Cell Metabolism, Volume 23

Supplemental Information

Adult NG2-Glia Are Required for Median

Eminence-Mediated Leptin Sensing

and Body Weight Control

Tina Djogo, Sarah C. Robins, Sarah Schneider, Darya Kryzskaya, Xiaohong Liu, AndrewMingay, Colleen J. Gillon, Joo Hyun Kim, Kai-Florian Storch, Ulrich Boehm, Charles W.Bourque, Thomas Stroh, Leda Dimou, and Maia V. Kokoeva

Light Dark Light Dark

Food

inta

ke (g

)

4 3

0 1 2

##   **   ns

##  

FEMALES  

Food

inta

ke (g

/h)

Light Dark

0.6

0.4

0.0

0.2

B

aCSF AraC

Food

inta

ke (g

/h)

Light Dark

0.4

0.0

0.2

Food

inta

ke (g

)

##   #  4 3

0 Light Dark Light Dark

1 2

***  ns

MALES  

A aCSF AraC

aCSF AraC

0.6

Light Dark aCSF AraC

Light Dark Dark Light

# ns

ns

# 40 30

10 20

Loco

mot

or a

ctiv

ity

(cnt

s x1

03/h

)

0

6

2

4

0

Loco

mot

or a

ctiv

ity

(cnt

s x1

03) C

Light Dark aCSF AraC

Light Dark Dark Light

# ns **  

#

D

Loco

mot

or a

ctiv

ity

(cnt

s x1

03/h

)

6

2 4

0

8 10

Loco

mot

or a

ctiv

ity

(cnt

s x1

03)

60

20 40

0

80

males females

E

Loco

mot

or a

ctiv

ity

(cnt

s x1

03/d

ay)

100

n.s.

**  

50

0 aCSF AraC aCSF AraC F   icv  aCSF  or  AraC  

 0     7   14  days  

running  wheel  access  in  constant  darkness  

G  

24  hrs   24  hrs  days  

days  

Percentiles: 0.2 2.5 12.0 25.8 41.2893 (80)

7 15 23 7 15 23 7

05/13/13

05/11/13

05/09/13

05/07/13

05/05/13

05/03/13

24  hrs   Percentiles: 0.2 8.7 20.0 33.7 48.8902 (72)

7 15 23 7 15 23 7

05/13/13

05/11/13

05/09/13

05/07/13

05/05/13

05/03/13

24  hrs   Percentiles: 0.2 6.3 20.5 38.0 53.8906 (76)

7 15 23 7 15 23 7

05/13/13

05/11/13

05/09/13

05/07/13

05/05/13

05/03/13

H  

aCSF  

AraC  Percentiles: 0.2 2.8 10.8 19.5 31.0896 (52)

7 15 23 7 15 23 7

05/13/13

05/11/13

05/09/13

05/07/13

05/05/13

05/03/13

I  

23.5

23.7

23.9

24.1

aCSF AraC (Gainers)

Per

iod

(h)  ns  

J  

0

300

600

900

aCSF AraC (Gainers)

Pow

er

 ns  

Figure S1

Figure S1. Food Intake, Locomotor Activity and Circadian Locomotor Analysis of AraC-infused Mice. Related to Figure 1. (A-E) Temporal pattern of food intake and locomotor activity. Both AraC-treated male (A) and female (B) mice showed elevated food intake specifically during the light period, whereas night-time feeding remained indifferent between groups. Locomotor activity was significantly reduced at night in AraC-treated females, whereas males showed only a trend in this regard (C, D). Traces show the 24 hr pattern of food intake, or activity averaged across 9 consecutive days (post-surgery day 33 to 42). Bar graphs depict day/night hourly averages of food intake (A, B) and activity (C, D), or total 24 h-locomotor activity (E). n = 4, **p <0.01, ***p < 0.001, paired t test; #p <0.05, ##p <0.01, ###p < 0.001, unpaired t test; n.s., not significant. Shown are mean ± SEM. (F-J) Running wheel activity analysis of aCSF- and AraC-infused animals in constant darkness. One day post icv-surgery, animals were provided with a running wheel and transferred to constant darkness (F). Animals were infused with aCSF or AraC for one week. Activity recording started on the first day of running wheel access. (G, H) Representative double-plotted actograms of running wheel activity in aCSF-infused mice (G) or AraC-infused mice which exhibited excessive weight gain (gainers, H). (I, J) Circadian locomotor activity period (I) and circadian amplitude (power from periodogram analyses) (J) in controls and AraC-infused gainers; ns, not significant; unpaired t test; n = 5-6; mean ± SEM.

Figure S2. NG2-glia Ablation upon AraC Infusion at Discrete Ventricular System Sites. Related to Figure 1. (A-L) Immunodetection of NG2 in mice infused with AraC into the lateral (D-F), third (G-I) or fourth ventricle (J-L). Control mice (A-C) were infused with aCSF into the right lateral ventricle. (M) Illustration of the infusion strategy. Blue arrow indicates the direction of CSF flow. Mice were sacrificed at the end of the infusion period. LV, lateral ventricle; 3V, third ventricle; 4V, fourth ventricle. Dashed lines demarcate the border of the NG2-glia depleted territory. Scale bars = 200 µm.

Frac

tion

of B

rdU

+ ce

lls (%

)

Hypo ME

A

3V

Iba1 BrdU

30 hr BrdU

Iba1

BrdU

3V

NG2 BrdU

30 hr BrdU

NG2

BrdU

C

B

ME ME

Iba1 BrdU

NG2 BrdU

Perfusion (16wpi)

NG2-glia GFP expressing

E Esco2 expressing

NG2-glia derivatives

Perfusion (1wpi) Tamoxifen

Sox

10,

Esc

owt/w

t , G

FP

Sox

10,

Esc

ofl/f

l ,GFP

E

v E E E E

E

E E

E

E E

E

E E E

E

E

E E

E v

E E

E

E E E

E

E

E E

E E

E E

E

E

v

v

E

E

E E E

E E

E

E

E E

E E E

E

E

v E

E

E

E v E E

E E E E

E

E

v E E

E v E E

E

E

E

D

Figure S3

Figure S3. Proliferating Glia in the Hypothalamus and Illustration of the Esco2–based Ablation Model. Related to Figure 2. (A, B) Immunodetection of proliferative microglia (A) or NG2-glia (B) in mice that received six i.p. BrdU injections within the 30 hrs prior to perfusion. Boxed areas are shown enlarged next to the image. (C) Relative distribution (in %) of cell types among the population of proliferating (BrdU-positive) cells found in the hypothalamus (Hypo) and ME. (D) Schematic depicting the TX treatment regimen to conditionally inactivate Esco2. Esco2 disruption causes cell cycle arrest in S-phase and thus death of dividing cells (hashed cells, bottom panel). However, once NG2-glia differentiate into OLG (depicted by hexagons), they become postmitotic and protected from Esco2 disruption-mediated cell death.

A B

0

5

10

15

20

1  

BW

gai

n (%

)

*** ***

ns

0

Pump implantation

perfusion

8 days

icv aCSF or AraC

gained 3.52 g gained 3.48 g

gained 3.86 g gained 3.92 g

3V 3V

3V 3V

gained 0.69 g

3V

pSTAT3

AraC HCD gained 0.14 g

3V

aCSF C D E gained 3.44 g

3V

gained 4.39 g

3V

HCD (ip vehicle) F

Figure S4

Figure S4. pSTAT3 Induction in the ArcN of AraC-treated and HCD-fed Mice. Related to Figure 3. (A, B) Percent weight gain (A) in mice on post-surgery day 8 (B), in comparison to mice that exhibited matching gains in body weight in response to a high-calorie diet (HCD) (n = 6-10). ***p<0.001, unpaired t test. Mean ± SEM. (C-E) Immunodetection of pSTAT3 in aCSF- (C), AraC-treated (D), and HCD (E) mice 45 min after i.p. leptin on day 6 of AraC infusion. Shown are fluorescent images from 2 representative animals each. Body weight increases relative to the day prior to pump implantation are indicated at top of the images. 3V, third ventricle. Scale bar, 100 µm.

A B

C E

F D

aCSF AraC

0  

5  

10  

15  

20  

fAH

P (m

V)

-­‐60  

-­‐40  

-­‐20  

0  

RM

P (m

V)

0.0  

0.5  

1.0  

1.5  

2.0  

Rm

(GO

hm)

0  

10  

20  

30  

40  

Ci (

pFar

ad)

G H I J

0  

50  

100  

150  

AP

ampl

itude

(mV

)

0.0  

0.5  

1.0  

1.5  

AP

dura

tion

(mse

c)

0  

50  

100  

150  

200  

Vm

ax (V

/sec

)

K L M aCSF AraC

aCSF, ob/ob mice AraC, ob/ob mice

0  

20  

40  

60  

80  

lep9n   MTII  

* ns

Food

inta

ke (%

of v

ehic

le)

O N

0  1  2  3  4  5  

lep9n   vehicle   MTII  

Food

inta

ke (g

)

**

ns

ns

90  

100  

110  

120  

0   5   10   15   20  

BW

(% o

f ini

tial)

Time after surgery (days)

P AraC, ob/ob mice aCSF, ob/ob mice

Figure S5

Figure S5. Basal Electrical Properties of ArcN LepR Neurons; Effect of AraC on ob/ob Mice. Related to Figure 4. (A-F) Baseline electrical activity traces from current-clamped, ArcN Tom+ neurons (A, B) and electrical responses to current injection (100 -350 pA, step size 50 pA) (C, E) with typical single action potential wave forms (D, F). (G-M) Basal electrical membrane properties of ArcN Tom+ neurons. Shown are mean ± SEM of the resting membrane potential (G), fast after-hyperpolarisation potential (H), membrane resistance (I), membrane capacitance (J), action potential amplitude (K), action potential duration (L) and maximum voltage (M); for all parameters tested no significant difference were found; unpaired t test; n = 6-8. (N-P) Attenuated anorexigenic response to leptin in AraC-treated ob/ob mice at 14 days following icv infusion. AraC- and aCSF-infused ob/ob mice received i.p. leptin, vehicle, or MTII 1 hr before the dark cycle onset and overnight food intake was measured. Food intake is given in grams (N) or relative to prior vehicle treatment (in %, O). (P) Body weights of AraC and vehicle treated ob/ob mice. Red bar indicates infusion period. n = 6-9; **p<0.01, *p<0.05, unpaired t test; ns, not significant. Mean ± SEM.

Figure S6. Female Body Weights and ME-ultrastructure. Related to Figure 5 and 6. (A-C) Changes in body weights of individual AraC- (A) and aCSF-infused (B) female mice. (C) Body weight averages (as % of initial) of mice shown in A and B. Red bar indicates AraC infusion period. Mean ± SEM. (D-F) Shown are representative electron micrographs of the ME region immunogold-labeled with GFP antibodies. Arrowheads, silver-enhanced gold particles; arrows, postsynaptic densities; T, synaptic terminal; A, axonal profile; S, spine-like protrusion; m, mitochondrion. Scale bars, 500nm.

Figure S7. Effect of ME Directed X-irradiation on NG2-glia, Body Weights, and LepR Processes. Related to Figure 7. (A-C) Validation of the beam-targeting approach. Immunodetection of phosphorylated H2AX in the hypothalamus of sham- (left) and x-irradiated (right) mice 1 hr after a single dose of 15 Gray. Images in (B) and (C) represent enlargements of the boxed areas in (A) and (B), respectively. Dashed lines demarcate the radiation beam path. (D, E) X-irradiation led to a significant reduction in ME NG2-glia, when assessed 7 days post treatment initiation. *p<0.05, unpaired t test. n = 3-5, mean ± SEM. (F, G) Body weights (F) and composition (G) of sham- and x-irradiated female mice. Body composition was determined at post-irradiation day 10 (G, top) and 178 (G, bottom). Fat, but not lean, mass of mitotic blocker treated animals was significantly elevated. n = 5-10; *p<0.05, **p<0.01, unpaired t test, mean ± SEM. (H-M) Images of the ME from sham- (H-J) and x-irradiated (K-M) mice were subjected to automated, quantitative object identification using Squassh3C. Identified objects are highlighted with color hues randomly assigned to individual objects (N-S), Mean object intensity (T) and total object surface (U) were calculated based on all objects, or only those above 50 or 100 pixels in size (V). n = 5, student t-test, * = p<0.05, **=p<0.01, ns = non-significant. Mean ± SEM. Movie S1. NG2-glia and LepR Processes Intermingle in the ME. Related to Figure 4. The movie is an animated z stack acquired from a ME section of a LepR-Cre x GFP mouse following immuno-labeling with antibodies against NG2 glia (red) and GFP (green). Z-stack images were acquired by confocal laser microscopy, DAPI nuclear stain is shown in blue. Arrows mark NG2-glia contacting GFP-positive processes.

EXTENDED EXPERIMENTAL PROCEDURES Mice LepR-Cre (LepRb-IRES-Cre) (DeFalco et al., 2001) or NG2-CreER (NG2creERTMBAC) (Zhu et al., 2011) mice were crossed with tdTomato (Ai9-lsl-tdTomato) (Madisen et al., 2010) or GFP (ROSA26-CAGS-tauGFP) (Wen et al., 2011) reporter mice to generate LepR-Cre x tdTomato, LepR-Cre x GFP, or NG2-CreER x tdTomato animals. Sox10-iCreERT2 mice (Simon et al., 2012) were crossed with a CAG-eGFP reporter line (Nakamura et al., 2006) and Esco2fl/fl mice (Whelan et al., 2012) to create Sox10, lox-Esco2fl/fl, GFP mice which carried a Sox10-iCreERT2 transgene, a GFP reporter allele and which were homozygous for the floxed Esco2 allele. C57BL/6J and ob/ob (B6.Cg-Lepob/J) mice were obtained from Jackson Laboratories. Viral Injections For viral injections, LepR-Cre x GFP mice were anesthetized with isoflurane and placed in a stereotaxic frame (Kopf Instruments). 0.3 µl of AAVdj-Ef1α-Flex-ArchT-GFP (Molecular Virology Support Core, Oregon Health and Science University; Addgene plasmid # 58851, (Han et al., 2011)) was bilaterally injected at a flow rate of 0.05 µl/min using a syringe pump (Harvard apparatus) at the following stereotaxic coordinates: anteroposterior -1.5mm, lateral +/- 0.4 mm, dorsoventral -6.1 mm. X-Irradiation Mice were anesthetized with KX cocktail and placed into a stereotaxic apparatus, which was then covered by a 1/8” thick lead shield. To target x-irradiation to the mediobasal hypothalamic region of the mouse brain, the exposed skull was covered by a lead sheet that featured a 2 mm diameter bore. The bore was aligned with a marking on the skull that was placed 1.7 mm posterior to bregma. In this setting, only a small tissue column (2 mm diameter) which includes the ME and mediobasal hypothalamus is exposed to the radiation beam while the mouse brain and body remains otherwise shielded. The mouse positioned in the shielded stereotaxic apparatus was then placed into an x-irradiator (Rad Source 2000 Biological Research Irradiator). Animals were exposed to a total of either 45 Gray or 60 Gray in 3 consecutive sessions on alternating days. Per session, animals were irradiated for 7.6 min or 10.10 min to receive a dose of 15 Gray or 20 Gray respectively. Sham-irradiated controls were identically treated, except for the use of an intact, bore-less lead sheet placed above their skulls to prevent irradiation exposure. Tamoxifen and BrdU Treatment Tamoxifen (Sigma) was dissolved in a 1:9 (v/v) ethanol:sunflower oil mix to a final concentration of 30 mg/ml. NG2-CreER x tdTomato mice were given an i.p. injection twice daily at a dose of 1.2mg tamoxifen 3 hr after lights on and 1.5mg tamoxifen 1 hr before lights off, for 5 consecutive days. Sox10-Cre, Esco2fl/fl and Sox10-Cre, Esco2wt/wt mice were given three times 10 mg tamoxifen in 1:9 (v/v) ethanol:corn oil by gavage every second day (30 mg/mouse in total). For intraperitoneal BrdU (Sigma) delivery, a solution of 6.25 µg/µL BrdU in saline was injected at a dose of 50 mg/kg body weight twice daily, unless otherwise specified. For icv delivery of BrdU in the context of AraC treatment, osmotic minipumps were filled with BrdU (Sigma) in aCSF to provide a dose of 30 µg/day. Immunohistochemistry Brain tissue was fixed and processed for immunohistochemical detection as previously described (Robins et al., 2013). Mice were deeply anesthetized with KX and transcardially perfused with 0.9% NaCl followed by 10% formalin. Brains were removed, postfixed in 10% formalin overnight, and then incubated with a 20% sucrose cryoprotectant solution. 25µm thick coronal sections were cut using a sliding microtome (Leica) and 5 series of sections were collected per

brain. For immunohistochemistry, sections were blocked with a solution of 3% goat serum/0.1% triton X-100 in PBS for 1 hr before being incubated with primary antibody in blocking solution at room temperature overnight. Sections containing BrdU underwent an additional antigen retrieval procedure prior to blocking, consisting of either a 10 minute incubation in a 500 U/ml DNase I (Sigma, cat# DN25) solution or a 1 hr incubation in 2N HCl at room temperature. Sections processed for pSTAT3 immunoreactivity were subjected to antigen retrieval treatment consisting of sequential incubation with 1% NaOH + 1% H2O2 (20 minutes), 0.3% glycine (10 minutes) and 0.06% SDS (10 minutes). Phospho-H2A.X labelling included an antigen retrieval step in which sections were incubated for one hour at 80˚C in 10mM sodium citrate buffer (pH 6.0) prior to blocking with 5% goat serum/0.1% triton X-100 in PBS. Primary antibodies were used at the following dilutions: rat monoclonal anti-BrdU (Accurate, cat# OBT00030) 1:500, rabbit polyclonal anti-GFP (Invitrogen, cat# A6544) 1:10,000, goat anti-GFP (Vector Labs, cat# A21271) 1:1000, mouse anti-phospho-histone H2A.X (Millipore, cat# 05-636) 1:700, rabbit polyclonal anti-Iba-1 (Wako, cat#019-19741) 1:500, rabbit polyclonal anti-NG2 (Millipore, cat# AB5320) 1:250, rabbit polyclonal anti-POMC precursor (Phoenix Pharmaceuticals, cat# M9C6) 1:2500, rabbit polyclonal anti-pSTAT3 (Cell Signalling cat# 9131) 1:2000. Alexa350, Alexa488, Alexa568 and Alexa647 conjugates (Life Technologies) were employed as secondary antibodies for fluorescence detection. Fluorescence Imaging and Data Analysis Epifluorescent images of brain sections were taken using a Zeiss Axio Image D2 microscope equipped with an AxioCam HRm camera. Confocal images were acquired using the visible laser scanning and spectral descan detection portion of a Carl Zeiss LSM 710 NLO system in conjunction with Zen black edition software (Zeiss). Images were acquired using a 20x dry or 63x oil objective (both Plan-Apochromat, Zeiss) at a resolution of 1600x1600 or 1024x1024 pixels, respectively. Z-stacks encompassing the entire thickness of the tissue slice were acquired at a pinhole size of 1 airy unit with a 50% overlap between adjacent optical sections. Three sections or more per animal were used for quantification, representing anterior, middle and posterior ME regions. Z-stacks were utilized to identify and quantify cells throughout the entire ME region, from the internal zone at the floor of the third ventricle and the base of the arcuate nucleus, to the edge of the external or pericapillary zone containing the fenestrated capillaries. For analysis of GFP+ dendrites following x-irradiation, a region of interest (ROI) encompassing the median eminence was selected from 20x confocal images of GFP, excluding the dorsal region, which contains the cell soma of tanycytes and neurons. The ROI was processed using Squassh3C software (Rizk et al., 2014) with a threshold of 0.60 for automated object recognition. Objects with a size greater than 200 pixels were excluded from further (Squassh) analysis, due to concerns that they may represent conglomerations of closely adjacent objects. Data output of the Squassh analysis comprised a list of the automatically identified objects along with their size and fluorescent signal intensities. Statistical differences were calculated using ANOVA. For the quantitative analysis of the immunohistochemical pSTAT3 signal in hypothalamic sections, fluorescence images were processed using a custom macro in Image J. Briefly, ROIs in sections were delineated by identifying objects in the ArcN or VMH above a set threshold; these corresponded to the somas of pSTAT3+ ArcN or VMH cells. Mean intensities of these soma signals were then calculated first for each ROI, and then the mean of all ROIs for each animal was computed. Ultrastructural Analysis Obese, AraC-infused LepR-Cre x GFP animals and normal weight (aCSF-infused) controls were perfused 7 weeks post cannulation surgery. Tissue processing and image acquisition was performed according to previously described methods (Stroh et al., 2006). Mice were deeply anaesthetized using KX and transcardially perfused with 0.9% NaCl followed by 2%

paraformaldehyde (PFA)/2% glutaraldehyde (GA). Brains were removed, post-fixed overnight in 2%PFA/2%GA, and the hypothalamic area was cut into 50µm thick coronal sections using a vibratome (1000 Plus Sectioning System, Vibratome) and further processed as described (Stroh et al., 2006). Sections were immuno-labelled with rabbit anti-GFP antiserum (1:10,000) followed by gold-conjugated goat anti-rabbit IgG (1:50, Aurion ImmunoGold Reagents, Electron Microscopy Sciences) and silver intensification treatment (Aurion R-Gent SE-EM kit, Electron Microscopy Sciences). Ultrathin (90nm), stained sections were imaged by transmission electron microscopy (Tecnai 12 120kV, FEI). Structures containing 3 or more silver grains were considered LepR-GFP positive. Eighteen 4800x images (~100µm2 each) per animal, representing the ME from the ventricular to the pericapillary zone, were subjected to analysis using Stereo Investigator and Neurolucida Explorer contour analysis software (MBF Bioscience). The number, perimeter and area enclosed by the contours were calculated (summed across all 18 images) and given as ratios of immuno-positive degenerating structures over all immuno-positive structures. Data was presented as mean ratios of sums ± standard error. LepR+ dendrites were classified as degenerating if they manifested two or more of the following previously documented structural aberrations attributed to degenerating neurites: an empty, electron-lucent cytoplasm devoid of organelles (Jaworski et al., 2011; Kaur et al., 2011), vacuolization (Jaworski et al., 2011; Kaur et al., 2011), membranous fragments, disintegrated organelles/cytoskeletal elements, dark debris (Adori et al., 2011; Ikonomidou et al., 1996; Jaworski et al., 2011), abnormal mitochondria with swollen and distended cristae, holes, discontinuous membranes, or an electron-dense burst morphology (Ikonomidou et al., 1996; Siskova et al., 2010). Western Blot Analysis Overnight fasted mice were injected with leptin in PBS (7.5µg/g body weight) or PBS alone (vehicle) 30 min prior to sacrifice. 500µm coronal brain sections containing the mediobasal hypothalamus were cut on a vibratome (Leica). The ArcN or VMH/DMH regions were then microdissected according to landmarks including the third ventricle, the optical fiber tracks, median eminence and the fornix. Dissected tissues were homogenized in lysis buffer consisting of 1x lysis buffer (Cell Signaling), inhibitor 2, inhibitor 3, PMSF and Aprotinin (all Sigma). Immuno-blotting was performed as previously described (Liu et al., 2014). Briefly, 25µg of protein extract was loaded onto 10% SDS-polyacrylamide gels, subjected to electrophoresis, and then transferred onto nitrocellulose membranes. Immuno-signal of STAT3 (1:10,000, cat#AB1542, Santa Cruz) and pSTAT3 (1:2000, cat#9131, New England Biolabs) antibodies were visualized using a FluorChem 8800 Imager (Alpha Innotech). Electrophysiology Recordings AraC-infused (lateral ventricle) LepR-Cre x tdTomato mice were anaesthetized two weeks after pump implantation with isoflurane and then decapitated. Brains were swiftly removed and submerged in cold aCSF solution saturated with a 95% O2/5% CO2 carbogen mix. Brains were cut into 300 µm coronal sections on a vibratome (Leica), and maintained in carbogenated aCSF at 32°C. Patch micropipettes were prepared by pulling glass capillary tubes (1.2mm o.d., A-M Systems Inc.) on a micropipette puller (P-87; Sutter Instrument, Novato, CA). Micropipettes were filled with a solution containing 128mM potassium gluconate, 10mM KCl, 1mM MgCl2, 10mM HEPES, 1mM EGTA, 0.3mM CaCl, 3mM MgATP and 0.3mM NaGTP (pH adjusted to 7.30 with NaOH). Pipette resistance in the bath was 3 to 7MΩ. LepR+ neurons of the ArcN or VMH regions were identified by tdTomato fluorescence, patch clamped, and the action potential firing frequency and membrane voltage were monitored under current clamp conditions. For experiments aimed at determining leptin responsiveness, after a stable baseline recording was achieved the bath solution was supplemented with leptin at a final concentration of 100nM for

four minutes. As the leptin effect on the electrical properties of patched LepR+ neurons did not diminish upon removal of leptin, a fresh slice was used for each recording. Whole cell current and voltage (d.c. 2KHz) was recorded using a Multiclamp amplifier (Axon 700B, Molecular Devices) attached to a Digidata 1550 interface (Molecular Devices) and digitized using pClamp 10 (Molecular Devices). Signals were analyzed offline using Clampfit 10 software. We classified neurons displaying changes to the baseline RMP of greater than 5mV following leptin administration as being leptin responsive. Body Composition, Food Intake, and Locomotor Activity Measurements Body composition of adult male and female mice was assessed using the EchoMRI system (Echo Medical Systems). Metabolic chambers (TSE systems) were used for assessing ambulatory activity and daily food intake of singly housed mice. Mice were adapted to the chamber environment for 5 days prior to data collection. Food and water was available ad libitum. Metabolic chamber monitoring was conducted for 9 consecutive days. Circadian Locomotor Activity Running wheel activity was recorded as described elsewhere (Blum et al., 2014). Actograms display wheel revolutions per 6 min interval. Animals were individually housed in light-tight cabinets and activity was recorded using ClockLab (Actimetrics, Wilmette, IL). χ2-periodgrams (Sokolove and Bushell, 1978) computed from the activity data (ClockLab, confidence level = 0.001) were used to determine period, defined by the highest (dominant) periodogram peak. Circadian amplitude was retrieved from the periodogram defined as power above the calculated threshold at the dominant period. SUPPLEMENTAL REFERENCES

Adori, C., Low, P., Ando, R.D., Gutknecht, L., Pap, D., Truszka, F., Takacs, J., Kovacs, G.G., Lesch, K.P., and Bagdy, G. (2011). Ultrastructural characterization of tryptophan hydroxylase 2-specific cortical serotonergic fibers and dorsal raphe neuronal cell bodies after MDMA treatment in rat. Psychopharmacology (Berl.) 213, 377-391.

Blum, I.D., Zhu, L., Moquin, L., Kokoeva, M.V., Gratton, A., Giros, B., and Storch, K.F. (2014). A highly tunable dopaminergic oscillator generates ultradian rhythms of behavioral arousal. Elife 3.

DeFalco, J., Tomishima, M., Liu, H., Zhao, C., Cai, X., Marth, J.D., Enquist, L., and Friedman, J.M. (2001). Virus-assisted mapping of neural inputs to a feeding center in the hypothalamus. Science 291, 2608-2613.

Han, X., Chow, B.Y., Zhou, H., Klapoetke, N.C., Chuong, A., Rajimehr, R., Yang, A., Baratta, M.V., Winkle, J., Desimone, R., et al. (2011). A high-light sensitivity optical neural silencer: development and application to optogenetic control of non-human primate cortex. Front. Syst. Neurosci. 5, 18.

Ikonomidou, C., Qin Qin, Y., Labruyere, J., and Olney, J.W. (1996). Motor neuron degeneration induced by excitotoxin agonists has features in common with those seen in the SOD-1 transgenic mouse model of amyotrophic lateral sclerosis. J. Neuropathol. Exp. Neurol. 55, 211-224.

Jaworski, T., Lechat, B., Demedts, D., Gielis, L., Devijver, H., Borghgraef, P., Duimel, H., Verheyen, F., Kugler, S., and Van Leuven, F. (2011). Dendritic degeneration, neurovascular defects, and inflammation precede neuronal loss in a mouse model for tau-mediated neurodegeneration. Am. J. Pathol. 179, 2001-2015.

Kaur, C., Viswanathan, S., and Ling, E.A. (2011). Hypoxia-induced cellular and vascular changes in the nucleus tractus solitarius and ventrolateral medulla. J. Neuropathol. Exp. Neurol. 70, 201-217.

Liu, L., Chowdhury, S., Fang, X., Liu, J.L., and Srikant, C.B. (2014). Attenuation of unfolded protein response and apoptosis by mReg2 induced GRP78 in mouse insulinoma cells. FEBS Lett. 588, 2016-2024.

Madisen, L., Zwingman, T.A., Sunkin, S.M., Oh, S.W., Zariwala, H.A., Gu, H., Ng, L.L., Palmiter, R.D., Hawrylycz, M.J., Jones, A.R., et al. (2010). A robust and high-throughput Cre reporting and characterization system for the whole mouse brain. Nat. Neurosci. 13, 133-140.

Nakamura, T., Colbert, M.C., and Robbins, J. (2006). Neural crest cells retain multipotential characteristics in the developing valves and label the cardiac conduction system. Circ. Res. 98, 1547-1554.

Rizk, A., Paul, G., Incardona, P., Bugarski, M., Mansouri, M., Niemann, A., Ziegler, U., Berger, P., and Sbalzarini, I.F. (2014). Segmentation and quantification of subcellular structures in fluorescence microscopy images using Squassh. Nat. Protoc. 9, 586-596.

Robins, S.C., Villemain, A., Liu, X., Djogo, T., Kryzskaya, D., Storch, K.F., and Kokoeva, M.V. (2013). Extensive regenerative plasticity among adult NG2-glia populations is exclusively based on self-renewal. Glia 61, 1735-1747.

Simon, C., Lickert, H., Gotz, M., and Dimou, L. (2012). Sox10-iCreERT2 : a mouse line to inducibly trace the neural crest and oligodendrocyte lineage. Genesis 50, 506-515.

Siskova, Z., Mahad, D.J., Pudney, C., Campbell, G., Cadogan, M., Asuni, A., O'Connor, V., and Perry, V.H. (2010). Morphological and functional abnormalities in mitochondria associated with synaptic degeneration in prion disease. Am. J. Pathol. 177, 1411-1421.

Sokolove, P.G., and Bushell, W.N. (1978). The chi square periodogram: its utility for analysis of circadian rhythms. J. Theor. Biol. 72, 131-160.

Stroh, T., Sarret, P., Tannenbaum, G.S., and Beaudet, A. (2006). Immunohistochemical distribution and subcellular localization of the somatostatin receptor subtype 1 (sst1) in the rat hypothalamus. Neurochem. Res. 31, 247-257.

Wen, S., Gotze, I.N., Mai, O., Schauer, C., Leinders-Zufall, T., and Boehm, U. (2011). Genetic identification of GnRH receptor neurons: a new model for studying neural circuits underlying reproductive physiology in the mouse brain. Endocrinology 152, 1515-1526.

Whelan, G., Kreidl, E., Wutz, G., Egner, A., Peters, J.M., and Eichele, G. (2012). Cohesin acetyltransferase Esco2 is a cell viability factor and is required for cohesion in pericentric heterochromatin. EMBO J. 31, 71-82.

Zhu, X., Hill, R.A., Dietrich, D., Komitova, M., Suzuki, R., and Nishiyama, A. (2011). Age-dependent fate and lineage restriction of single NG2 cells. Development 138, 745-753.