What are the contributions in this paper?

Q: What are the contributions in this paper?

In this paper, the adjusted error rates ( a9 ) were P = 0.004 and 0.01, respectively, for the 13 subsequent planned pairwise comparisons in Fig. 4a and for the 5 subsequent planned comparisons inFig. 4b.

(Open Access) Leptin activates anorexigenic POMC neurons through a neural network in the arcuate nucleus (2001) | Michael A. Cowley

letters to nature

480 NATURE

VOL 411

24 MAY 2001

www.nature.com

adjusted error rate (

a9) for the six subsequent planned pairwise comparisons was 0.009.

n = 8 PIs for each group.

PIs are distributed normally. Hence, these data were analysed by analysis of variance

with subsequent pairwise comparisons adjusted to maintain an experiment-wise error rate

a = 0.05. In Fig. 3, the error rate for the one subsequent planned comparison was 0.05.

In Fig. 4, to maintain an experiment-wise error rate of a = 0.05, the adjusted error rates

(a9) were P = 0.004 and 0.01, respectively, for the 13 subsequent planned pairwise

comparisons in Fig. 4a and for the 5 subsequent planned comparisons in Fig. 4b. The

number of PIs included in each group mean is listed above the corresponding bar in all

®gures.

Received 25 January; accepted 28 March 2001.

1. Scoville, W. B. & Milner, B. Loss of recent memory after bilateral hippocampal lesions.

J. Neuropsychiat. Clin. Neurosci. 12, 103±113 (2000).

2. Milner, B., Squire, L. R. & Kandel, E. R. Cognitive neuroscience and the study of memory. Neuron 20,

445±468 (1998).

3. Mayford, M. et al. Control of memory formation through regulated expression of a CaMKII

transgene. Science 274, 1678±1683 (1996).

4. Shimizu, E., Tang, Y. P., Rampon, C. & Tsien, J. Z. NMDA receptor-dependent synaptic reinforcement

as a crucial process for memory consolidation. Science 290, 1170±1174 (2000).

5. de Belle, J. S. & Heisenberg, M. Associative odor learning in Drosophila abolished by chemical ablation

of mushroom bodies. Science 263, 692±695 (1994).

6. Hammer, M. & Menzel, R. Multiple sites of associative odor learning as revealed by local brain

microinjections of octopamine in honeybees. Learn. Mem. 5, 146±156 (1998).

7. Kitamoto, T. Conditional modi®cation of behavior in Drosophila by targeted expression of a

temperature-sensitive shibire allele in de®ned neurons. J. Neurobiol. 47, 81±92 (2001).

8. Rybak, J. & Menzel, R. Anatomy of the mushroom bodies in the honey bee brain: the neuronal

connections of the a-lobe. J. Comp. Neurol. 334, 444±465 (1993).

9. Strausfeld, N. J. & Li, Y. Organization of olfactory and multimodal afferent neurons supplying the

calyx and pedunculus of the cockroach mushroom bodies. J. Comp. Neurol. 409, 603±625 (1999).

10. Ito, K. et al. The organization of extrinsic neurons and their implications in the functional roles of the

mushroom bodies in Drosophila melanogaster Meigen. Learn. Mem. 5, 52±77 (1998).

11. Connolly, J. B. et al. Associative learning disrupted by impaired Gs signaling in Drosophila mushroom

bodies. Science 274, 2104±2106 (1996).

12. Zars, T., Fischer, M., Schulz, R. & Heisenberg, M. Localization of a short-term memory in Drosophila.

Science 288, 672±675 (2000).

13. Poodry, C. A., Hall, L. & Suzuki, D. T. Developmental properties of Shibire: a pleiotropic mutation

affecting larval and adult locomotion and development. Dev. Biol. 32, 373±386 (1973).

14. Poodry, C. A. & Edgar, L. Reversible alteration in the neuromuscular junctions of Drosophila

melanogaster bearing a temperature-sensitive mutation, shibire. J. Cell Biol. 81, 520±527 (1979).

15. Ramaswami, M., Rao, S., van der Bliek, A., Kelly, R. B. & Krishnan, K. S. Genetic studies on dynamin

function in Drosophila. J. Neurogenet. 9, 73±87 (1993).

16. Koenig, J. H. & Ikeda, K. Disappearance and reformation of synaptic vesicle membrane upon

transmitter release observed under reversible blockage of membrane retrieval. J. Neurosci. 9, 3844±

3860 (1989).

17. Chen, M. S. et al. Multiple forms of dynamin are encoded by shibire,aDrosophila gene involved in

endocytosis. Nature 351, 583±586 (1991).

18. van der Bliek, A. M. & Meyerowitz, E. M. Dynamin-like protein encoded by the Drosophila shibire gene

associated with vesicular traf®c. Nature 351, 411±414 (1991).

19. Damke, H., Baba, T., Warnock, D. E. & Schmid, S. L. Induction of mutant dynamin speci®cally blocks

endocytic coated vesicle formation. J. Cell Biol. 127, 915±934 (1994).

20. Grant, D., Unadkat, S., Katzen, A., Krishnan, K. S. & Ramaswami, M. Probable mechanisms

underlying interallelic complementation and temperature-sensitivity of mutations at the shibire locus

of Drosophila melanogaster. Genetics 149, 1019±1030 (1998).

21. Brand, A. H. & Perrimon, N. Targeted gene expression as a means of altering cell fates and generating

dominant phenotypes. Development 118, 401±415 (1993).

22. Yang, M. Y., Armstrong, J. D., Vilinsky, I., Strausfeld, N. J. & Kaiser, K. Subdivision of the Drosophila

mushroom bodies by enhancer-trap expression patterns. Neuron 15, 45±54 (1995).

23. Joiner, M. A. & Grif®th, L. C. Mapping of the anatomical circuit of CaM kinase-dependent courtship

conditioning in Drosophila. Learn. Mem. 6, 177±192 (1999).

24. Benzer, S. Behavioral mutants of Drosophila isolated by countercurrent distribution. Proc. Natl Acad.

Sci. USA 58, 1112±1119 (1967).

25. Tully, T., Preat, T., Boynton, S. C. & Del Vecchio, M. Genetic dissection of consolidated memory in

Drosophila. Cell 79, 35±47 (1994).

26. Wang, Y. T. & Linden, D. J. Expression of cerebellar long-term depression requires postsynaptic

clathrin-mediated endocytosis. Neuron 25, 635±647 (2000).

27. Carroll, R. C. et al. Dynamin-dependent endocytosis of ionotropic glutamate receptors. Proc. Natl

Acad. Sci. USA 96, 14112±14117 (1999).

28. Dudai, Y., Corfas, G. & Hazvi, S. What is the possible contribution of Ca

-stimulated adenylate

cyclase to acquisition, consolidation and retention of an associative olfactory memory in Drosophila?

J. Comp. Physiol. A 162, 101±109 (1988).

29. Waddell, S., Armstrong, J. D., Kitamoto, T., Kaiser, K. & Quinn, W. G. The amnesiac gene product is

expressed in two neurons in the Drosophila brain that are critical for memory. Cell 103, 805±813 (2000).

30. Dubnau, J. & Tully, T. Functional anatomy: from molecule to memory. Curr. Biol.11, R240±R243 (2001).

Supplementary information is available on Nature's World-Wide Web site

(http://www.nature.com) or as paper copy from the London editorial of®ce of Nature.

Acknowledgements

We thank H. Cline, E. Drier, J. Huang, M. Regulski and K. Svoboda for comments on the

manuscript. This work was supported by the NIH (J.T.D. and T.T.) and the NSF (T.K.).

Correspondence and requests for materials should be addressed to J.D.

(e-mail: dubnau@cshl.org).

.................................................................

Leptin activates anorexigenic POMC

neurons through a neural network

in the arcuate nucleus

Michael A. Cowley*

, James L. Smart*

, Marcelo Rubinstein

Marcelo G. Cerda

, Sabrina Diano§, Tamas L. Horvath§k, Roger D. Cone*

& Malcolm J. Low*

* The Vollum Institute, Oregon Health Sciences University, Portland,

Oregon 97201-3098, USA

Instituto de Investigaciones en Ingenierõ

a Gene

tica y Biologõ

a Molecular,

CONICET and Department of Biology, School of Sciences,

University of Buenos Aires 1428, Argentina

§ Reproductive Neurosciences Unit, Department of Obstetrics and Gynecology and

k Department of Neurology, Yale Medical School, New Haven, Connecticut 06520,

USA

These authors contributed equally to this work

..............................................................................................................................................

The administration of leptin

to leptin-de®cient humans, and the

analogous Lep

/Lep

mice, effectively reduces hyperphagia and

obesity

2,3

. But common obesity is associated with elevated leptin,

which suggests that obese humans are resistant to this adipocyte

hormone. In addition to regulating long-term energy balance,

leptin also rapidly affects neuronal activity

4±6

. Proopiomelano-

cortin (POMC) and neuropeptide-Y types of neurons in the

arcuate nucleus of the hypothalamus

are both principal sites of

leptin receptor expression and the source of potent neuropeptide

modulators, melanocortins and neuropeptide Y, which exert

opposing effects on feeding and metabolism

8,9

. These neurons

are therefore ideal for characterizing leptin action and the

mechanism of leptin resistance; however, their diffuse distribu-

tion makes them dif®cult to study. Here we report electrophysio-

logical recordings on POMC neurons, which we identi®ed by

targeted expression of green ¯uorescent protein in transgenic

mice. Leptin increases the frequency of action potentials in the

anorexigenic POMC neurons by two mechanisms: depolarization

through a nonspeci®c cation channel; and reduced inhibition by

local orexigenic neuropeptide-Y/GABA (g-aminobutyric acid)

neurons. Furthermore, we show that melanocortin peptides

have an autoinhibitory effect on this circuit. On the basis of our

results, we propose an integrated model of leptin action and

neuronal architecture in the arcuate nucleus of the hypothalamus.

Previous studies suggest that leptin does not have equal effects on

all neuronal subtypes. Acute leptin treatment presumably activates

POMC, but not neuropeptide Y (NPY) neurons in the arcuate

nucleus of the hypothalamus (ARC), because c-Fos protein is

increased only in the former population and Socs3 messenger

RNA is increased in both

. Furthermore, a population of ARC

neurons seems to be inhibited directly by leptin, but the peptide

phenotype of these neurons has not been directly established

11±14

.In

addition to leptin receptors, both POMC and NPY neurons express

a receptor, MC3-R, for POMC-derived melanocortin peptides

The physiological role of this receptor is not well understood,

although MC3-R null mice have increased adiposity compared

with wild-type mice

To test the hypothesis that leptin selectively activates POMC

neurons, we ®rst generated a strain of transgenic mice expressing

green ¯uorescent protein (EGFP; Clontech) under the transcrip-

tional control of mouse Pomc genomic sequences, including a

region located between -13 kilobases (kb) and -2 kb that is required

for accurate neuronal expression

(Fig. 1a). Bright green ¯uores-

cence (509 nm) was seen in the two central nervous system regions

where POMC is produced: the ARC and the nucleus of the solitary

tract (data not shown).

letters to nature

NATURE

VOL 411

24 MAY 2001

www.nature.com 481

Under ultraviolet (450±480 nm) excitation, POMC neurons were

clearly distinguished from adjacent, non-¯uorescent neurons (Fig.

1b) visualized under infrared optics. Double immuno¯uorescence

showed that there was more than 99% cellular colocalization of

EGFP and POMC peptides in the ARC (Fig. 1c). There was close

apposition of both tyrosine-hydroxylase- and NPY-stained termi-

nals on EGFP-expressing POMC neurons, but no evidence of

colocalization of the tyrosine hydroxylase or NPY immunoreactiv-

ity with EGFP (data not shown). Total ¯uorescent cell counts

of coronal hypothalamic sections determined that there were

3,148 6 62 (mean 6 s.e.m.; n = 3) POMC-EGFP neurons distrib-

uted through the whole ARC

(Fig. 1d). POMC neurons are located

both medially and ventrally in mouse ARC, in contrast to a

predominantly lateral position in rat ARC.

POMC-EGFP neurons in hypothalamic slices had a resting

membrane potential of -40 to -45 mV and exhibited frequent

spontaneous action potentials. The non-selective opioid agonist

met-enkephalin (Met-Enk, 30 mM; Sigma) caused a rapid (35±

40 s), reversible hyperpolarization (10±20 mV) of the membrane

potential of POMC cells (n = 10) and prevented spontaneous action

potential generation (Fig. 2a). In normal (2.5 mM K

) Krebs

solution, the reversal potential of the inwardly rectifying opioid

current was about -90 mV, whereas in 6.5 mM K

Krebs solution the

reversal potential shifted to about -60 mV (n = 3; Fig. 2b). The m-

opioid receptor (MOP-R) antagonist CTAP (1 mM; Phoenix Phar-

maceuticals) completely prevented the current induced by Met-Enk

in POMC cells (n = 3; Fig. 2c). These characteristics indicate that the

opioid current was caused by both activation of MOP-R and

increased ion conductance through G-protein-coupled, inwardly

rectifying potassium channels

. The similarity of opioid responses

in EGFP-labelled POMC neurons to those of guinea-pig

–13 –2 0 9

EcoRI EcoRINotl

Length (kb)

transcriptional start

translational stop

EGFP

123

Bregma –1.06 mm –1.22 mm

–1.58 mm

–1.34 mm

–1.70 mm

–1.46 mm

–1.82 mm –1.94 mm

Bregma –2.06 mm –2.18 mm –2.46 mm

–2.54 mm –2.80 mm–2.70 mm –2.92 mm

IR-DIC

Ultraviolet

Colocalization

β-endorphin

EGFP

2468

SOX

PaPo

PeP

Arc

Gen

A12

Arc

VMH

SubI

VMH

PMD

PMV

VMH

AHP

VMH

PMV

DTM

SMT

SuMM

PMD

PMV

SuM

SuMM

sumx

VTM

MRe

SMT

MRe

–2.30 mm

Figure 1 Generation of transgenic mice expressing EGFP in ARC POMC neurons.

a, Structure of the POMC-EGFP transgene. b, Identi®cation of a single POMC neuron (red

arrowhead on recording electrode tip) by infrared-differential interference contrast

(IR-DIC) microscopy (upper) and EGFP ¯uorescence (lower) in a living ARC slice before

electrophysiological recordings. c, Colocalization (yellow) of EGFP (green) and b-

endorphin immunoreactivity (red) in ARC POMC neurons. Scale bars, 50 mm(b, c).

d, Distribution of EGFP-positive neuronal soma throughout the ARC nucleus. Green open

circles = 5 cells, green ®lled circles = 10 cells.

2,450

–10

–20

–30

–40

–50

–60

–70

–80

2,200 2,250 2,300 2,350 2,400

Time (s)

Potential (mV)

Met-Enk (30 µM)

–30

–120 –110 –100 –90

–80 –70 –60 –50 –40

–30

–20

–10

Met-Enk

Met-Enk CTAP

Voltage (mV)

–130

–130 –110

–90

–70

–50

–75

–50

–25

2.5 mM [K

]

6.5 mM [K

]

Voltage (mV)

Net current (pA)

–150

Net current (pA)

Figure 2 Activation of MOP-Rs hyperpolarizes the EGFP-labelled POMC neurons by

opening G-protein-coupled inwardly rectifying potassium channels. a, Met-enkephalin

(Met-Enk) hyperpolarizes POMC neurons and inhibits all action potentials. Horizontal bar

indicates the time when 30 mM Met-Enk was bath applied to the slice. b, Met-Enk current

and reversal potential is shifted by extracellular K

concentration. c, Met-Enk activates

MOP-Rs on POMC neurons. A Met-Enk (30 mM) current was observed, and the MOP-R-

speci®c antagonist CTAP (1 mM) was applied for 1 min. After CTAP, Met-Enk elicited no

current. Figure is representative of three experiments.

letters to nature

482 NATURE

VOL 411

24 MAY 2001

www.nature.com

mouse

POMC cells, identi®ed by post-recording immunohisto-

chemistry, suggests that expression of the EGFP transgene does not

compromise expression of receptors nor their coupling to second

messenger systems in POMC neurons.

We next determined the direct effects of leptin on identi®ed

POMC cells in slice preparations. Leptin (0.1±100 nM) depolarized

72 out of 77 POMC cells by 3±30 mV (Fig. 3a; depolarization at

100 nM leptin, 9.7 6 1.2 mV (mean 6 s.e.m.; n = 45)) in 2±10 min,

in a concentration-responsive manner (Fig. 3b). There were two

components to the depolarization and neither were fully reversible

within 40 min. First, the depolarization was due to a small inward

current that reversed at about -20 mV (Fig. 3c), suggesting the

involvement of a nonspeci®c cation channel

. Second, leptin

treatment decreased the GABA-mediated tone onto POMC cells.

GABA-mediated inhibitory postsynaptic currents (IPSCs) were

observed in POMC cells, and leptin (100 nM) decreased their

frequency by 25% (Fig. 3d) in 5 out of 15 cells, indicating that it

may act presynaptically to reduce GABA release (leptin had no effect

on IPSCs in 10 out of 15 POMC neurons).

The effect on IPSC frequency occurred with a similar lag to the

effect on membrane potential. Thus, leptin not only directly

depolarizes POMC neurons but also acts at GABA-secreting nerve

terminals to reduce the release of GABA onto POMC neurons,

allowing them to adopt a more depolarized resting potential. The

consistent depolarization of POMC cells by leptin was speci®c

because leptin had no effect on 5 out of 13 adjacent non-¯uorescent

cells tested (Fig. 3e), whereas it hyperpolarized ®ve (Fig. 3f) and

depolarized three other non-POMC neurons in the ARC (data

not shown). The electrophysiological effects of leptin reported

here are consistent with leptin's biological actionsÐleptin rapidly

causes release of a-melanocyte-stimulating hormone (a-MSH)

from rat hypothalami

, presumably by activating POMC neurons.

Previous reports of neuronal hyperpolarization by leptin

11,12

and

the colocalization of GABA and NPY

in subpopulations of ARC

neurons led us to speculate that leptin hyperpolarizes NPY/GABA

cells that directly innervate POMC neurons, and thus reduces

GABA-mediated drive onto POMC cells. Both leptin and NPY Y2

receptors are expressed on NPY neurons in the ARC

7,23

. Further-

more, activation of Y2 receptors inhibits NPY release from NPY

neurons

, and presumably would also diminish GABA release from

NPY/GABA terminals. This provided us with an alternative phar-

macological approach, independent of leptin, to test the proposed

innervation of POMC neurons by GABA-secreting NPY neurons.

Indeed, NPY (100 nM; Bachem) decreased the frequency of

GABA-mediated IPSCs by 55% in 3 min in all 12 POMC cells

tested (Fig. 4a). Both NPY and leptin still inhibited IPSCs in the

presence of tetrodotoxin (TTX; 6/6 and 3/5 cells, respectively),

indicating that some of the inhibition of IPSCs was occurring

through direct effects at presynaptic nerve terminals. POMC neu-

rons express the NPY Y1 receptor

, and NPYalso hyperpolarized all

POMC neurons tested by an average of 9 6 6mV(n = 3; data not

shown).

To con®rm the origin of GABA-mediated innervation on POMC

neurons from NPY/GABA terminals, we also tested the effect of the

highly selective MC3-R agonist

D-Trp

-gMSH (ref. 25) on local

GABA release.

D-Trp

-gMSH (7 nM) increased the frequency of

GABA-mediated IPSCs (280 6 90%) recorded from three out of

four POMC neurons (Fig. 4b). It had no effect on one cell. The

positive effect of MC3-R activation, together with the negative

effects of NPY and leptin, shows the dynamic range of the NPY/

GABA synapse onto POMC neurons and points to the important

role of this synapse in modulating signal ¯ow in the ARC.

D-Trp

gMSH (7 nM) also hyperpolarized (-5.5 6 2.4 mV) 9 out of 15

POMC neurons tested, and decreased the frequency of action

potentials (Fig. 4c); the remaining cells showed no signi®cant

response to

D-Trp

-gMSH. We could not determine whether these

effects were entirely due to increased GABA release onto the POMC

cells, or whether there was an additional postsynaptic action of

D-

Trp

-gMSH on POMC neurons, roughly half of which also express

the MC3-R

. Thus, MC3-R acts in a similar autoreceptor manner to

MOP-Rs on POMC neurons, diminishing POMC neuronal activity

in response to elevated POMC peptides.

To determine that the IPSCs in POMC neurons were due to local

innervation by NPY/GABA cells, we carried out multi-label immuno-

–25

–50

–75

–100

500 550

Time (s)

Leptin (100 nM)

600

–100 –80 –60 –40 -20

–60

–50

–40

–30

–20

–10

Voltage (mV)

Net current (pA)

log [Leptin (M)]

–11

–10 –9 –8 –7

12.5

10.0

7.5

5.0

2.5

–2.5

= 5.9 nM

Depolarization (mV)

Leptin (100 nM)

020

1.5

1.0

0.5

Time (min)

–25

–50

–75

–100

500 550

Time (s)

Potential (mV)

Leptin (100 nM)

600

IPSC frequency (Hz)

–40

–60

–80

–100

600 700

Time (s)

Potential (mV)

Leptin (100 nM)

800

–20

Figure 3 Leptin depolarizes POMC neurons through a nonspeci®c cation channel, and

decreases the GABA-mediated tone onto POMC cells. a, Leptin depolarizes POMC

neurons and increases the frequency of action potentials within 1±10 min of addition.

Figure is representative of recordings made from 77 POMC neurons. b, Leptin causes a

concentration-dependent depolarization of POMC cells. Depolarization by leptin was

determined at 0.1, 1, 10, 50 and 100 nM (effector concentration at half-maximum

response, EC

= 5.9 nM) in 8, 7, 9, 3 and 45 cells, respectively. c, Leptin depolarizes

POMC cells by activating a nonspeci®c cation current. Figure is representative of the

response in ten cells. d, Leptin decreases the frequency of IPSCs in POMC cells. Figure is

an example of ®ve cells in which leptin (100 nM) decreased the frequency of IPSCs.

e, Leptin had no effect on ®ve adjacent non-¯uorescent ARC neurons. f, Leptin

hyperpolarized ®ve non-¯uorescent ARC neurons.

letters to nature

NATURE

VOL 411

24 MAY 2001

www.nature.com 483

histochemistry using light and electron microscopy. Although

independent NPY

and GABA

innervation of POMC cells has

been reported, colocalization of NPY and GABA in nerve terminals

forming synapses onto POMC cells has not been shown. Similar to

the rat

, a dense innervation of POMC cells by NPYaxon terminals

was detected in the mouse (Fig. 4d). Electron microscopy con®rmed

the co-expression of NPYand GABA in axon terminals, and showed

that these boutons established synapses on the perikarya of all 15

ARC POMC neurons analysed (Fig. 4e).

These observations allow us to propose a detailed model of

regulation of this circuit including: dual mechanisms of leptin

action in the ARC, interactions between NPY/GABA and POMC

neurons, and autoregulatory feedback from opioid and melano-

cortin peptides as well as NPY (Fig. 4f). In this model, leptin directly

depolarizes the POMC neurons while simultaneously hyper-

polarizing the somata of NPY/GABA neurons, and diminishes

release from NPY/GABA terminals. This diminished GABA release

disinhibits the POMC neurons. Together, the direct and indirect

effects of leptin result in an activation of POMC neurons and an

increased frequency of action potentials. In addition, both POMC

and NPY neurons express autoreceptors for some of their respective

neuropeptide products (b-endorphin or a-MSH, and NPY, respec-

tively) and activation of these autoreceptors may provide ultrashort

feedback loops that further modulate the effects of leptin on POMC

neurons. The effects of NPY and melanocortin agonists on IPSC

frequency are similar to the effect of NPY on IPSCs in the

paraventricular nucleus of the hypothalamus

Thus, there seem to be two classes of neurons accounting for

leptin sensitivity in the brain: those activated (depolarized) to

release anorexigenic peptides; and those inhibited (hyperpolarized)

with a consequent reduction in release of orexigenic peptides.

Identifying viable POMC neurons with EGFP provides an invalu-

able tool for determining complex actions on these important

neurons. Furthermore, an understanding of the acute effects of

leptin provides a model system for studying the mechanisms

underlying the development of leptin resistance

. M

Methods

Generation of POMC-EGFP mice

The EGFP cassette contains its own Kozak consensus translation initiation site, along with

SV40 polyadenylation signals downstream of the EGFP coding sequences, which directs

proper processing of the 39 end of the EGFP mRNA. We introduced the EGFP cassette by

standard techniques into the 59 untranslated region of exon 2 of a mouse Pomc genomic

clone containing 13 kb of 59 and 2 kb of 39 ¯anking sequences

. The transgene was

microinjected into pronuclei of one-cell-stage embryos of C57BL/6J mice (Jackson

Laboratories) as described

. One founder was generated and bred to wild-type C57BL/6J

to produce N

hemizygous mice. In addition, we also generated N

and subsequent

generations of mice homozygous for the transgene. The mice are fertile and have normal

growth and development.

Immuno¯uorescence and GFP colocalization

Anaesthetized mice were perfused transcardially with 4% paraformaldehyde, and free-

¯oating brain sections were prepared with a vibratome. We processed sections for

immuno¯uoresence and colocalization of GFP ¯uoresence using standard techniques.

Primary antisera and their ®nal dilutions were rabbit anti-b-endorphin, 1:2500 (v/v);

rabbit anti-NPY, 1:25,000 (v/v) (Alanex Corp.); rabbit anti-ACTH, 1:2000 (v/v); and

mouse anti-TH, 1:1000 (v/v) (Incstar). After rinsing, sections were incubated with

10 mg ml

-1

biotinylated horse anti-mouse/rabbit immunoglobulin-g (IgG) (Vector

Laboratories) followed by Cy-3 conjugated streptavidin, 1:500 (v/v) (Jackson Immuno-

research Laboratories). Photomicrographs were taken on an Axioscop (Zeiss) using

¯uoroscein isothiocyanate (FITC) and rhodamine isothiocyanate (RITC) ®lter sets

(Chroma Technology Corp.).

Electrophysiology

We cut 200-mm thick coronal slices from the ARC of 4-week-old male POMC-EGFP mice.

Slices were maintained at 35 8C in Krebs solution ((in mM): 126 NaCl, 2.5 KCl, 1.2 MgCl

2.4 CaCl

.2H

O, 1.2 NaH

O, 21.4 NaHCO

, 11.1 glucose), and were saturated with

95% O

5% CO

for 1 h before recordings. Recordings were made in Krebs solution at

35 8C. Slices were visualized on an Axioskop FS2 (Zeiss) through standard infrared optics

and using epi¯uoresence through a FITC ®lter set (see Fig. 1c). Whole-cell recordings were

made from ¯uorescent neurons using an Axopatch 1D ampli®er (Axon Instruments) and

Clampex 7 (Axon Instruments).

We determined resting membrane potentials using an event-detection protocol on a

PowerLab system (AD Instruments) to average expanded traces of the membrane

potential. Drugs were applied to the bath over the times indicated. The resting membrane

potential was stable for up to 1 h in cells treated with Krebs solution alone (data not

shown). I±V relationships for the Met-Enk currents were established using a step protocol

(-60 mV holding potential, sequentially pulsed (40 ms) from -120 to -50 mV; cells were

returned to -60 mV for 2 s between voltage steps). The protocol was repeated after

addition of Met-Enk. The net current was the difference between the two I±V relation-

51015

D-Tr p

MSH (7 nM)

Time (min)

IPSC frequency (Hz)

Time (min)

NPY (100 nM)

0123456

1.5

1.0

0.5

IPSC frequency (Hz)

–25

–50

–75

–100

100

Time (s)

Potential (mV)

200 300 400

D-Tr p

MSH (7 nM)

Target site

NPY/AGRP

POMC

GABA

POMC

NPY

Y2-R

Y1-R

MC3-R

GABA

MC3-R

MOP-R

MC3-R

Lep-R

Figure 4 The GABA-mediated inputs to POMC cells are from NPY neurons that co-express

GABA. a, NPY decreases the frequency of mini-IPSCs in POMC neurons. b,

D-Trp

gMSH

(7nM)Ða dose that selectively activates MC3-RÐincreases the frequency of GABA-

mediated IPSCs in POMC neurons. c,

D-Trp

-gMSH hyperpolarizes POMC neurons.

Panels a±c are representative results. d, Expression of NPY in nerve terminals adjacent to

POMC neurons in the ARC. NPY nerve terminals (black, green arrowheads); POMC

neuronal soma (brown). Scale bar, 10 mm. e, Expression of GABA and NPY in nerve

terminals synapsing onto POMC neurons in the ARC. GABA immunoreactivity (10-nm gold

particles, red arrowheads) and NPY immunoreactivity (25-nm gold particles, green

arrows) are in separate vesicle populations colocalized in synaptic boutons that make

direct contact with the soma of POMC neurons (DAB contrasted with uranyl acetate and

lead citrate, diffuse black in cytoplasm). Scale bar, 1 mm. f, Model of leptin regulation of

NPY/GABA and POMC neurons in the ARC (see text).

letters to nature

484 NATURE

VOL 411

24 MAY 2001

www.nature.com

ships. This protocol was repeated in Krebs solution with 6.5 mM K

. To identify the

postsynaptic leptin current, I±V relationships were performed similarly with slow voltage

ramps (5 mV s

-1

from -100 to -20 mV) before and 10 min after adding leptin (100 nM).

GABA-mediated IPSCs were recorded using a CsCl internal electrode solution ((in

mM): 140 CsCl, 10 HEPES, 5 MgCl

, 1 BAPTA, 5 Mg-ATP, 0.3 Na-GTP). Both mini IPSCs

and large amplitude (presumably multisynaptic) IPSCs were observed in the untreated

slices. TTX (1 mM) abolished large IPSCs. We acquired data before and after drug addition

at a -50-mV holding potential in 2-s sweeps every 4 s for the times indicated in the ®gures.

Mini-postsynaptic currents were analysed using Axograph 4 (Axon Instruments). IPSCs

and excitatory postsynaptic currents (EPSCs) were distinguished on the basis of their

decay constants; in addition, picrotoxin (100 mM) blocked all IPSCs. POMC neurons

receive a low EPSC tone, and the frequency was not modulated by any of the treatments

described here.

Immunostaining for light and electron microscopy

We carried out double immunocytochemistry for NPY and POMC using different colour

diaminobenzidine (DAB) chromogens on ®xed mouse hypothalami according to

published protocols

. For electron microscopy, pre-embedding immunostaining for b-

endorphin was done with an ABC Elite kit (Vector Laboratories) and a DAB reaction,

followed by post-embedding labelling of GABA and NPY using rabbit anti-GABA, 1:1000

(v/v), and gold-conjugated (10-nm) goat anti-rabbit IgG or sheep anti-NPY and gold-

conjugated (25-nm) goat anti-sheep IgG. Sections were contrasted with saturated uranyl

acetate (10 min) and lead citrate (20±30 s), and examined using a Philips CM-10 electron

microscope.

Received 19 December 2000; accepted 12 March 2001.

1. Zhang, Y. et al. Positional cloning of the mouse obese gene and its human homologue. Nature 372,

425±432 (1994); erratum ibid. 374, 479 (1995).

2. Farooqi, I. S. et al. Effects of recombinant leptin therapy in a child with congenital leptin de®ciency.

N. Engl. J. Med. 341, 879±884 (1999).

3. Camp®eld, L. A., Smith, F. J., Guisez, Y., Devos, R. & Burn, P. Recombinant mouse OB protein:

evidence for a peripheral signal linking adiposity and central neural networks. Science 269, 546±549

(1995).

4. Kim, M. S. et al. The central melanocortin system affects the hypothalamo-pituitary thyroid axis and

may mediate the effect of leptin. J. Clin. Invest. 105, 1005±1011 (2000).

5. Haynes, W. G., Morgan, D. A., Walsh, S. A., Mark, A. L. & Sivitz, W. I. Receptor-mediated regional

sympathetic nerve activation by leptin. J. Clin. Invest. 100, 270±278 (1997).

6. Haynes, W. G., Morgan, D. A., Djalali, A., Sivitz, W. I. & Mark, A. L. Interactions between the

melanocortin system and leptin in control of sympathetic nerve traf®c. Hypertension 33, 542±547

(1999).

7. Hakansson, M. L., Brown, H., Ghilardi, N., Skoda, R. C. & Meister, B. Leptin receptor immunor-

eactivity in chemically de®ned target neurons of the hypothalamus. J. Neurosci. 18, 559±572 (1998).

8. Kalra, S. P. et al. Interacting appetite-regulating pathways in the hypothalamic regulation of body

weight. Endocr. Rev. 20, 68±100 (1999).

9. Cone, R. D. The central melanocortin system and energy homeostasis. Trends Endocrinol. Metab. 10,

211±216 (1999).

10. Elias, C. F. et al. Leptin differentially regulates NPY and POMC neurons projecting to the lateral

hypothalamic area. Neuron 23, 775±786 (1999).

11. Glaum, S. R. et al. Leptin, the obese gene product, rapidly modulates synaptic transmission in the

hypothalamus. Mol. Pharmacol. 50, 230±235 (1996).

12. Spanswick, D., Smith, M. A., Groppi, V. E., Logan, S. D. & Ashford, M. L. Leptin inhibits hypothalamic

neurons by activation of ATP-sensitive potassium channels. Nature 390, 521±525 (1997).

13. Lee, K., Dixon, A. K., Richardson, P. J. & Pinnock, R. D. Glucose-receptive neurones in the rat

ventromedial hypothalamus express KATP channels composed of Kir6.1 and SUR1 subunits.

J. Physiol. (Lond.) 515, 439±452 (1999).

14. Shiraishi, T., Sasaki, K., Niijima, A. & Oomura, Y. Leptin effects on feeding-related hypothalamic and

peripheral neuronal activities in normal and obese rats. Nutrition 15, 576±579 (1999).

15. Bagnol, D. et al. Anatomy of an endogenous antagonist: relationship between Agouti-related protein

and proopiomelanocortin in brain. J. Neurosci. [online] (cited 25 Aug. 99) ,http://www.jneurosci.

org/cgi/content/full/19/18/RC26. (1999).

16. Butler, A. A. et al. A unique metabolic syndrome causes obesity in the melanocortin-3 receptor-

de®cient mouse. Endocrinology 141, 3518±3521 (2000).

17. Young, J. I. et al. Authentic cell-speci®c and developmentally regulated expression of pro-opiome-

lanocortin genomic fragments in hypothalamic and hindbrain neurons of transgenic mice. J. Neurosci.

18, 6631±6640 (1998).

18. Franklin, K. B. J. & Paxinos, G. The Mouse Brain in Stereotaxic Coordinates (Academic, San Diego,

1997).

19. Kelly, M. J., Loose, M. D. & Ronnekleiv, O. K. Opioids hyperpolarize b-endorphin neurons via

m-receptor activation of a potassium conductance. Neuroendocrinology 52, 268±275 (1990).

20. Slugg, R. M., Hayward, M. D., Ronnekleiv, O. K., Low, M. J. & Kelly, M. J. Effect of the m-opioid

agonist DAMGO on medial basal hypothalamic neurons in b-endorphin knock-out mice.

Neuroendocrinology 72, 208±217 (2000).

21. Powis, J. E., Bains, J. S. & Ferguson, A. V. Leptin depolarizes rat hypothalamic paraventricular nucleus

neurons. Am. J. Physiol. 274, R1468±R1472 (1998).

22. Horvath, T. L., Bechmann, I., Naftolin, F., Kalra, S. P. & Leranth, C. Heterogeneity in the neuropeptide

Y-containing neurons of the rat arcuate nucleus: GABAergic and non-GABAergic subpopulations.

Brain Res. 756, 283±286 (1997).

23. Broberger, C., Landry, M., Wong, H., Walsh, J. N. & Hokfelt, T. Subtypes Y1 and Y2 of the

neuropeptide Y receptor are respectively expressed in pro-opiomelanocortin- and neuropeptide-Y-

containing neurons of the rat hypothalamic arcuate nucleus. Neuroendocrinology 66, 393±408 (1997).

24. King, P. J., Widdowson, P. S., Doods, H. N. & Williams, G. Regulation of neuropeptide Y release by

neuropeptide Y receptor ligands and calcium channel antagonists in hypothalamic slices.

J. Neurochem. 73, 641±646 (1999).

25. Grieco, P., Balse, P. M., Weinberg, D., MacNeil, T. & Hruby, V. J.

D-Amino acid scan of gamma-

melanocyte-stimulating hormone: importance of Trp(8) on human MC3 receptor selectivity. J. Med.

Chem. 43, 4998±5002 (2000).

26. Csiffary, A., Gorcs, T. J. & Palkovits, M. Neuropeptide Y innervation of ACTH-immunoreactive

neurons in the arcuate nucleus of rats: a correlated light and electron microscopic double

immunolabeling study. Brain Res. 506, 215±222 (1990).

27. Horvath, T. L., Naftolin, F. & Leranth, C. GABAergic and catecholaminergic innervation of

mediobasal hypothalamic beta-endorphin cells projecting to the medial preoptic area. Neuroscience

51, 391±399 (1992).

28. Cowley, M. A. et al. Integration of NPY, AGRP, and melanocortin signals in the hypothalamic

paraventricular nucleus: evidence of a cellular basis for the adipostat. Neuron 24, 155±163 (1999).

29. Halaas, J. L. et al. Physiological response to long-term peripheral and central leptin infusion in lean

and obese mice. Proc. Natl Acad. Sci. USA 94, 8878±8883 (1997).

Acknowledgements

We wish to thank V. J. Hruby for the D-Trp

-gMSH, O. K. Ronnekliev, R. G. Allen and

M. R. Brown for antisera and J. T. Williams and J. M. Brundege for advice. This work was

supported by the NIH, a Fogarty International Research Collaborative Award, the

International Scholar Program of the Howard Hughes Medical Institute, and Agencia

Nacional de Promocio

n Cienti®ca y Technolo

gica.

Correspondence and requests for materials should be addressed to R.D.C.

(e-mail: cone@ohsu.edu) or M.J.L. (e-mail: low@ohsu.edu).

.................................................................

Calmodulin bifurcates the local

signal that modulates

P/Q-type Ca

channels

Carla D. DeMaria*, Tuck Wah Soong

, Badr A. Alseikhan*,

Rebecca S. Alvania* & David T. Yue*

* The Johns Hopkins University School of Medicine, Departments of Biomedical

Engineering and Neuroscience, Program in Molecular and Cellular Systems

Physiology, 720 Rutland Avenue, Baltimore, Maryland 21205, USA

National Neuroscience Institute, 11 Jalan Tan Tuck Seng, Singapore 308433, and

Department of Physiology, National University of Singapore

..............................................................................................................................................

Acute modulation of P/Q-type (a

) calcium channels by neuro-

nal activity-dependent changes in intracellular Ca

concentra-

tion may contribute to short-term synaptic plasticity

1±3

potentially enriching the neurocomputational capabilities of

the brain

4,5

. An unconventional mechanism for such channel

modulation has been proposed

6,7

in which calmodulin (CaM)

may exert two opposing effects on individual channels, initially

promoting (`facilitation') and then inhibiting (`inactivation')

channel opening. Here we report that such dual regulation

arises from surprising Ca

-transduction capabilities of CaM.

First, although facilitation and inactivation are two competing

processes, both require Ca

-CaM binding to a single `IQ-like'

domain on the carboxy tail of a

; a previously identi®ed `CBD'

CaM-binding site

6,7

has no detectable role. Second, expression of a

CaM mutant with impairment of all four of its Ca

-binding sites

(CaM

1234

) eliminates both forms of modulation. This result

con®rms that CaM is the Ca

sensor for channel regulation,

and indicates that CaM may associate with the channel even

before local Ca

concentration rises. Finally, the bifunctional

capability of CaM arises from bifurcation of Ca

signalling by the

lobes of CaM: Ca

binding to the amino-terminal lobe selectively

initiates channel inactivation, whereas Ca

sensing by the carboxy-

terminal lobe induces facilitation. Such lobe-speci®c detection

provides a compact means to decode local Ca

signals in two

ways, and to separately initiate distinct actions on a single

molecular complex.

To simplify the dissection of the molecular mechanisms, we

studied recombinant P/Q-type (a

d) channels expressed

in mammalian HEK293 cells. Figure 1a shows that Ca

-dependent

Leptin activates anorexigenic POMC neurons through a neural network in the arcuate nucleus

Figures

Citations

Central nervous system control of food intake and body weight

Gut hormone PYY 3-36 physiologically inhibits food intake

BRIEF COMMUNICATION ARISING: Gut hormone PYY3-36 physiologically inhibits food intake

Inhibition of food intake in obese subjects by peptide YY3-36.

The Distribution and Mechanism of Action of Ghrelin in the CNS Demonstrates a Novel Hypothalamic Circuit Regulating Energy Homeostasis

References

The Mouse Brain in Stereotaxic Coordinates

Positional cloning of the mouse obese gene and its human homologue

Targeted gene expression as a means of altering cell fates and generating dominant phenotypes.

Recombinant mouse OB protein: evidence for a peripheral signal linking adiposity and central neural networks.

Effects of recombinant leptin therapy in a child with congenital leptin deficiency.

Related Papers (5)

Central nervous system control of food intake

Positional cloning of the mouse obese gene and its human homologue

Targeted disruption of the melanocortin-4 receptor results in obesity in mice

Antagonism of Central Melanocortin Receptors in Vitro and in Vivo by Agouti-Related Protein

Role of melanocortinergic neurons in feeding and the agouti obesity syndrome

Frequently Asked Questions (1)

Q1. What are the contributions in this paper?