocularity of white adipocytes. Due to the morphological similarity to brown fat, these UCP1-expressing cells were termed ‘‘beige’’ or ‘‘brite’’ adipocytes (Walde´n et al., 2012; Wu et al., 2012). Despite their morphological and functional similarities, beige and brown adipocytes have clearly distinct developmental origins (Long et al., 2014; Rosen and Spiegelman, 2014; Seale et al., 2008). Interestingly, brown adipose tissue of adult humans contains a significant population of cells with a beige adipose molecular signature, and these cells could prevent obesity and type II diabetes (Jespersen et al., 2013; Lee et al., 2014; Lidell et al., 2013; Sharp et al., 2012; Shinoda et al., 2015; Wu et al., 2012). There is a general agreement that all beige adipocytes are capable of UCP1-dependent thermogenesis, and UCP1 is universally used as a marker of beige adipocytes. However, the evidence for UCP1-dependent thermogenesis in beige adipocytes came primarily from experiments using subcutaneous (inguinal) adipose tissue (Shabalina et al., 2013). In contrast, only a small fraction of beige adipocytes of abdominal white fat depots, such as epididymal and retroperitoneal, may express UCP1 (Himms-Hagen et al., 2000; Loncar et al., 1988), and this phenomenon has never been adequately addressed. Moreover, beige adipocytes seem to use alternative UCP1-independent thermogenic mechanisms, as demonstrated in UCP1/ mice (Granneman et al., 2003; Ukropec et al., 2006). Creatine-dependent ADP/ATP substrate cycling has recently been illustrated to serve as a thermogenic pathway in inguinal beige adipocytes that co-exists with classic UCP1-dependent thermogenesis (Kazak et al., 2015). Thus, the molecular mechanisms of thermogenesis in beige fat require further exploration. UCP1-dependent H+ leak current (IH) across the inner mitochondrial membrane (IMM) short-circuits the mitochondrial voltage (DJ) and allows brown fat mitochondria to generate heat at the expense of ATP production. Until recently, there were no methods for direct IH measurement, which severely Cell Metabolism 25, 811–822, April 4, 2017 ª 2017 Elsevier Inc. 811

limited our understanding of the molecular mechanisms of mitochondrial thermogenesis. Recently, we directly measured and characterized UCP1-dependent IH in the interscapular brown fat using the patch-clamp technique (Fedorenko et al., 2012). Classic studies from the 1970s demonstrated that UCP1-dependent thermogenesis is activated by long-chain fatty acids (FAs) and that UCP1 activity is equivalent to transport of H+ across the IMM (Nicholls and Locke, 1984). However, the mechanism of FA-dependent UCP1 operation had remained elusive (Cannon and Nedergaard, 2004). By directly recording UCP1 currents across the whole IMM of brown fat, we found that FA anions (FA) activate UCP1 by serving as its transport substrates, and UCP1 operates as an unusual FA/H+ symporter (Fedorenko et al., 2012). However, in contrast to short-chain FA, physiologically relevant long-chain FA cannot dissociate from UCP1 due to strong hydrophobic interactions established by their carbon tails and thus serve as continuously attached substrates. Therefore, in the presence of long-chain FAs, UCP1 effectively operates as an FA-activated H+ carrier (Bertholet and Kirichok, 2017). We also demonstrated that in addition to their transport function, long-chain FAs competitively release tonic inhibition of UCP1 by cytosolic purine nucleotides. Finally, although classic studies postulated that UCP1 is activated by long-chain FAs released by hydrolysis of cytoplasmic lipid droplets (triggered by b3-adrenergic receptor activation), we discovered that UCP1 can also be controlled by FAs generated within the IMM via phospholipid hydrolysis by a non-canonical Ca2+-independent phospholipase A2 (PLA2) (Fedorenko et al., 2012). Here we use mitochondrial patch clamping to compare IH properties in interscapular brown fat and beige fat of two distinct white fat depots: inguinal and epididymal. We demonstrate that b3-adrenergic receptor stimulation (with the selective agonist CL316.243, abbreviated ‘‘CL’’ here) leads to development of beige adipocytes in both inguinal and epididymal fat, as indicated by fragmentation of cytosolic lipid droplets, robust mitochondrial biogenesis, and upregulation of the thermogenic gene program. However, while all new CL-induced inguinal beige mitochondria exhibit UCP1-dependent IH similar to that of brown fat, the large majority of newly formed epididymal beige mitochondria (85%) have no UCP1-dependent IH. Thus, in contrast to the current belief that all beige adipocytes are capable of UCP1-dependent thermogenesis, our data clearly demonstrate the existence of UCP1-positive and UCP1-negative beige fat cells. UCP1-positive beige adipocytes, based on the properties of their IH, appear to have similar potential for UCP1-dependent thermogenesis as brown adipocytes, except that their IH is less sensitive to inhibition by purine nucleotides. Although UCP1-negative beige adipocyte mitochondria are devoid of UCP1-dependent thermogenesis, they run robust creatine-dependent substrate cycling (Kazak et al., 2015). This may be the principal mechanism of thermogenesis employed by UCP1-negative beige fat. RESULTS ‘‘Browning’’ of Inguinal and Epididymal Fat after Chronic b3-Adrenergic Receptor Stimulation To induce beige adipocytes in white fat depots, we injected mice intraperitoneally with a selective b3-adrenergic receptor agonist 812 Cell Metabolism 25, 811–822, April 4, 2017

CL for 10 days. This procedure has previously been shown to induce brown-like adipocytes in various white fat depots (Granneman et al., 2003, 2005; Himms-Hagen et al., 2000). Histological analysis showed strong remodeling in both inguinal and epididymal depots (Figure 1A). This white fat transformation is described as ‘‘browning,’’ and its most obvious morphological characteristic is an increase in number and reduction in size of cytoplasmic lipid droplets. Indeed, white adipocytes of vehicletreated mice were primarily unilocular (Figure 1A, left panels), while CL treatment resulted in accumulation of multilocular adipocytes (Figure 1A, right panels). Although the browning of epididymal fat was not as strong as that of inguinal fat, it was clearly distinguishable from vehicle-treated fat with the appearance of lipid droplet multilocularity (Figure 1A). Another key parameter used to describe the browning of white fat is mitochondrial biogenesis. Robust mitochondrial biogenesis was detected by the increased abundance of the transcriptional coactivator PGC1a (a key regulator of mitochondrial biogenesis); the mitochondrial respiratory chain protein COX IV; and a component of the mitochondrial protein import machinery, TOM20, in both inguinal and epididymal depots after CL injection (Figures 1B–1D). Interestingly, while UCP1 protein expression was induced in the inguinal fat upon CL injection, it was undetectable in epididymal fat (Figures 1B–1D). The interscapular brown fat did not show any significant change in the expression level of these proteins (Figure S1A). Interestingly, UCP1 appeared non-essential for the process of browning because robust mitochondrial biogenesis, as judged by PGC1a, COX IV, and TOM20 expression, was observed in both epididymal and inguinal depots of UCP1/ mice (Figures S1C and S1D). Similar to wild-type (WT) mice, CL injection did not induce mitochondrial biogenesis in interscapular brown fat of UCP1/ mice (Figure S1B). Finally, we assessed induction of the thermogenic program in both inguinal and epididymal fat by examining relative mRNA levels for Cidea, Dio2, Prdm16, Ucp1, and Pgc1a. CL injection highly increased the expression levels of Cidea, Dio2, and Ucp1 in both tissues (Figure S2). Interestingly, although in epididymal fat UCP1 protein was undetectable with western blot in both control and CL-injected mice (Figures 1B and 1D), CL injection caused a dramatic increase in Ucp1 mRNA (Figure S2). Thus, thermogenic gene profiling further confirmed the induction of browning in both epididymal and inguinal fat. Direct Measurement of Mitochondrial H+ Leak in Beige Adipocytes: UCP1-Positive and UCP1-Negative Cells We next applied the patch-clamp technique to the vesicles of the whole IMM (mitoplasts) (Fedorenko et al., 2012; Fieni et al., 2012) isolated from inguinal and epididymal beige fat mitochondria to directly measure H+ currents carried by UCP1. Our standard mitochondrial isolation protocol produced no mitochondria from inguinal and epididymal fat of vehicle-treated mice (due to the very low amount of mitochondria in white fat). However, we isolated significant amounts of mitochondria from both tissues of CL-injected animals, as was expected based on the mitochondrial biogenesis (Figures 1B–1D). We next recorded the mitochondrial H+ leak (IH) across the whole IMM in mitoplasts isolated from inguinal and epididymal fat of CL-injected mice and compared it to IH of other tissues, such as brown fat, heart, and

VEHICLE

A

Figure 1. Browning of Inguinal and Epididymal Fat Depots upon Chronic b3-Adrenergic Receptor Stimulation

CL316.243

INGUINAL

100 μM

(A) Representative inguinal and epididymal fat sections of mice treated with vehicle or CL316.243 compound were stained with hematoxylin and eosin; scale bar, 100 mm. A magnified view is shown on the right corner of each picture; scale bar, 300 mm. (B) Representative immunoblots showing the effect of b3-adrenergic receptor stimulation on the protein levels of Na+/K+ ATPase, PGC1a, UCP1, and COXIV in inguinal fat (left) and epididymal fat (right). See also Figures S1 and S2. (C and D) PGC1a, TOM20, COX IV, and UCP1 protein levels in inguinal (C) and epididymal (D) fat, based on the data presented in (B). Data shown as mean ± SEM; n = 5–9. See also Figures S1 and S2.

EPIDIDYMAL

300 μM

B

Epididymal Fat

Inguinal Fat Vehicle CL316.243

Vehicle

113

Na+/K+ ATPase

95

PGC1α

35

UCP1

CL316.243

Ratio TOM 20 / Na-K ATPase , (a.u.)

Ratio COX IV / TOM 20, (a.u.)

UNDETECTABLE

Ratio UCP1 / TOM 20, (a.u.)

UNDETECTABLE

Ratio TOM 20 / Na-K ATPase , (a.u.)

Ratio COX IV / TOM 20, (a.u.)

UNDETECTABLE

Ratio UCP1 / TOM 20, (a.u.)

UNDETECTABLE

UNDETECTABLE

Ratio PGC1α / Na-K ATPase , (a.u.)

Ratio PGC1α / Na-K ATPase , (a.u.)

estingly, there were two clearly distinct populations of epididymal beige fat mitoplasts with no ‘‘intermediate’’ cases: (1) COX IV 15 having no UCP1 current and (2) having 17 TOM20 UCP1-dependent IH with an amplitude comparable to that found in brown and PGC1α TOM20 COX IV UCP1 inguinal fat mitoplasts (Figures 2C and C **** 2E). As expected (given the fact that 1.5 1.5 1.5 1.5 ** UCP1 expression is limited to fat tissues), 1.0 1.0 1.0 1.0 under the same conditions IH was undeInguinal Fat 0.5 0.5 0.5 0.5 tectable in mitoplasts isolated from heart and skeletal muscle (Figure 2D). Although 0 0 0 0 Veh CL Veh CL Veh CL Veh CL IH was missing in the majority of mitoplasts from epididymal beige fat as well as in all TOM20 UCP1 PGC1α COX IV D mitoplasts from heart and skeletal muscle, ** 1.5 1.5 ** 1.5 1.5 we recorded Cl currents (likely carried by 1.0 1.0 1.0 1.0 the inner membrane anion channel, IMAC) Epididymal Fat (Borecky´ et al., 1997) in all these mitoplasts 0.5 0.5 0.5 0.5 as a positive control (Figure S3). 0 0 0 0 Veh CL Veh CL These experiments demonstrate that Veh CL Veh CL all inguinal beige adipocytes have functional UCP1 and thus can be classified skeletal muscle (Figure 2). IH was induced by application of a as UCP1 positive. In contrast, epididymal beige adipocytes voltage ramp from 160 mV (on the matrix side of the IMM as are clearly divided into two distinct subpopulations, UCP1 compared to the cytosolic side) to +100 mV to cover all possible positive and UCP1 negative. To gain insight into spatial distriphysiological voltages across the IMM. The bath and pipette so- bution of UCP1-positive and UCP1-negative beige adipocytes lutions were formulated to record H+ currents and contained only in inguinal and epididymal fat, we performed immunohistosalts that dissociate into large anions and cations that are nor- chemical analyses of both tissues. Mitochondrial immunostaining in slices of inguinal and epididymal fat revealed an mally impermeant through ion channels or transporters. All inguinal beige fat mitoplasts (n = 35) used in electrophys- increase of mitochondrial biomass after CL treatment in both iological experiments had a large IH (Figures 2A and 2E), tissues (Figures 3 and S4), which was in agreement with similar to that observed in brown fat mitoplasts (Figures 2B western blot analysis for markers of mitochondrial biogenesis and 2E). IH in both inguinal beige fat and brown fat was in- (Figures 1B–1D). However, UCP1 protein expression differed hibited by 1 mM of a classic UCP1 inhibitor guanosine diphos- dramatically between the two tissues. In control mice, inguinal phate (GDP) and disappeared in UCP1/ mice (Figures 2A depot UCP1 was largely undetectable, while CL injection and 2B), clearly demonstrating that in both tissues IH was made virtually all cells in this depot UCP1 positive (Figures 3A, 3C, and S4A). In contrast, in the epididymal fat depot CL mediated by UCP1. In contrast, only about 15% (4 out of 27) of mitoplasts of treatment only produced isolated islands of UCP1-positive adepididymal beige fat had large GDP-sensitive IH, while the major- ipocytes, while the majority of mitochondria-rich beige adipoity of mitoplasts had no measurable H+ current (Figure 2C). Inter- cytes were UCP1 negative (Figures 3B, 3D, and S4B). These Cell Metabolism 25, 811–822, April 4, 2017 813

Inguinal fat

A 0 mV

850 ms

+100 mV

-160 160 mV

Brown fat

B 0 mV

C

+100 mV

850 ms

Epididymal fat

0 mV

850 ms

(A) Representative IH recorded from inguinal beige fat mitoplasts isolated from WT (upper panel) and UCP1/ mice (lower panel). IH traces are shown before (control, black) and after (red) the addition of 1 mM GDP to the bath solution. The voltage protocol is indicated at the top. (B) Representative IH recorded from brown fat mitoplasts isolated from WT (upper panel) and UCP1/ mice (lower panel). IH traces are shown before (control, black) and after (red) the addition of 1 mM GDP to the bath solution. (C) Representative IH recorded from two distinct types of epididymal beige fat mitoplasts isolated from WT mice: UCP1 positive (upper panel) and UCP1 negative (lower panel). IH traces are shown before (control, red) and after (black) the addition of 1 mM GDP to the bath solution. (D) Representative IH recorded from mitoplasts isolated from skeletal muscle (upper panel) and heart (lower panel). IH traces are shown before (control, black) and after (red) the addition of 1 mM GDP to the bath solution. (E) Bar graph showing average IH current density in inguinal beige fat (Ing), brown fat (BAT), heart, skeletal muscle (SM), and two distinct types of beige adipocyte from epididymal beige fat (UCP1 negative [Epi1] and positive [Epi2]). Data shown as mean ± SEM. See also Figure S3.

-160 mV

-160 160 mV

WT

WT 1 mM GDP

+100 mV

WT 1 mM GDP

1 mM GDP

control

control control

UCP1 -/-

UCP1 -/-

WT

1 mM GDP

1 mM GDP

1 mM GDP

control

E

Skeletal muscle 0 mV

-160 160 mV V

850 ms

+100 mV

1 mM GDP

25 pA/pF 200 ms

control pHi 7.0

Heart 0 mV -160 160 mV V

850 ms

+100 mV pHo 7.0

Proton current, pA/pF

D

control

control

Figure 2. Mitochondrial IH in Beige Adipocytes of Inguinal and Epididymal Fat as Compared to Other Tissues

125

100 0 75

50 0 25

1 mM GDP 0 control

Ing BAT Epi 1 Epi 2 Heart SM

n=13 n=10 n=23

results are in agreement with electrophysiological data (Figure 2) and confirm the existence of two populations of beige adipocytes in epididymal fat, UCP1 positive and UCP1 negative. Despite the lack of UCP1 expression, the majority of UCP1-negative adipocytes had robust OXPHOS immunostaining and multilocularity, the two key parameters used to describe beige adipocytes (Figures 3B, 3D, and S4B). Interestingly, UCP1-positive epididymal beige adipocytes appear to be organized in clusters rather than being randomly spread across the tissue. This suggests that UCP1-positive cells are likely to originate from a few precursor cells and are not the result of white adipocyte interconversion. Mitochondrial size can be indicative of the thermogenic capacity of a fat cell, with brown adipocyte mitochondria being fragmented and round (especially upon norepinephrine stimulation), while white fat mitochondria appear larger and filamentous (Cinti, 2009; Wikstrom et al., 2014). Interestingly, the average size of mitochondria in UCP1-negative and UCP1-positve epididymal adipocytes was similar (Figure S5), which further argues for the morphological and functional similarity of these cells. There are two outstanding questions regarding the molecular mechanism of mitochondrial thermogenesis in UCP1-positive and UCP1-negative beige adipocytes: (1) Although UCP1-positive adipocytes have UCP1-dependent H+ leak, are the proper814 Cell Metabolism 25, 811–822, April 4, 2017

ties of this leak identical to those of classic brown fat? (2) Do UCP1-negative beige adipocytes possess an alternative mechanism for mitochondrial thermogenesis? We addressed these two questions with the experiments presented below. n=4

n=6

n=8

The Mechanism by Which FAs Control H+ Current through UCP1 Is Identical in Beige and Brown Fat UCP1 is a brown fat-specific transport protein of the IMM that belongs to the superfamily of mitochondrial solute carriers SLC25 (Nicholls and Locke, 1984; Palmieri, 2004). UCP1 increases H+ conductance of the IMM to short-circuit mitochondrial voltage and make brown fat mitochondria produce heat rather than ATP. UCP1 has two principal physiological regulators: free FAs that activate H+ current via UCP1 and purine nucleotides that inhibit it. There has been a long-standing debate regarding the mechanism by which FAs and purine nucleotides exert their control over UCP1 (Cannon and Nedergaard, 2004). Regarding FAs, opinions put forward by different research groups range widely from the ideas that FAs are absolutely required for the H+ transport ability of UCP1 (Garlid et al., 2000; Klingenberg and Huang, 1999) to the notion that UCP1 is a constitutively active H+ carrier and FAs are only needed to competitively remove purine nucleotide inhibition (Shabalina et al., 2004). The primary reason why the mechanism of FA-dependent UCP1 operation remained elusive was the lack of direct methods to study UCP1. By directly measuring UCP1-dependent H+ leak in brown fat using the mitochondrial patch-clamp

INGUINAL

A

VEHICLE

OXPHOS

UCP1

50 μM

CL316.243

B

EPIDIDYMAL VEHICLE

CL316.24

Figure 3. Mitochondrial Biogenesis and UCP1 Expression in Inguinal and Epididymal Depots after Chronic b3-Adrenergic Stimulation (A and B) Confocal micrographs of inguinal fat (A) and epididymal fat (B) of mice treated with vehicle or CL and immunolabeled with UCP1 (upper panels) and with respiratory chain complex antibodies (MitoProfile Total OXPHOS, lower panels); scale bar, 50 mm. See also Figures S4 and S5. (C and D) Magnified views of the areas within the white rectangles from (A) and (B). Inguinal fat (C), epididymal fat (D). Representative UCP1-positive (right) and UCP1-negative (left) beige adipocytes of epididymal fat are shown by white circles.

OXPHOS

UCP1

from the recording solution (Fedorenko et al., 2012). This FA production was likely due to the activity of phospholipase A2 INGUINAL EPIDIDYMAL C D (PLA2) associated with the IMM (Fedor150 μM enko et al., 2012). Interestingly, similar robust PLA2-like activity was present within the IMM of inguinal beige fat. Indeed, after removal of endogenous membrane FAs and complete deactivation of UCP1 with MbCD, the subsequent removal of MbCD led to reactivation of the UCP1 current, indicating FA regeneration within the IMM (Figure 4B). This suggests that similar to technique, we have recently gained insight into the FA-depen- brown fat, UCP1 activity in beige fat may be affected not only dent mechanism of UCP1 operation (Bertholet and Kirichok, by FAs originating from cytoplasmic lipid droplets (Cannon and 2017; Fedorenko et al., 2012). Here we use the same approach Nedergaard, 2004) but also by FAs produced within the IMM. In the experiments shown in Figure 2, we compared therto determine FA dependence of beige fat UCP1. Since brown and beige adipocytes have different developmental origins (Wu mogenic capacities of beige and brown fat mitochondria by et al., 2012) and represent two distinct cell types, beige fat mito- measuring the amplitude of UCP1-dependent IH in both tissues chondria may present a different environment for UCP1, which (Figures 2A, 2B, and 2E). However, the IH amplitude in this comparison could be affected by different amounts of endogenous can modify its functional properties. To compare the biophysical properties of IH between UCP1- FAs present within the IMM of beige and brown fat. Therefore, positive beige and interscapular brown fat, we used mitoplasts to eliminate the effect of endogenous FAs (which can be different isolated from inguinal beige adipocytes, as they are much in beige and brown fat), we applied exogenous FAs on the backmore abundant than UCP1-positive beige adipocytes of epidid- ground of MbCD (10 mM MbCD mixed with 0.5 mM oleic acid, ymal fat. OA) to directly compare IH of beige and brown fat mitoplasts in Similar to brown fat UCP1 (Fedorenko et al., 2012), UCP1- the presence of the same amount of activating exogenous dependent beige fat IH was observed immediately after breaking FAs. Under these conditions, the difference in the IH amplitude in to the mitoplast (Figure 4A control trace, as well as control between brown and beige fat was even smaller (Figures 4C traces in Figures 2A and 2B, upper panels) and appeared to be and 4D) than observed under conditions when UCP1 was acticonstitutively active, similar to what was suggested previously vated by endogenous FAs (Figures 2A, 2B, and 2E). These (Shabalina et al., 2004). However, UCP1 was not constitutively data demonstrate that the thermogenic potential of UCP1-posiactive because addition of methyl-b-cyclodextrin (MbCD), which tive mitochondria from beige and brown fat is similar. We have previously shown that brown fat UCP1 operates as binds FAs and extracts them from the membrane, led to complete deactivation of IH (Figure 4A). After extraction of endoge- an FA/H+ cotransporter (Fedorenko et al., 2012). Interestingly, nous FAs with MbCD and complete IH deactivation, the current in contrast to short-chain FA that are transported by UCP1 could be reactivated by addition of 1.5 mM arachidonic acid. across the membrane, long-chain FA cannot dissociate from These results exactly replicate our previous data using brown UCP1 due to strong hydrophobic interactions but move within fat mitoplasts (Fedorenko et al., 2012), and demonstrate that the UCP1 translocation pathway, triggering H+ transport (BerFAs are required for UCP1-dependent IH and that the IMM con- tholet and Kirichok, 2017; Fedorenko et al., 2012). Therefore, in the presence of long-chain FAs (the only physiologically relevant tains endogenous FAs that activate UCP1. In the case of brown fat, we demonstrated that after depletion FA activators), FA anion current is not observed and UCP1 effecof endogenous FAs from the IMM using cyclodextrins, mem- tively operates as an FA-activated H+ carrier (Fedorenko et al., brane FAs are rapidly regenerated upon cyclodextrin removal 2012) (Figure S7). Cell Metabolism 25, 811–822, April 4, 2017 815

A

Figure 4. FA Dependence of Beige Fat UCP1

B

C

E

FA currents through UCP1 can be observed in isolation from H currents by using low-pKa FA analogs that cannot be protonated and are always negatively charged in a physiological pH range (Fedorenko et al., 2012). Long-chain low-pKa FA induce transient UCP1 currents due to their inability to dissociate from UCP1 because of hydrophobic interactions (Figure S7B), while short-chain low-pKa FAs that establish much weaker hydrophobic interactions are carried by UCP1 across the membrane (Figure S7A). We found that beige fat UCP1 operates in the same manner. Arachidonyl sulfonate, a long-chain low-pKa FA analog, induced GDP-sensitive transient currents in beige fat mitoplasts (Figure 4E, left panel). In contrast, C6-sulfonate, a low-pKa short chain FA analog, is simply transported by UCP1 across the membrane (Figure 4E, right panel). Thus, beige fat UCP1 transports FA in the same way as brown fat UCP1. +

Beige Fat UCP1 Is Less Sensitive to Purine Nucleotide Inhibition than Brown Fat UCP1 Two principal physiological regulators of UCP1 act in opposition: FAs and Mg2+-free purine nucleotides. In vivo, UCP1 is tonically inhibited by cytosolic purine nucleotides (mainly ATP4), and FAs have to overcome this purine nucleotide inhibition to activate H+ leak via UCP1 (Nicholls, 2006; Nicholls and Lindberg, 1973). However, whether FAs remove purine nucleo816 Cell Metabolism 25, 811–822, April 4, 2017

D

(A) IH (control, red) is deactivated by 10 mM MbCD and reactivated by 1.5 mM arachidonic acid (AA, blue) at pH 7. The voltage protocol is indicated at the top. (B) Left panel: representative time course of IH amplitude in control (1), and upon the application (2) and subsequent washout (3) of 10 mM MbCD at pH 8.0. Right panel: IH traces recorded at times 1, 2, and 3 as indicated in the left panel. (C) Representative IH upon extraction of endogenous long-chain FAs with 10 mM MbCD (black) and after application of 0.5 mM OA mixed with 10 mM MbCD (red). The experiment was performed in brown fat (left panel) and inguinal beige fat (right panel). (D) Bar graph showing average IH current density in inguinal beige fat (ING) and brown fat (BAT) based on the experiments shown in (C). Data shown as mean ± SEM. (E) Representative IH recorded after the extraction of endogenous membrane FAs with 10 mM MbCD (control, black), after subsequent application of the indicated concentration of Cn-sulfonate (red), and upon adding 1 mM GDP (blue), at symmetrical pH 6.0. The structure of the activating Cn-sulfonate is shown near the currents induced: arachidonyl sulfonate (C20, left panel) and hexanesulfonate (C6, right panel). See also Figure S7.

tide inhibition by direct competition has remained controversial (Huang, 2003; Klingenberg, 2010; Nicholls, 2006; Rial et al., 1983; Shabalina et al., 2004; Winkler and Klingenberg, 1994). We have demonstrated that similar to purine nucleotides, long-chain FAs can only bind to UCP1 on the cytosolic face of the IMM and there is apparent competition between long-chain FAs and purine nucleotides for binding to UCP1 (Fedorenko et al., 2012). This suggests that FAs competitively remove purine nucleotide inhibition, and the transport site where FAs bind may partially overlap with the inhibitory purine nucleotide binding site. Thus, the role of purine nucleotides is to increase the threshold for FA binding to its transport site to insure that UCP1 H+ leak is not activated by low background concentrations of FAs. To compare affinities of purine nucleotide binding to UCP1 of beige and brown fat, we activated UCP1-dependent IH in beige and brown fat with the same FA concentration and determined affinities with which cytosolic ATP4 inhibited brown and beige fat IH (Figure 5). To eliminate the effect of endogenous membrane FAs and achieve precise control over FA concentration, IH was activated with exogenous OA applied on a background of 10 mM MbCD (see above). Interestingly, under these conditions beige fat UCP1 appeared to have about three times lower affinity for ATP than brown fat UCP1. In beige fat, IH activated by 0.5 mM OA/10 mM MbCD was inhibited by ATP with an IC50 of 20.5 ± 1.4 mM, whereas the IC50 in brown fat was 7.5 ± 1.4 mM (Figure 5). Thus, we conclude that brown and beige fat UCP1 have different affinities for purine nucleotide binding.

A

B

Figure 5. Inhibition of UCP1 by Purine Nucleotides in Brown and Beige Fat (A) The dose dependence of IH inhibition by ATP4 in brown (red) and beige fat (blue). Current amplitudes were measured upon stepping from 0 to 160 mV; see (B). (B) Representative IH traces in various concentrations of ATP4 on the cytosolic face of the IMM. Brown (left panel) and beige (right panel) are shown. IH was activated with 0.5 mM oleic acid (OA) mixed with 10 mM MbCD. ± SEM, n = 5.

Beige and Brown Fat Mitochondria Have Distinct OXPHOS Profiles Based on electrophysiological analysis of UCP1 currents, mitochondria of inguinal beige fat appear to be similar to mitochondria of classic brown fat (except for lower sensitivity of beige fat UCP1 to cytosolic ATP4). To further address whether inguinal beige fat and brown fat mitochondria are similar in terms of their bioenergetics, we compared expression profiles of the respiratory chain complexes using western blot. This analysis was performed specifically in mitochondria (not whole tissue) isolated from beige and brown fat tissues of CL-injected mice (Figure 6). The OXPHOS antibody cocktail allowed us to quantify the expression level of the four complexes I to IV of the respiratory chain as well as the ATP synthase, and then to establish an ‘‘OXPHOS profile’’ of brown and beige fat mitochondria. The OXPHOS profile of interscapular brown fat mitochondria had three prominent bands: ATP synthase, complex II, and complex IV (Figure 6B). In contrast, the most prominent band detected in beige fat mitochondria (regardless of whether they came from the inguinal or epididymal depot) was ATP synthase (Figure 6A). Thus, the OXPHOS signatures of brown and beige fat mitochondria are distinct. It is well established that brown fat mitochondria have a low amount of the ATP synthase, arguably to prioritize UCP1-dependent thermogenesis over ATP synthesis (Cannon and Nedergaard, 2004). However, in contrast to brown fat, ATP synthase was the dominant band in the OXPHOS profile of beige fat. Indeed, the ratio of ATP synthase to complex II (Figure 6C, left panel) or complex IV (Figure 6C, right panel) appeared to be significantly higher in beige fat as compared to brown fat. Interestingly, although the overall OXPHOS profile of beige fat mitochondria appeared similar regardless of the origin (inguinal or epididymal), epididymal beige fat had a greater abundance of ATP synthase as compared to inguinal fat.

Together, these data indicate that beige fat mitochondria, regardless of whether they are UCP1 positive (inguinal) or UCP1 negative (85% of epididymal beige fat), have characteristic OXPHOS profiles that differ significantly from the OXPHOS profile of classic brown fat. Creatine-Dependent Respiration in Epididymal Beige Fat Recently, a UCP1-independent thermogenic pathway that relies on creatine-dependent substrate cycling has been demonstrated in inguinal beige mitochondria (Kazak et al., 2015). This alternative thermogenic mechanism appears to utilize creatine-dependent futile cycling of ATP synthesis and hydrolysis and thus depends on ATP synthase activity. This cycle drives hydrolysis of a molar excess of ATP with respect to creatine, resulting in a surplus of oxygen consumption under ADP-limiting conditions. Creatine-dependent increase in ADP-limited respiration was observed in inguinal beige fat mitochondria following 1 week of cold exposure (Kazak et al., 2015). Considering that epididymal beige mitochondria have a similar abundance of ATP synthase as inguinal beige mitochondria, they possibly could exhibit this activity following chronic CL administration. We therefore examined the relationship between creatine and ADP-limiting respiration in beige fat mitochondria isolated from the epididymal and inguinal adipose depots. CKMT2 is the principal mitochondrial creatine kinase (CK) isoform in thermogenic fat (Kazak et al., 2015), and is therefore thought to be involved in creatine-dependent substrate cycling. The level of Ckmt2 mRNA was also significantly upregulated in epididymal, but not inguinal, fat (Figures S6A and S6B). However, CL injection caused a significant increase in CKMT2 protein in both epididymal and inguinal fat (Figures 7A and 7B). These data indicate that CKMT2 protein Cell Metabolism 25, 811–822, April 4, 2017 817

A

Epididymal fat

Inguinal fat

B

52 45 37

CV-ATP5A CIII-UQCRC2 CIV-MTCO1

25 15

CII-SDHB CI-NDUFB8

17

TOM20

*

8.0

6.0 4.0

2.0

0

n=5

n=5

n=5

BAT

EPI

ING

Ratio ATP synt / Complex IV, (a.u.)

Ratio ATP synt / Complex II, (a.u.)

C

Brown fat

4.0

(A and B) Representative immunoblots showing protein levels of five mitochondrial respiratory protein complexes: NDUFB8 (complex I, CI), SDHA (complex II, CII), core 2 subunit (complex III, CIII), CIV-I subunit (complex IV, CIV), and ATP5 subunit alpha (complex V, CV) along with the loading control (TOM20) in mitochondria of beige (A) and brown (B) fat. (C) Histograms representing the protein levels of ATP synthase relative to the levels of complex II (left panel) and complex IV (right panel), based on the data presented in (A) and (B). Data shown as mean ± SEM; n = 5.

3.0 2.0

DISCUSSION

1.0 0

n=5 BAT

abundance is primarily regulated post-transcriptionally. Upon CL injection, mitochondrial CK activity was similar in epididymal and inguinal fat (Figure 7C). Phospho1, a phosphatase that may potentiate the creatine-driven substrate cycle through the regulation of downstream high-energy phosphate metabolism (Kazak et al., 2015), was also dramatically increased at the mRNA level in both epididymal and inguinal fat (Figures S6A and S6B). The mRNA levels of other proteins involved in creatine metabolism, Ckmt1, Slc6a8, Gamt, and Gatm, were also affected by CL injection, but the changes were not consistent between epididymal and inguinal fat (Figures S6A and S6B). To assess the activity of the creatine-driven substrate cycle, following CL treatment, beige fat mitochondria were supplemented with pyruvate and malate in the presence of sub-saturating levels of ADP (0.1 mM). Similar to what was previously observed in inguinal beige mitochondria following cold exposure (Kazak et al., 2015), addition of a sub-stoichiometric concentration of creatine to these organelles resulted in activation of creatine-dependent respiration (Figures 7D and 7E, left panels). We next examined the effect of creatine on oxygen consumption in epididymal beige fat mitochondria following CL treatment. Similar to inguinal beige mitochondria, creatine also strongly stimulated respiration of epididymal beige mitochondria under ADP-limiting conditions (Figures 7D and 7E, right panels). On the basis of the mitochondrial phosphate/oxygen (P/O) ratio (Watt et al., 2010), the molar excess of ADP phosphorylation can be estimated from the oxygen consumption rate stimulated by creatine addition. This calculation allows for stoichiometry between excess ADP production and added creatine to be inferred. Employing this analysis indicated that creatine drove a molar excess liberation of ADP to a similar extent for inguinal (18.45 ± 1.237) and epididymal (16.08 ± 1.24) beige mitochondria. Thus, beige fat mitochondria from subcutaneous and visceral depots have a similarly high capacity for creatinedependent substrate cycling. 818 Cell Metabolism 25, 811–822, April 4, 2017

Figure 6. Distinct OXPHOS Profiles of Brown and Beige Fat Mitochondria

UCP1-Positive and UCP1-Negative Beige Adipocytes n=5 n=5 Here we provided the first patch-clamp charEPI ING acterization of the mitochondrial thermogenic H+ leak (IH) in beige adipocytes from two distinct white fat depots: inguinal and epididymal. In contrast to respirometry studies of the H+ leak, this method (1) measures IH directly as H+ current across the IMM, with high amplitude and time resolution; (2) fully controls transmembrane voltage and solution composition on both sides of the IMM, thus allowing for IH measurement under controlled conditions; and (3) measures IH in a single mitochondrion (rather than averages over all mitochondria present in the respiration chamber). Thus, a direct comparison of individual mitochondria can be applied in the same or different tissues (Fedorenko et al., 2012). Using this method, we directly compared properties of IH in individual mitochondria from beige fat, brown fat, heart, and skeletal muscle under identical conditions. All tested mitochondria of inguinal beige fat had large UCP1dependent IH with the amplitude comparable to that of interscapular brown fat (Figure 2). In contrast, only about 15% of mitochondria of epididymal beige fat had UCP1-dependent IH. IH was undetectable in the remaining 85% of epididymal beige fat mitochondria, and in this respect they were identical to mitochondria of heart and skeletal muscle that express no UCP1 (Figure 2). Therefore, based on these electrophysiological data as well as on immunocytochemistry experiments (Figures 3 and S4), two distinct populations of beige fat cells can be identified, either UCP1 positive or UCP1 negative. We classify the UCP1-negative adipocytes as beige fat because they are similar to classic UCP1-positive beige adipocytes in several important ways. Multilocularity (Figure 1), robust mitochondrial biogenesis (Figures 1 and 3), activation of thermogenic gene program (Figure S2), distinctive OXPHOS profile (Figure 6), and creatine cycle (Figure 7) are the key similarities between the UCP1-negative and UCP1-positive cells. Interestingly, although UCP1 is not detectable on the protein level in UCP1-negative adipocytes (Figures 1, 2, and 3), Ucp1 mRNA is upregulated in these cells as compared to white fat (Figure S2). From all these data, it follows that the multilocular UCP1-negative cells are a type of a thermogenic beige adipocyte. In contrast, when multilocular UCP1-negative cells are compared

B

Veh

CL Na+/K+ ATPase

113 50

CKMT2

Inguinal Fat Veh

ATPase

50

D

F

CL Na+/K+

113

CKMT2

Inguinal Fat

Inguinal Fat

Inguinal Fat 5.0

***

4.0 3.0 2.0 1.0 0

Veh

Epididymal Fat CKMT2 / Na,K ATPase, (a.u.)

Epididymal Fat

CKMT2 / Na,K ATPase, (a.u.)

A

CL

Epididymal Fat

E

2.5

C

CK activity

*

2.0 1.5 1.0 0.5 0

Veh

CL

Inguinal Fat

ING

EPI

Epididymal Fat

Epididymal Fat

Figure 7. Creatine Stimulates Respiration in Inguinal and Epididymal Beige Fat Mitochondria (A) Representative immunoblots showing the effect of b3-adrenergic receptor stimulation on the protein level of Na+/K+ ATPase and CKMT2 in epididymal fat (upper panel) and inguinal fat (lower panel). See also Figure S6. (B) Histograms representing the protein level of CKMT2 relative to the level of Na+/K+ ATPase in inguinal fat (left panel) and epididymal fat (right panel). Data shown as mean ± SEM; n = 4. (C) CK activity of inguinal fat (ING) and epididymal fat (EPI) mitochondria after b3-adrenergic receptor stimulation. Data shown as mean ± SEM; n = 6. (D) Oxygen consumption by inguinal fat (left panel) and epididymal fat (right panel) mitochondria following treatment with vehicle or creatine (0.01 mM). Traces exhibit oxygen consumption of mitochondria during state 4 and following the addition of a limiting amount of ADP (0.1 mM). Data shown as mean ± SEM; n = 7 separate mitochondrial preparations for inguinal and epididymal fat. (E) The vehicle- and creatine-dependent oxygen consumption traces under ADP-limiting (0.1 mM) conditions are the same as those shown in (D), and are compared to the oxygen consumption traces following addition of saturating amounts of ADP (1 mM). Data shown as mean ± SEM; n = 7 separate mitochondrial preparations for inguinal and epididymal fat under ADP-limiting conditions. ADP saturating rates were obtained from two separate mitochondrial preparations. (F) Creatine-dependent basal and ADP-limiting oxygen consumption rate (OCR) in inguinal and epididymal beige fat mitochondria. The OCR of vehicle-treated beige fat mitochondria was subtracted from creatine-treated beige organelles. Data shown as mean ± SEM; n = 7 mitochondrial preparations for inguinal (left panel) and epididymal (right panel) adipose each. These data were obtained from the raw O2 traces shown in (D).

to white adipocytes, striking differences in cytoplasmic lipid droplet morphology, mitochondrial abundance, and thermogenic gene expression become obvious. Thus, multilocular

UCP1-negative adipocytes cannot be classified as white adipocytes because with their thermogenic program they burn fat, while white adipocytes by definition are a fat storage. Cell Metabolism 25, 811–822, April 4, 2017 819

Thus, we demonstrate the existence of a UCP1-negative thermogenic cell within white fat depots. Because this type of adipocyte does not express UCP1, it was previously overlooked as a potential target for obesity treatment. A better understanding of the thermogenic pathways employed by the UCP1-negative beige fat cells, such as the creatine cycle, can help to develop new approaches to the treatment of the metabolic syndrome. This approach may be especially useful for reduction of visceral fat depots, where UCP1 expression level is low. UCP1-Dependent Thermogenesis in Beige Fat UCP1-positive beige adipocytes, based solely on the amplitude of UCP1 current in these cells as compared to brown fat (Figures 2 and 4), appear to have a thermogenic potential similar to that of brown fat. However, functional properties of beige and brown fat UCP1 are not identical. There are two principal physiological regulators of UCP1: FAs and Mg2+-free purine nucleotides. As far as FAs are concerned, we found beige and brown fat UCP1 to be indistinguishable: (1) in both tissues FAs are required for H+ current via UCP1 and serve as UCP1 transport substrates, (2) the amplitude of the FA-dependent IH via UCP1 is comparable between beige and brown fat, and (3) the IMM of both beige and brown fat has PLA2 activity that generates FA and can contribute to UCP1 activation along with FAs derived from the hydrolysis of cytoplasmic lipid droplets. In contrast, purine nucleotides inhibit beige fat UCP1 with three times higher IC50 as compared to brown fat (Figure 5). This indicates that FAs would overcome purine nucleotide inhibition to activate UCP1-dependent thermogenesis more easily in beige fat compared to brown fat. Presently, it would be difficult to assess the physiological significance of this phenomenon or to explain its molecular mechanism. However, we can hypothesize that beige fat UCP1 must have a lower affinity for ATP due to a potentially higher level of local ATP concentration in the mitochondrial intermembrane space associated with a greater abundance of ATP synthase expression than brown fat (Figure 6). One possible explanation for the different affinities of brown and beige fat UCP1 for purine nucleotides could be different lipid compositions of the IMM of beige and brown fat, such as variations in the cardiolipin concentration. Cardiolipin binds to UCP1 (Lee et al., 2015) and has been shown to affect purine nucleotide binding to UCP1 (Klingenberg, 2010). Alternatively, it has recently been demonstrated that cysteine oxidation state regulates sensitivity to activation of UCP1-dependent respiration upon adrenergic stimulus in brown adipose tissue and brown adipocytes (Chouchani et al., 2016). Since this activation depends on release of inhibitory purine nucleotide binding of UCP1, the difference in binding between beige and brown fat observed herein could be explained by a potentially different oxidative state of beige and brown fat mitochondria. However, the actual mechanism behind the different affinities of brown and beige fat UCP1 for purine nucleotides remains to be established.

pocytes. In UCP1-positive inguinal beige adipocytes, this type of thermogenesis co-exists with classic UCP1-dependent thermogenesis (Kazak et al., 2015), while in UCP1-negative beige adipocytes it is the only mechanism of mitochondrial thermogenesis illustrated so far (Figures 7 and S6). Interestingly, in contrast to UCP1-dependent thermogenesis, creatine-driven thermogenesis is unlikely to be activated by b3-adrenergic receptor stimulation on the acute basis. Rather, once beige adipocytes are generated via chronic b3-adrenergic stimulation, creatinedriven thermogenesis in these cells would contribute to the basal metabolic rate. Thus, creatine-driven thermogenesis and UCP1dependent thermogenesis of beige fat are non-redundant and should vary in their regulatory mechanisms and physiological roles. UCP1-negative beige adipocytes, where creatine cycle is the only thermogenic mechanism, appear to be incapable of fast adaptive thermogenesis and may rather be responsible for slow modulation of the basal metabolic rate of white fat depots depending on the tonic level of norepinephrine. STAR+METHODS Detailed methods are provided in the online version of this paper and include the following: d d d d

d

KEY RESOURCES TABLE CONTACT FOR REAGENT AND RESOURCE SHARING EXPERIMENTAL MODEL AND SUBJECT DETAILS METHOD DETAILS B Isolation of mitochondria and mitoplasts B Patch-clamp recording B Patch-clamp data acquisition and analysis B Histology and immunohistochemistry B Immunoblots B Mitochondrial respiration B Creatine kinase activity B Gene expression analysis QUANTIFICATION AND STATISTICAL ANALYSIS

SUPPLEMENTAL INFORMATION Supplemental Information includes seven figures and can be found with this article online at http://dx.doi.org/10.1016/j.cmet.2017.03.002. AUTHOR CONTRIBUTIONS A.M.B. and Y.K. conceived the project and designed all experiments except for respirometry. A.M.B. performed electrophysiological experiments. A.M.B., M.G.B., I.P., G.L.W., and A.B. performed biochemical, immunocytochemistry, and gene expression profiling experiments. L.K., E.T.C., and B.M.S. designed and performed respiration experiments. S.K. consulted on CL injection of mice and immunocytochemistry experiments. A.M.B., Y.K., L.K., E.T.C., and B.M.S. wrote the manuscript. Y.K. and B.M.S. supervised all research. A.M.B., L.K., E.T.C., S.K., B.M.S., and Y.K. discussed the results and commented on the manuscript. ACKNOWLEDGMENTS

Creatine-Driven Thermogenesis in Beige Fat We demonstrate that regardless of UCP1 expression, creatinedriven respiration is a distinctive feature of beige adipocytes and can, in principle, be used along with mitochondrial biogenesis and multilocularity of lipid droplets to distinguish beige adi820 Cell Metabolism 25, 811–822, April 4, 2017

We thank the Histology and Light Microscopy Core at Gladstone Institutes for histology processing, and Nikon Imaging Center at UCSF for use of microscopy equipment. We thank members of the Y.K. lab for helpful discussions. This work was supported by NIH grants 5R01GM107710 (Y.K.) and DK 31405 (B.M.S.), the JPB Foundation (B.M.S.), the Canadian Institutes of Health

Research (L.K.), and the Human Frontier Science Program (E.T.C.). B.M.S. serves as a Consultant to Calico Life Sciences, LLC. Received: June 8, 2016 Revised: December 27, 2016 Accepted: March 4, 2017 Published: April 4, 2017 REFERENCES Bertholet, A.M., and Kirichok, Y. (2017). UCP1: A transporter for H(+) and fatty acid anions. Biochimie 134, 28–34. €ck, S. (2015). Human brown adipose tissue: what we Betz, M.J., and Enerba have learned so far. Diabetes 64, 2352–2360. Borecky´, J., Jezek, P., and Siemen, D. (1997). 108-pS channel in brown fat mitochondria might Be identical to the inner membrane anion channel. J. Biol. Chem. 272, 19282–19289.

Kazak, L., Chouchani, E.T., Jedrychowski, M.P., Erickson, B.K., Shinoda, K., Cohen, P., Vetrivelan, R., Lu, G.Z., Laznik-Bogoslavski, D., Hasenfuss, S.C., et al. (2015). A creatine-driven substrate cycle enhances energy expenditure and thermogenesis in beige fat. Cell 163, 643–655. Klingenberg, M. (2010). Wanderings in bioenergetics and biomembranes. Biochim. Biophys. Acta 1797, 579–594. Klingenberg, M., and Huang, S.G. (1999). Structure and function of the uncoupling protein from brown adipose tissue. Biochim. Biophys. Acta 1415, 271–296. Kozak, L.P., Koza, R.A., and Anunciado-Koza, R. (2010). Brown fat thermogenesis and body weight regulation in mice: relevance to humans. Int. J. Obes. 34 (Suppl 1 ), S23–S27. Lee, P., Smith, S., Linderman, J., Courville, A.B., Brychta, R.J., Dieckmann, W., Werner, C.D., Chen, K.Y., and Celi, F.S. (2014). Temperature-acclimated brown adipose tissue modulates insulin sensitivity in humans. Diabetes 63, 3686–3698.

Cannon, B., and Nedergaard, J. (2004). Brown adipose tissue: function and physiological significance. Physiol. Rev. 84, 277–359.

Lee, Y., Willers, C., Kunji, E.R., and Crichton, P.G. (2015). Uncoupling protein 1 binds one nucleotide per monomer and is stabilized by tightly bound cardiolipin. Proc. Natl. Acad. Sci. USA 112, 6973–6978.

Chouchani, E.T., Kazak, L., Jedrychowski, M.P., Lu, G.Z., Erickson, B.K., Szpyt, J., Pierce, K.A., Laznik-Bogoslavski, D., Vetrivelan, R., Clish, C.B., et al. (2016). Mitochondrial ROS regulate thermogenic energy expenditure and sulfenylation of UCP1. Nature 532, 112–116.

Lidell, M.E., Betz, M.J., Dahlqvist Leinhard, O., Heglind, M., Elander, L., Slawik, M., Mussack, T., Nilsson, D., Romu, T., Nuutila, P., et al. (2013). Evidence for two types of brown adipose tissue in humans. Nat. Med. 19, 631–634.

Cinti, S. (2009). Transdifferentiation properties of adipocytes in the adipose organ. Am. J. Physiol. Endocrinol. Metab. 297, E977–E986.

Loncar, D., Afzelius, B.A., and Cannon, B. (1988). Epididymal white adipose tissue after cold stress in rats. II. Mitochondrial changes. J. Ultrastruct. Mol. Struct. Res. 101, 199–209.

Cousin, B., Cinti, S., Morroni, M., Raimbault, S., Ricquier, D., Pe´nicaud, L., and Casteilla, L. (1992). Occurrence of brown adipocytes in rat white adipose tissue: molecular and morphological characterization. J. Cell Sci. 103, 931–942.

Long, J.Z., Svensson, K.J., Tsai, L., Zeng, X., Roh, H.C., Kong, X., Rao, R.R., Lou, J., Lokurkar, I., Baur, W., et al. (2014). A smooth muscle-like origin for beige adipocytes. Cell Metab. 19, 810–820.

Cypess, A.M., Lehman, S., Williams, G., Tal, I., Rodman, D., Goldfine, A.B., Kuo, F.C., Palmer, E.L., Tseng, Y.H., Doria, A., et al. (2009). Identification and importance of brown adipose tissue in adult humans. N. Engl. J. Med. 360, 1509–1517.

Nicholls, D.G. (2006). The physiological regulation of uncoupling proteins. Biochim. Biophys. Acta 1757, 459–466.

Cypess, A.M., Weiner, L.S., Roberts-Toler, C., Franquet Elı´a, E., Kessler, S.H., Kahn, P.A., English, J., Chatman, K., Trauger, S.A., Doria, A., and Kolodny, G.M. (2015). Activation of human brown adipose tissue by a b3-adrenergic receptor agonist. Cell Metab. 21, 33–38. de Jong, J.M., Larsson, O., Cannon, B., and Nedergaard, J. (2015). A stringent validation of mouse adipose tissue identity markers. Am. J. Physiol. Endocrinol. Metab. 308, E1085–E1105. Fedorenko, A., Lishko, P.V., and Kirichok, Y. (2012). Mechanism of fatty-aciddependent UCP1 uncoupling in brown fat mitochondria. Cell 151, 400–413. Fieni, F., Lee, S.B., Jan, Y.N., and Kirichok, Y. (2012). Activity of the mitochondrial calcium uniporter varies greatly between tissues. Nat. Commun. 3, 1317.

Nicholls, D.G., and Lindberg, O. (1973). Brown-adipose-tissue mitochondria. The influence of albumin and nucleotides on passive ion permeabilities. Eur. J. Biochem. 37, 523–530. Nicholls, D.G., and Locke, R.M. (1984). Thermogenic mechanisms in brown fat. Physiol. Rev. 64, 1–64. Palmieri, F. (2004). The mitochondrial transporter family (SLC25): physiological and pathological implications. Pflugers Arch. 447, 689–709. Rial, E., Poustie, A., and Nicholls, D.G. (1983). Brown-adipose-tissue mitochondria: the regulation of the 32000-Mr uncoupling protein by fatty acids and purine nucleotides. Eur. J. Biochem. 137, 197–203. Rosen, E.D., and Spiegelman, B.M. (2014). What we talk about when we talk about fat. Cell 156, 20–44.

Garlid, K.D., Jab urek, M., Jezek, P., and Varecha, M. (2000). How do uncoupling proteins uncouple? Biochim. Biophys. Acta 1459, 383–389.

Seale, P., Bjork, B., Yang, W., Kajimura, S., Chin, S., Kuang, S., Scime`, A., Devarakonda, S., Conroe, H.M., Erdjument-Bromage, H., et al. (2008). PRDM16 controls a brown fat/skeletal muscle switch. Nature 454, 961–967.

Granneman, J.G., Burnazi, M., Zhu, Z., and Schwamb, L.A. (2003). White adipose tissue contributes to UCP1-independent thermogenesis. Am. J. Physiol. Endocrinol. Metab. 285, E1230–E1236.

Shabalina, I.G., Jacobsson, A., Cannon, B., and Nedergaard, J. (2004). Native UCP1 displays simple competitive kinetics between the regulators purine nucleotides and fatty acids. J. Biol. Chem. 279, 38236–38248.

Granneman, J.G., Li, P., Zhu, Z., and Lu, Y. (2005). Metabolic and cellular plasticity in white adipose tissue I: effects of beta3-adrenergic receptor activation. Am. J. Physiol. Endocrinol. Metab. 289, E608–E616.

Shabalina, I.G., Petrovic, N., de Jong, J.M., Kalinovich, A.V., Cannon, B., and Nedergaard, J. (2013). UCP1 in brite/beige adipose tissue mitochondria is functionally thermogenic. Cell Rep. 5, 1196–1203.

Guerra, C., Koza, R.A., Yamashita, H., Walsh, K., and Kozak, L.P. (1998). Emergence of brown adipocytes in white fat in mice is under genetic control. Effects on body weight and adiposity. J. Clin. Invest. 102, 412–420.

Sharp, L.Z., Shinoda, K., Ohno, H., Scheel, D.W., Tomoda, E., Ruiz, L., Hu, H., Wang, L., Pavlova, Z., Gilsanz, V., and Kajimura, S. (2012). Human BAT possesses molecular signatures that resemble beige/brite cells. PLoS ONE 7, e49452.

Himms-Hagen, J., Melnyk, A., Zingaretti, M.C., Ceresi, E., Barbatelli, G., and Cinti, S. (2000). Multilocular fat cells in WAT of CL-316243-treated rats derive directly from white adipocytes. Am. J. Physiol. Cell Physiol. 279, C670–C681. Huang, S.G. (2003). Binding of fatty acids to the uncoupling protein from brown adipose tissue mitochondria. Arch. Biochem. Biophys. 412, 142–146.

Shinoda, K., Luijten, I.H., Hasegawa, Y., Hong, H., Sonne, S.B., Kim, M., Xue, R., Chondronikola, M., Cypess, A.M., Tseng, Y.H., et al. (2015). Genetic and functional characterization of clonally derived adult human brown adipocytes. Nat. Med. 21, 389–394.

Jespersen, N.Z., Larsen, T.J., Peijs, L., Daugaard, S., Homøe, P., Loft, A., de Jong, J., Mathur, N., Cannon, B., Nedergaard, J., et al. (2013). A classical brown adipose tissue mRNA signature partly overlaps with brite in the supraclavicular region of adult humans. Cell Metab. 17, 798–805.

Tseng, Y.H., Kokkotou, E., Schulz, T.J., Huang, T.L., Winnay, J.N., Taniguchi, C.M., Tran, T.T., Suzuki, R., Espinoza, D.O., Yamamoto, Y., et al. (2008). New role of bone morphogenetic protein 7 in brown adipogenesis and energy expenditure. Nature 454, 1000–1004.

Cell Metabolism 25, 811–822, April 4, 2017 821

Ukropec, J., Anunciado, R.P., Ravussin, Y., Hulver, M.W., and Kozak, L.P. (2006). UCP1-independent thermogenesis in white adipose tissue of coldacclimated Ucp1-/- mice. J. Biol. Chem. 281, 31894–31908.

Watt, I.N., Montgomery, M.G., Runswick, M.J., Leslie, A.G., and Walker, J.E. (2010). Bioenergetic cost of making an adenosine triphosphate molecule in animal mitochondria. Proc. Natl. Acad. Sci. USA 107, 16823–16827.

van Marken Lichtenbelt, W.D., Vanhommerig, J.W., Smulders, N.M., Drossaerts, J.M., Kemerink, G.J., Bouvy, N.D., Schrauwen, P., and Teule, G.J. (2009). Cold-activated brown adipose tissue in healthy men. N. Engl. J. Med. 360, 1500–1508.

Wikstrom, J.D., Mahdaviani, K., Liesa, M., Sereda, S.B., Si, Y., Las, G., Twig, G., Petrovic, N., Zingaretti, C., Graham, A., et al. (2014). Hormone-induced mitochondrial fission is utilized by brown adipocytes as an amplification pathway for energy expenditure. EMBO J. 33, 418–436.

Virtanen, K.A., Lidell, M.E., Orava, J., Heglind, M., Westergren, R., Niemi, T., €ck, S., and Nuutila, P. (2009). Taittonen, M., Laine, J., Savisto, N.J., Enerba Functional brown adipose tissue in healthy adults. N. Engl. J. Med. 360, 1518–1525.

Winkler, E., and Klingenberg, M. (1994). Effect of fatty acids on H+ transport activity of the reconstituted uncoupling protein. J. Biol. Chem. 269, 2508–2515.

Walde´n, T.B., Hansen, I.R., Timmons, J.A., Cannon, B., and Nedergaard, J. (2012). Recruited vs. nonrecruited molecular signatures of brown, ‘‘brite,’’ and white adipose tissues. Am. J. Physiol. Endocrinol. Metab. 302, E19–E31.

822 Cell Metabolism 25, 811–822, April 4, 2017

Wu, J., Bostro¨m, P., Sparks, L.M., Ye, L., Choi, J.H., Giang, A.H., Khandekar, M., Virtanen, K.A., Nuutila, P., Schaart, G., et al. (2012). Beige adipocytes are a distinct type of thermogenic fat cell in mouse and human. Cell 150, 366–376. Wu, J., Cohen, P., and Spiegelman, B.M. (2013). Adaptive thermogenesis in adipocytes: is beige the new brown? Genes Dev. 27, 234–250.

STAR+METHODS KEY RESOURCES TABLE

REAGENT or RESOURCE

SOURCE

IDENTIFIER

Rabbit polyclonal against UCP1

Abcam

ab10983; RRID: AB_2241462

Mouse monoclonal against OXPHOS MitoProfile

Abcam

ab110413; RRID: AB_2629281

Rabbit polyclonal against PGC1a

Abcam

ab54481; RRID: AB_881987

Antibodies

Mouse monoclonal against HSP60

Cayman

10011430; RRID: AB_10342684

Mouse monoclonal against COX IV

Abcam

ab33985; RRID: AB_879754

Rabbit monoclonal against Sodium-Potassium ATPase

Abcam

ab76020; RRID: AB_1310695

Rabbit polyclonal against TOM20

Santa Cruz

sc-11415; RRID: AB_2207533

Rabbit polyclonal against CKMT2

Abcam

ab71722; RRID: AB_1268273

Chemicals, Peptides, and Recombinant Proteins CL 316.243 hydrate

Sigma-Aldrich

Cat# C5976

Sucrose BioXtra

Sigma-Aldrich

Cat# S7903

HEPES BioXtra

Sigma-Aldrich

Cat# H7523

Ethylene glycol-bis(2-aminoethylether)-N,N,N0 , N0 -tetraacetic acid (EGTA)

Sigma-Aldrich

Cat# 03780

Trizma base

Sigma-Aldrich

Cat# T1503

D-Mannitol

Sigma-Aldrich

Cat# M4125

Potassium chloride

Sigma-Aldrich

Cat# 60128

EmbryoMax 0.1% Gelatin Solution

Emd Millipore

Cat# ES-006-B

Tetramethylammonium hydroxide solution

Sigma-Aldrich

Cat# 87729

Magnesium chloride

Sigma-Aldrich

Cat# 63068

D-Gluconic acid solution

Sigma-Aldrich

Cat# G1951

MES monohydrate

Sigma-Aldrich

Cat# 69892

Guanosine 50 -diphosphate tris salt from Saccharomyces cerevisiae

Sigma-Aldrich

Cat# G7252

Adenosine 50 -triphosphate disodium salt hydrate

Sigma-Aldrich

Cat# A6419

Oleic acid

Sigma-Aldrich

Cat# O1008

Methyl-b-cyclodextrin

Sigma-Aldrich

Cat# C4555

Sodium 1-hexanesulfonate monohydrate

Sigma-Aldrich

Cat# 52862

Arachidonic Acid sulfonate (sodium salt)

Cayman Chemical

Cat# 9001886

Formalin Solution, 10% (Histological)

Fisher Scientific

Cat# SF98-4

Ethanol, Pure, 200 Proof (100%)

Koptec

Cat# 64-17-5

Sodium Citrate Dihydrate, Fisher BioReagents

Fisher Bioreagents

Cat# BP327-1

Trito X-100

Sigma-Aldrich

Cat# T9284

TWEEN 20

Sigma-Aldrich

Cat# P7949

IGEPAL CA-630

Sigma-Aldrich

Cat# I8896

10% SDS Solution

Biorad

Cat# 1610416

Sodium deoxycholate

Sigma-Aldrich

Cat# D6750

Sodium chloride

Sigma-Aldrich

Cat# S7653

EDTA

Tecknova

Cat# E0306

Tris Buffer

UBS

Cat# 77514

cOmplet, Mini Protease Inhibitor Cocktail

Roche

Cat# 11836153001

QIAzol Lysis Reagent

QIAGEN

Cat# 79306

Sodium Pyruvate

Sigma-Aldrich

Cat# P5280

L-()-Malic acid

Sigma-Aldrich

Cat# M1000

Creatine monohydrate

Sigma-Aldrich

Cat# C3630 (Continued on next page)

Cell Metabolism 25, 811–822.e1–e4, April 4, 2017 e1

Continued REAGENT or RESOURCE

SOURCE

IDENTIFIER

Bovine Serum Albumin

Sigma-Aldrich

Cat# 6003

Sigma-Aldrich

Cat# A2754

SeaHorse Bioscience

http://www.agilent.com/en-us/promotions/ xftechnologyoverview

0

Adenosine 5 -diphosphate sodium salt Critical Commercial Assays XFe24 Extracellular Flux Analyzer Creatine Kinase Activity Assay Kit

Sigma-Aldrich

Cat# MAK116

RNeasy Mini Kit

QIAGEN

Cat# 74104

High-Capacity cDNA Reverse Transcription Kit

Applied Biosystems

Cat# 4368813

SYBR GreenE qPCR SuperMix Universal

Thermo Fisher Scientific

Cat# 11762500

Mouse: C57BL/6J

The Jackson Laboratory

JAX:000664

Mouse: B6.129-Ucp1tm1Kz/J

The Jackson Laboratory

JAX: 003124

Different sources

See table in Gene expression analysis section

Molecular Devices

https://www.moleculardevices.com/systems/ conventional-patch-clamp/pclamp-10-software

Origin 7.5

OriginLab

http://www.originlab.com/

GraphPad Prism 7

GraphPad Software

https://www.graphpad.com/

ImageJ Software

ImageJ

https://imagej.net/

Wave 2.2.0

Agilent

http://www.agilent.com/en-us/support/cell-analysis(seahorse)/seahorse-xf-software

Experimental Models: Organisms/Strains

Oligonucleotides Mouse qRT-PCR Primer Sequences Software and Algorithms PClamp 10

CONTACT FOR REAGENT AND RESOURCE SHARING Further information and requests for reagents may be directed to and will be fulfilled by the Lead Contact, Yuriy Kirichok (yuriy. [email protected]). EXPERIMENTAL MODEL AND SUBJECT DETAILS All animal experiments were performed according to procedures approved by the UCSF’s Institutional Animal Care and Use Committee and adhered to NIH standards. C57BL/6J wild-type and UCP1(/) mice (2-5 months old, male) from the Jackson Laboratory were maintained on a standard rodent chow diet with 12 hr light and dark cycles. CL316.243 (Sigma) at 1mg/kg was injected intraperitoneally into mice daily for 10 days. METHOD DETAILS Isolation of mitochondria and mitoplasts Wild-type and UCP1(/) mice were sacrificed by CO2 asphyxiation followed by cervical dislocation. Mitoplasts were prepared from heart, skeletal muscle, brown fat, inguinal fat, and epididymal fat. The selected mouse tissue was isolated, rinsed and homogenized in ice-cold medium containing 250cmM sucrose, 10cmM HEPES, and 1cmM EGTA (pH adjusted to 7.2 with Trizma base), using a glass grinder with six slow strokes of a Teflon pestle rotating at 275 (soft tissues) or 600 (fibrous tissues) rotations per minute. The homogenate was centrifuged at 700cg for 5–10cmin to pellet nuclei and unbroken cells. For some tissues, the first nuclear pellet was resuspended in the same solution and homogenized again to increase the yield of mitochondria. Mitochondria were collected by centrifugation of the supernatant at 8,500cg for 10cmin. Mitoplasts were produced from mitochondria using a French press. Briefly, mitochondria were suspended in a solution containing 140cmM sucrose, 440cmM D-mannitol, 5cmM HEPES, and 1cmM EGTA (pH adjusted to 7.2 with Trizma base), and then subjected to a French press at 2,000cpsi to rupture the outer membrane. Mitoplasts were pelleted at 10,500cg for 15cmin and resuspended for storage in 500cml of solution containing 750cmM KCl, 100cmM HEPES and 1cmM EGTA (pH adjusted to 7.2 with Trizma base). Mitochondria and mitoplasts were prepared at 0–4c C and stored on ice for up to 5ch. Immediately before the electrophysiological experiments, 15–50 mL of the mitoplast suspension was added to 500 mL solution containing 150 mM KCl, 10 mM HEPES, and 1 mM EGTA (pH adjusted to 7.0 with Trizma base) and plated on 5 mm coverslips pretreated with 0.1% gelatin to reduce mitoplast adhesion. e2 Cell Metabolism 25, 811–822.e1–e4, April 4, 2017

Patch-clamp recording Mitoplasts used for patch-clamp experiments were 3–5 mm in diameter and typically had membrane capacitances of 0.5–1.2 pF. Gigaohm seals were formed in the bath solution containing 150 mM KCl, 10 mM HEPES, and 1 mM EGTA, pH 7 (adjusted with Trizma base). A 3M KCl agar salt bridge was used as the bath reference electrode. Voltage steps of 250–500 mV and 1–50 ms were applied to obtain the whole-mitoplast configuration, as monitored by the appearance of capacitance transients. The access resistance and membrane capacitance of mitoplasts were determined with the Membrane Test tool of pClamp 10 (Molecular Devices). Mitoplasts were stimulated every 5 s. Pipettes were filled with 130 mM tetramethylammonium hydroxide (TMA), 1.5 mM EGTA, 2 mM magnesium chloride, 150 mM HEPES (or MES), (pH adjusted to 6.0–8.0 with D-gluconic acid, tonicity adjusted to 360 mmol/kg with sucrose). Typically, pipettes had resistances of 25–40 MU, and the access resistance was 40–75 MU. The calculated voltage-clamp error associated with the access resistance did not exceed 10 mV. Whole-mitoplast UCP1 current was recorded in the bath solution containing 150 mM HEPES (or MES) and 1 mM EGTA (pH adjusted to 6.0–8.0 with Trizma base, tonicity adjusted to 300 mmol/kg with sucrose). All experiments were performed under continuous perfusion of the bath solution. Patch-clamp data acquisition and analysis Data acquisition and analysis were performed using PClamp 10 (Molecular Devices) and Origin 7.5 (OriginLab). All electrophysiological data presented were acquired at 10 kHz and filtered at 1 kHz. Amplitudes of H+ currents were measured 25 ms after application of the 160 mV voltage step. Histology and immunohistochemistry Tissues (BAT, inguinal fat, epididymal fat) were dissected, fixed in 10% formalin for 24 hr, and subsequently stored in 70% ethanol prior to paraffin embedding (5 mice per group). Paraffin-embedded 7 mm slices and hematoxylin/eosin staining were produced by the Gladstone Histology and Light Microscopy Core (Gladstone Institutes, San Francisco, CA). For immunocytochemistry, slices were subjected to citrate-based antigen retrieval and then permeabilized by incubation in 0.5% Triton X-100 in PBS for 30 min. Slices were incubated for 2 hr at room temperature with rabbit polyclonal antibodies against UCP1 at a dilution of 1:100 (Abcam, ab10983) and mouse OXPHOS antibodies at a dilution of 1:50 (Abcam, ab110413). After extensive washing in PBS, secondary antibodies anti-rabbit Alexa 488 and anti-mouse Alexa 555 (Invitrogen) were added for 45 min. Slices were mounted with Mowiol. Images were captured using a CSU-W1 spinning disk with Borealis upgrade on a Nikon Ti inverted microscope. Mitochondrial size was measured using ImageJ software and confocal images ( 3 100). Average mitochondrial size was determined from 1500 mitochondria of 44 UCP1-positive beige adipocytes and 2500 mitochondria of 65 UCP1-negative adipocytes. Immunoblots For western blot analysis, tissues were lysed in RIPA buffer (1% Igepal, 0.1% sodium dodecyl sulfate, 0.5% sodium deoxycholate, 150 mM NaCl, 1 mM EDTA, 50 mM Tris-HCl [pH 7.4] and a cocktail of proteases inhibitors). Lysates were resolved by SDS-PAGE; transferred to PVDF membrane (Millipore); and probed with anti-UCP1 (Abcam, ab10983), anti-PGC1 a (Abcam, ab54481), antiHSP60 (Cayman, 10011430), anti-COXIV (Abcam, ab33985), anti-Sodium Potassium ATPase antibody (Abcam, ab76020) antiTOM20 (Santa Cruz, sc-11415), OXPHOS cocktail (Abcam, ab110413) and anti-CKMT2 (ab71722). Mitochondrial respiration Mitochondrial respiration was determined using an XFe24 Extracellular Flux Analyzer (Seahorse Bioscience) using 15 mg mitochondrial protein in a buffer containing 4% BSA, 10mM Pyruvate, 5mM Malate, 1mM GDP. Mitochondria were plated and centrifuged 2,000 g for 20 min to promote adherence to the V7 cell culture microplate. ADP addition was used at final concentration of 0.1 mM for subcutaneous and epididymal mitochondria. Calculation of Creatine:ADP Stoichiometry was done as previously reported (Kazak et al., 2015). Briefly, creatine was added to mitochondria at a concentration of 0.01 mM in a total volume of 0.555ml in the XF well. Thus, the total pmoles of creatine in the entire well was 5,550 pmoles. The XF microchamber volume during measurement of the oxygen consumption rate is 2 – 7 ml. Based on a conservative estimate of a 7 mL microchamber volume, this is equivalent to a total of 77 pmoles creatine in the microchamber. The total pmoles of oxygen consumed was measured over a 3 min period, based on the oxygen consumption rate (measured as pmoles/min). The quantity of ADP molecules phosphorylated to ATP was calculated based on the mitochondrial P/O ratio (Watt et al., 2010). Creatine kinase activity Creatine kinase activity from sucrose-gradient isolated mitochondria (0.25 mg total protein) was examined using the Creatine Kinase Activity Assay Kit (Sigma) according to the manufacturer’s instructions. Gene expression analysis Epididymal and inguinal fat tissues were dissected from male mice (9 per group), immediately followed by total RNA extraction using QIAZOL (QIAGEN) and purified with RNeasy Mini kit (QIAGEN). Total RNA was reverse transcribed using a High-Capacity cDNA Reverse Transcription kit (Applied Biosystems). Obtained cDNA was diluted to 15ng/uL and used for qPCR analysis. 15ng of cDNA and final concentration of 0.2uM of each primer were mixed with SYBR GreenERTM qPCR Supermix Universal (Thermo Fisher Scientific) according to manufacturer’s protocol. Reactions were performed in a 96-well format using Stratgene Cell Metabolism 25, 811–822.e1–e4, April 4, 2017 e3

Mx3000 real time PCR system (Agilent Technologies). Relative mRNA levels were calculated using the comparative CT method and normalized to TFIIB mRNA. A complete list of primers and sequences can be found below.

Mouse qRT-PCR Primer Sequences Gene

Sequences 50 / 30

Sequence Source

TFIIB

Forward, TGGAGATTTGTCCACCATGA; Reverse, GAATTGCCAAACTCATCAAAACT

(de Jong et al., 2015)

Ucp1

Forward, ACTGCCACACCTCCAGTCATT; Reverse, CTTTGCCTCACTCAGGATTGG

(Kazak et al., 2015)

Dio2

Forward, CATGCTGACCTCAGAAGGGC; Reverse, CCCAGTTTAACCTGTTTGTAGGCA

NCBI/Primer Blast tool

Slc6a8

Forward, TGCATATCTCCAAGGTGGCAG; Reverse, CTACAAACTGGCTGTCCAGA

(Kazak et al., 2015)

Gamt

Forward, GCAGCCACATAAGGTTGTTCC; Reverse, CTCTTCAGACAGCGGGTACG

(Kazak et al., 2015)

Gatm

Forward, CAGACACAAATTGGCCGCTC; Reverse, CCCAGGTAGTTTGTAACCTGGC

NCBI/Primer Blast tool

Ckmt1

Forward, GGCCTCAAAGAGGTGGAGAA; Reverse, CAGGATCTTTGGGAAGCGGT

NCBI/Primer Blast tool

Ckmt2

Forward, GCATGGTGGCTGGTGATGAG; Reverse, AAACTGCCCGTGAGTAATCTTG

(Kazak et al., 2015)

Cidea

Forward, TGCTCTTCTGTATCGCCCAGT; Reverse, GCCGTGTTAAGGAATCTGCTG

(Tseng et al., 2008)

Pgc1a

Forward, CCCTGCCATTGTTAAGACC; Reverse, TGCTGCTGTTCCTGTTTTC

(Kazak et al., 2015)

Prdm16

Forward, CCCACCAGACTTCGAGCTAC; Reverse, ATCCGTCAGCATCTCCCATC

NCBI/Primer Blast tool

Phospho1

Forward, AAGCACATCATCCACAGTCCCTC; Reverse, TTGGTCTCCAGCTGTCATCCAG

(Kazak et al., 2015)

QUANTIFICATION AND STATISTICAL ANALYSIS Student’s t test was used to determine the significance between two groups with GraphPad Prism 7. Unless otherwise specified, *p < 0.05, **p < 0.01, and ***p < 0.001. Errors bars plotted on graphs are presented as the mean ± SEM.

e4 Cell Metabolism 25, 811–822.e1–e4, April 4, 2017

Cell Metabolism, Volume 25

Supplemental Information

Mitochondrial Patch Clamp of Beige Adipocytes Reveals UCP1-Positive and UCP1-Negative Cells Both Exhibiting Futile Creatine Cycling Ambre M. Bertholet, Lawrence Kazak, Edward T. Chouchani, Marta G. Bogaczy nska, Ishan Paranjpe, Gabrielle L. Wainwright, Alexandre Bétourné, Shingo Kajimura, Bruce M. Spiegelman, and Yuriy Kirichok

PGC1α

PGC1α

55

Tubulin

35

UCP1

15

COX IV

COX IV / HSP60, (a.u.)

95

1.5

1.0

0.5 0

1.0

CL

0.5

0

Veh

PGC1α

Veh

CL

35

UCP1

15

COX IV

17

TOM20

2.5

TOM20 / Na,K ATPase, (a.u.)

PGC1α

95

PGC1α / Na,K ATPase, (a.u.)

ATPase

TOM20

CL

2.0 1.5 1.0 0.5 0

COX IV 0.6

2.5 COX IV / TOM20, (a.u.)

CL316.243 Na+/K+

113

2.0 1.5 1.0 0.5

Veh

CL

0.4

0.2

0

0

Veh

Veh

CL

CL

Epididymal Fat (UCP1-/-)

PGC1α

15

COX IV

17

TOM20

1.0

0.5

TOM20 / Na,K ATPase, (a.u.)

35

UCP1

0

Veh

Inguinal Fat (UCP1-/-) Vehicle CL316.243

95

PGC1α

35

UCP1

15

COX IV

17

TOM20

1.5

2.0 1.5 1.0 0.5 0

0.6 0.4 0.2 0

Veh

CL

0.5

Veh

CL

*

3.0

0.4

0.2

CL

COX IV

0.6

*

1.0

TOM20

1.0 0.8

*

0

Veh

PGC1α PGC1α / Na,K ATPase, (a.u.)

113

Na+/K+ ATPase

COX IV

*

2.5

CL

TOM20 / Na,K ATPase, (a.u.)

95

PGC1α

1.5 PGC1α / Na,K ATPase, (a.u.)

ATPase

TOM20

COX IV / TOM20, (a.u.)

Na+/K+

COX IV / TOM20, (a.u.)

CL316.243

UNDETECTABLE

Vehicle

D

2.0

1.0

Brown Fat (UCP1-/-) Vehicle

113

3.0

0

Veh

C

UCP1 1.5

HSP60

55

B

COX IV 4.0

2.0

UCP1 / HSP60, (a.u.)

Brown Fat (WT) Saline CL316.243 PGC1α / HSP60, (a.u.)

A

2.0

1.0

0

0

Veh

CL

Veh

Figure S1. Effect of chronic β3-adrenergic receptor stimulation on brown adipose tissue. Related to Figure 1.

CL

A

Epididymal Fat

**

250

Relative mRNA (Normlaized to TFIIB)

200 50

** ****

40 30 20 10

* 0

Veh CL Cidea

B

Veh CL

Veh CL

Dio2

Prdm16

Veh

CL

Ucp1

Veh

CL

Pgc1α

Inguinal Fat

***

300

Relative mRNA (Normlaized to TFIIB)

250

***

120 100 80

60 40 20

** *

0

Veh CL Cidea

Veh

CL

Dio2

Veh

CL

Prdm16

Veh

CL

Ucp1

Figure S2. Thermogenic profile of epididymal and inguinal fat. Related to Figure 1

Veh

CL

Pgc1α

A

0 mV

850 ms

+100 mV

B

0 mV

850 ms

+100 mV

C

Inguinal fat (UCP1-/-)

850 ms

+100 mV

-160 mV

-160 mV

-160 mV

0 mV

Brown fat (UCP1-/-)

Epididymal fat (WT)

150 mM Cl-

150 mM Cl-

150 mM Cl-

control

control

control pHi 7.0

50 pA/pF 200 ms

pHo 7.0

D

0 mV

850 ms

+100 mV

E

0 mV

850 ms

+100 mV

F

-160 mV

-160 mV

Skeletal muscle (WT)

150 mM Cl-

control

Heart (WT)

150 mM Cl-

control

Chloride current (pA/pF)

250 200 150 100 50 0

Figure S3. Mitochondrial Cl- current in various mouse tissues. Related to Figure 3.

n=9

n=8

n=16

n=6

n=5

BAT

ING

EPI

Heart

SM

UCP1-/-

UCP1-/-

WT

A

INGUINAL VEHICLE

EPIDIDYMAL VEHICLE

CL316.24 3

OXPHOS

UCP1

20 μM

CL316.243

B

Figure S4. Mitochondrial biogenesis and UCP1 expression in inguinal and epididymal depots after chronic β3adrenergic stimulation. Related to Figure 3.

A 4

OXPHOS

1

4

1

3

2

20 μM

UCP1

3

2

B Area 1

OXPHOS Area 3

(UCP1+)

(UCP1-)

OXPHOS

1 μM

4 μM

OXPHOS Area 4

(UCP1+)

(UCP1-)

Mean size of mitochondria (μm)

C

Area 2

OXPHOS

2.0

* 1.5 1.0 0.5 n=44 0

n=65

UCP1+ UCP1-

Figure S5. Mitochondrial size in UCP1+ and UCP1- beige adipocytes. Related to Figure 3

Epididymal Fat

A 10

** Relative mRNA (Normlaized to TFIIB)

8

6

4

* 2

*

* 0

Veh CL Veh CL Phospho1 Slc6a8

B

Veh

CL

Gamt

Veh CL

Veh CL

Gatm

Ckmt1

Veh

CL

Ckmt2

Inguinal Fat 8

Relative mRNA (Normlaized to TFIIB)

** **

6

4

**

2

0

Veh CL Veh CL Phospho1 Slc6a8

Veh

CL

Gamt

Veh

CL

Gatm

Veh

CL

Ckmt1

Figure S6. mRNA levels of proteins involved in creatine metabolism. Related to Figure 7

Veh

CL

Ckmt2

A UCP1 transports short-chain FA anions +50mV

+50mV

FA-

0mV -50mV

-

short-chain FA- current

FA-

UCP1

FA-

10 pA/pF

-

250 ms

B UCP1 traps long-chain FA anions long-chain

FA-

current

FA-

-

FA-

UCP1

-

FA-

C UCP1 operates as a FA-dependent H+ carrier in the presence of protonateable long-chain FA H+ current FA-

H+ +

-

FA-

UCP1

H+ +

-

FA-

+

H+

Figure S7. The mechanism by which FA control UCP1 currents in beige and brown fat. Related to Figure 4.

SUPPLEMENTAL FIGURE LEGENDS Figure S1. Effect of chronic β3-adrenergic receptor stimulation on brown adipose tissue. Related to Figure 1 (A) Representative immunoblot showing the effect of β3-adrenergic receptor stimulation on the expression of Tubulin, PGC1α, UCP1, COX IV, and HSP60 proteins in brown fat. Histograms show PGC1α, COX IV and UCP1 protein levels relative to the levels of HSP60. Data shown as mean ± SEM, n = 3. (B-D) Representative immunoblots showing the effect of β3-adrenergic receptor stimulation on the expression of Na+/K+ ATPase, PGC1α, UCP1, COX IV and TOM20 proteins in brown fat (B), inguinal fat (C), and epididymal fat (D) of UCP1-/- mice. Histograms show PGC1α, TOM20 and COX IV protein levels for each immunoblot (B-D). Data shown as mean ± SEM, n = 4.

Figure S2. Thermogenic profile of epididymal and inguinal fat. Related to Figure 1 Quantitative RT-PCR of Cidea, Dio2, Prdm16, Ucp1, Pgc1α from epididymal fat (A) and inguinal fat (B) showing the effect of β3-adrenergic receptor stimulation on the relative mRNA level for these genes. Data shown as mean ± SEM. n = 6 to 9.

Figure S3. Mitochondrial Cl- current in various mouse tissues. Related to Figure 3. (A-E) Representative Cl- currents recorded in mitoplasts isolated from inguinal beige fat (UCP1-/-, A), brown fat (UCP1-/-, B), epididymal beige fat (UCP1-negative mitoplast, WT, C), skeletal muscle (WT, D), and heart (WT, E). Control traces were recorded in HEPES/Tris solution and Cl- currents (blue) were induced by addition of KCl to the cytosolic face of the IMM (blue, see methods). The voltage protocol is indicated at the top. (F) Bar graph showing average current density of Cl- in brown fat (BAT, UCP1-/-), inguinal beige fat (ING, UCP1-/-), epididymal beige fat (EPI, UCP1-negative mitoplasts, WT), heart (WT), skeletal muscle (SM, WT). Data shown as mean ± SEM.

Figure S4. Mitochondrial biogenesis and UCP1 expression in inguinal and epididymal depots after chronic β3-adrenergic stimulation. Related to Figure 3. (A and B) Confocal micrographs of inguinal fat and epididymal fat of mice treated with vehicle or CL316.243 and immunolabeled with UCP1 (upper panels) and with respiratory chain complex antibodies (MitoProfile® Total OXPHOS, lower panels); scale bar, 20 μm).

Figure S5. Mitochondrial size in UCP1+ and UCP1- beige adipocytes. Related to Figure 3 (A) A representative confocal micrograph of epididymal fat of CL-treated mice, immunolabeled with respiratory chain complex antibodies (MitoProfile® Total OXPHOS, left panel) and with UCP1 antibody (right panel); scale bar, 20 μm. (B) Magnified views of areas 1 and 2 (UCP1 positive adipocytes), and areas 3 and 4 (UCP1 negative adipocytes) of panel A. Mitochondrial networks are shown. (C) Histogram showing the mean size of mitochondria of UCP1 positive (UCP1+) and UCP1 negative (UCP1-) adipocytes based on panel B. Data shown as mean ± SEM.

Figure S6. mRNA levels of proteins involved in creatine metabolism. Related to Figure 7 (A and B) Quantitative RT-PCR of Phospho1, Slc6a8, Gamt, Gatm, Ckmt1 and Ckmt2 from epididymal (A) and inguinal fat (B), showing the effect of β3-adrenergic receptor stimulation. Data shown as mean ± SEM. n = 6 to 9.

Figure S7. The mechanism by which FA control UCP1 currents in beige and brown fat. Related to Figure 4 (A) Right panel: an example of steady UCP1 current induced by short-chain low-pKa FA analogs added on the cytosolic face of the IMM. Voltage protocol is shown above. Left panel: the mechanism of this current. Shortchain FA- are transported by UCP1 across the IMM. (B) Right panel: an example of transient UCP1 current induced by long-chain low-pKa FA analogs added on the cytosolic face of the IMM. Left panel: the mechanism of this current. A long-chain FA- analog is translocated by UCP1 similar to short-chain FA-, however the long carbon tail of FA- establishes strong hydrophobic interaction with UCP1 to prevent FA- dissociation. Thus, the negatively charged FA- shuttles within the UCP1 translocation pathway in response to the transmembrane voltage, producing transient currents. (C) Right panel: an example of steady H+ current via UCP1 induced by regular long-chain FA added on the cytosolic face of the IMM. Left panel: the mechanism of this current. UCP1 operates as a symporter that transports one FA- and one H+ per the transport cycle. The H+ and the FA- are translocated by UCP1 upon conformational change, and H+ is released on the opposite side of the IMM, while the FA- stays associated with UCP1 due to the hydrophobic interactions established by its carbon tail. The FA- anion then returns to initiate another H+ translocation cycle. Charge is translocated only in step 3 when the LCFA anion returns without the H+.