Ablation of Sax2 gene expression prevents diet-induced obesity

Ablation of Sax2 gene expression prevents diet-inducedobesityRuth Simon1,2, Stefan Britsch2* and Andrew Bergemann1*

1 Department of Pathology, Mount Sinai School of Medicine, New York, NY, USA

2 Institute for Molecular and Cellular Anatomy, University of Ulm, Germany

Introduction

Obesity is increasingly becoming a major health hazard

throughout all industrialized societies. Easy access to

high caloric food and a sedentary life style are the

main causes for the increase in obesity and related

health risks, including diabetes mellitus and cardiovas-

cular diseases. Great efforts are being made to under-

stand the regulation of energy homeostasis and to find

ways of reducing the obesity epidemic and related

health risks. The regulation of energy homeostasis

occurs by a complex circuitry in the brain, particularly

in the hypothalamus and brainstem. These brain

circuitries integrate and coordinate several types of sig-

nals from the periphery, as well as from other parts of

the brain, including neurotransmitters, hormones and

nutrients, and translate them into feeding behavior,

thereby controlling energy uptake and expenditure.

Peripheral signals, including adiposity signals arising

from adipose tissue and the pancreas, as well as signals

from the gastrointestinal tract, interact with specific

neurons of the hypothalamus and the brainstem,

Keywords

brainstem; diet-induced obesity; energy

homeostasis; food uptake; neural circuitry

Correspondence

R. Simon, Institute for Molecular and

Cellular Anatomy, University of Ulm, Albert-

Einstein-Allee 11, D-89081 Ulm, Germany

Fax: +49 731 500 23102

Tel: +49 731 500 23225

E-mail: [email protected]

*These authors contributed equally to this

work

(Received 19 July 2010, revised 5

November 2010, accepted 11 November

2010)

doi:10.1111/j.1742-4658.2010.07960.x

Regulation of energy homeostasis is mainly mediated by factors in the

hypothalamus and the brainstem. Understanding these regulatory mecha-

nisms is of great clinical relevance in the treatment of obesity and related

diseases. The homeobox gene Sax2 is expressed predominantly in the

brainstem, in the vicinity of serotonergic neurons, and in the ventral neural

tube starting during early development. Previously, we have shown that the

loss of function of the Sax2 gene in mouse causes growth retardation start-

ing at birth and a high rate of postnatal lethality, as well as a dramatic

metabolic phenotype. To further define the role of Sax2 in energy homeo-

stasis, age-matched adult wild-type, Sax2 heterozygous and null mutant

animals were exposed to a high-fat diet. Although food uptake among the

different groups was comparable, Sax2 null mutants fed a high-fat diet

exhibited a significantly lower weight gain compared to control animals.

Unlike their counterparts, Sax2 null mutants did not develop insulin resis-

tance and exhibited significantly lower leptin levels under both standard

chow and high-fat diet conditions. Furthermore, neuropeptide Y, an

important regulator of energy homeostasis, was significantly decreased in

the forebrain of Sax2 null mutants on a high-fat diet. These data strongly

suggest a critical role for Sax2 gene expression in diet-induced obesity.

Sax2 gene expression may be required to allow the coordinated crosstalk

of factors involved in the maintenance of energy homeostasis, possibly regu-

lating the transcription of specific factors involved in energy balance.

Abbreviations

BAT, brown adipose tissue; H&E, hematoxylin and eosin; 5-HT, serotonin; NPY, neuropeptide Y; PAS, periodic acid-Schiff;

POMC, pro-opiomelanocortin; WAT, white adipose tissue.

FEBS Journal 278 (2011) 371–382 ª 2010 The Authors Journal compilation ª 2010 FEBS 371

respectively [1]. Adiposity signals, such as leptin and

insulin, interact specifically in a reciprocal way with

two neuron groups located in the arcuate nucleus of

the hypothalamus: the orexigenic neurons expressing

neuropeptide Y (NPY) and the anorectic neurons that

express pro-opiomelanocortin (POMC). High levels of

leptin and insulin prevent food intake by suppressing

the expression of NPY mRNA and by activating

POMC mRNA expression, whereas low levels activate

NPY mRNA expression, which in turn inhibits the

expression of POMC mRNA, leading to an increase in

appetite and potentially to obesity [2–6]. In addition,

there are NPY and POMC expressing neurons located

in nuclei of the brainstem involved in the regulation of

energy homeostasis, as well as receptors for leptin and

insulin allowing the crosstalk of the hypothalamus with

the brainstem and vice versa [7,8]. The brainstem nuclei

in turn receive information from the gastrointestinal

tract, through signals such as ghrelin and peptide YY,

relaying them to nuclei in the hypothalamus [9–14].

Possible candidates for the crosstalk between brainstem

and hypothalamus are serotonin (5-HT) and the mela-

nocortin pathway. Heisler et al. [15] reported specific

serotonin receptors, 5-HT2CR and 5-HT1BR, located on

POMC and NPY neurons, respectively. Antagonists,

particularly to 5-HT2CR, regulate energy balance by

activating the melanocortin pathway [16,17]. In turn,

melanocortin 4 receptors and NPY receptors are

located in the midbrain, the pons and the ventral

medulla, further suggesting an interaction between sero-

tonergic neurons and NPY and POMC neurons in the

hypothalamus [16,18,19]. The homeobox gene Sax2,

which is expressed predominantly in the brainstem,

plays a critical role in the regulation of serotonin, NPY

and POMC activities during early postnatal develop-

ment. Taken together with the dramatic metabolic phe-

notype exhibited by Sax2 null mutants, these data

strongly suggest an important function for Sax2 in the

regulation of energy homeostasis [20].

In the present study, we report that adult Sax2 null

mutants (Sax2) ⁄ )) are resistant to diet-induced obesity,

although their food uptake is comparable to wild-type

(Sax2+ ⁄ +) as well as Sax2 heterozygous (Sax2) ⁄ +)

animals. Sax2) ⁄ ) animals on a high-fat diet exhibit

normal glucose metabolism and do not develop insulin

resistance. In addition, NPY mRNA levels in the fore-

brain of Sax2) ⁄ ) on a high-fat diet are down-regu-

lated, whereas leptin levels are decreased independent

of the diet. The data obtained in the present study sug-

gest that glucose metabolism and energy storage path-

ways are indirectly affected by a lack of Sax2 gene

expression, most likely through an impairment of food

absorption.

Results

Comparison of body weight and food uptake of

adult wild-type, Sax2 heterozygous and null

mutants

During early postnatal development all Sax2) ⁄ ) ani-

mals show significant growth retardation independent

of their gender [21]. Examination of both male and

female adult Sax2) ⁄ ) animals revealed a significantly

smaller size compared to age-matched wild-type or

Sax2 heterozygous counterparts. Although there was

no difference in body weight between male Sax2) ⁄ +

and Sax2+ ⁄ + animals (n = 8; P > 0.05), the female

counterparts exhibited a significant weight difference,

as analyzed by a two-tailed Student’s t-test (n = 14;

P < 0.05; Fig. 1A), implying a heterozygous pheno-

type for females as a result of a dosage effect. Com-

paring Sax2+ ⁄ + and Sax2) ⁄ + animals to Sax2) ⁄ )

mice of the same gender revealed a more dramatic dif-

ference in body weight (females: Sax2+ ⁄ + and

Sax2) ⁄ +, n = 14; Sax2) ⁄ ), n = 16; Sax2+ ⁄ + ⁄Sax2) ⁄ +, P < 0.05; Sax2+ ⁄ + ⁄Sax2) ⁄ ), P < 0.0005

and Sax2) ⁄ + ⁄Sax2) ⁄ ), P < 0.005; males: Sax2+ ⁄ +

and Sax2) ⁄ +, n = 8; Sax2) ⁄ ), n = 7; Sax2+ ⁄ + ⁄Sax2) ⁄ ) and Sax2) ⁄ + ⁄Sax2) ⁄ ), P < 0.0001). In addi-

tion, the difference in average weight between control

and mutant animals was 2.5-fold greater for males

than females (Fig. 1A). To determine the cause of

these weight differences, we determined the average

daily food uptake of male and female animals of all

three genotypes over a period of 10 days. As shown in

Fig. 1B, the amount of food uptake of male Sax2) ⁄ )

did not differ significantly from their counterparts

(Sax2+ ⁄ + and Sax2) ⁄ +, n = 10; Sax2) ⁄ ), n = 6;

P > 0.05; Fig. 1B). This is different for female ani-

mals. Although there was no significant difference

when comparing Sax2) ⁄ + animals with Sax2+ ⁄ + or

Sax2) ⁄ ), there was a small, but significant difference

between Sax2+ ⁄ + and Sax2) ⁄ ) females (Sax2+ ⁄ + and

Sax2) ⁄ +, n = 10; Sax2) ⁄ ), n = 11; Sax2+ ⁄ + ⁄Sax2) ⁄ ) P < 0.05; Fig. 1B). Comparing the food

uptake between male and female animals of the corre-

sponding genotype revealed a significantly higher

amount (P < 0.05) for male Sax2+ ⁄ + and Sax2) ⁄ )

compared to the female counterparts, whereas there

was no difference between male and female Sax2) ⁄ +

animals. Furthermore comparing the ratio of food

uptake to body weight of the different genotypes and

genders revealed no significant differences for female

animals but a hyperphagic behavior for Sax2) ⁄ ) male

animals (P < 0.0005; Fig. 1C). These data imply a

gender-specific role for Sax2 in relation to food uptake

Sax2 expression required for diet-induced obesity R. Simon et al.

372 FEBS Journal 278 (2011) 371–382 ª 2010 The Authors Journal compilation ª 2010 FEBS

and ⁄or metabolism. Although, for female animals, the

slightly reduced food intake might account for the

weight difference, this is not the case for the male ani-

mals. The discrepancy between food uptake and body

weight suggests that Sax2) ⁄ ) mice either utilize energy

sources less efficiently for storage at least in the case of

male mutants or undergo higher energy expenditure.

To address the latter possibility, we measured the body

temperature of adult animals for 5 days in succession

at the same time of day. We found that Sax2 mutant

animals exhibited a significant lower body temperature

compared to control animals (Fig. 1D). The low body

temperature could be an indicator for fasting condi-

tions of the mutant mice, which does not correlate

with the food uptake data. In addition, we examined

the relative mRNA expression levels of specific mark-

ers involved in thermogenesis and metabolism, such as

uncoupling protein 1, peroxisome proliferator-activated

receptor c and peroxisome proliferator-activated recep-

tor coactivator 1a, amongst others, in white (WAT)

and brown (BAT) adipose tissue by quantitative real-

time RT-PCR. The mRNA expression level of these

markers did not exhibit a difference between control

and mutant tissues, suggesting that energy expenditure

as well as energy storage are not affected by a loss of

Sax2 expression (data not shown).

Sax2 null mutants are resistant to diet-induced

obesity

To further identify a role for the Sax2 gene in the reg-

ulation of energy homeostasis Sax2+ ⁄ +, Sax2) ⁄ +, as

well as Sax2) ⁄ ) animals, were exposed to a high-fat

diet. Age-matched male and female animals of all three

genotypes were single housed and exposed either to

standard chow or a high-fat diet for 6–11 weeks. The

weight gain of all animals was determined weekly at

the same time of day and day of the week. Animals

fed a standard diet showed a small increase in body

weight, with no significant differences between Sax2) ⁄ )

animals and their Sax2+ ⁄ + counterparts (females:

Sax2+ ⁄ +, n = 6, Sax2) ⁄ ) and Sax2) ⁄ +, n = 5;

males: Sax2+ ⁄ + and Sax2) ⁄ +, n = 5; Sax2) ⁄ ),

n = 2; Fig. 2A,B, left panel). By contrast Sax2+ ⁄ +

A B

C D

Fig. 1. Determination of body weight, food uptake and body temperature of Sax2 + ⁄ +, Sax2 ) ⁄ + and Sax2 ) ⁄ ) mice. Body weight (A), average

daily food uptake (B) and food uptake per gram of body weight (C) of single housed male and female Sax2 + ⁄ +, Sax2 ) ⁄ + and Sax2 ) ⁄ ) mice

at the age of 4 months (D) Determination of body temperature of adult Sax2 + ⁄ + and Sax2 ) ⁄ ) animals. All symbols with error bars are the

mean ± SEM and asterisks indicate the statistical significance: *P < 0.05; **P < 0.005; ***P < 0.0005; ****P < 0.0001.

R. Simon et al. Sax2 expression required for diet-induced obesity


and Sax2) ⁄ + animals on the high-fat diet exhibited a

dramatic weight gain compared to animals on standard

chow, as well as Sax2) ⁄ ) animals on a high-fat diet

(females: Sax2+ ⁄ +, n = 11; Sax2) ⁄ +, n = 12; Sax2) ⁄ ),

n = 7; males: Sax2+ ⁄ + and Sax2) ⁄ +, n = 5; Sax2) ⁄ ),

n = 3; Fig. 2A, B, right panel). Both male and female

Sax2) ⁄ ) animals gained weight during the first week on

the high-fat diet (females 6.4% and males 15.1% of the

initial body weight) but substantially less than their

Sax2+ ⁄ + and Sax2) ⁄ + counterparts. Although female

Sax2) ⁄ ) animals gained weight over the subsequent

6 weeks (12.6% of the initial body weight compared to

25.5% of the Sax2+ ⁄ + animals) (Fig. 2C), male Sax2) ⁄ )

animals unexpectedly lost weight after 3 weeks on a

high-fat diet (Fig. 2D). These data strongly suggest

that Sax2 gene expression is required for diet-induced

obesity and that its role is gender specific.

Histological analysis of adult WAT, BAT and liver

tissue

Previously, we have shown that the postnatal Sax2) ⁄ )

phenotype exhibits a lack of fat incorporation in WAT

and BAT, as well as low glycogen storage in the liver

[20]. The lack of energy storage most likely contributes

to the premature death of the majority of Sax2) ⁄ )

mice during early postnatal development [21]. To

determine whether the Sax2) ⁄ ) animals surviving to

adulthood exhibit a similar phenotype, we examined

WAT, BAT and liver tissues of Sax2+ ⁄ + and Sax2) ⁄ )

mice at the age of 6 months (Fig. 3A–H). In addition,

we performed histological staining assays on tissues

obtained from female animals fed a high-fat diet for

11 weeks to determine whether a special diet could res-

cue the phenotype (Fig. 3I–P). Hematoxylin and eosin

(H&E) staining analysis of animals fed standard chow

revealed very little fat incorporation into epididymal

WAT of Sax2) ⁄ ) mice compared to Sax2+ ⁄ + WAT

(Fig. 3A,B). Although there was an increase of fat

incorporation in WAT of Sax2) ⁄ ) mice on a high-fat

diet, the incorporation was far less than that into

WAT of Sax2+ ⁄ + animals. Indeed, the high-fat diet

did not even rescue incorporation into mutants to the

levels seen in wild-type animals on a standard diet, as

indicated by the smaller size of the adipose cells

(Fig. 3I,J,B). Analysis of BAT revealed a similar

A B

C D

Fig. 2. Sax2 expression is required for diet-induced obesity. (A–D) Determination of weight gain of single housed female (A, C) and male

(B, D) Sax2+ ⁄ +, Sax2) ⁄ + and Sax2) ⁄ ) mice at 6 weeks (A, B, D) or 11 weeks (C) on a high-fat diet. HFD, high-fat diet; STD, standard chow;

All symbols with error bars are the mean ± SEM and asterisks indicate the statistical significance: *P < 0.01 (A, B); P < 0.05 and 0.01 (C);

P < 0.05 to 0.0005 (D); **P < 0.005; ***P < 0.0005.



pattern. Although BAT of Sax2+ ⁄ + animals on a

high-fat diet for 11 weeks demonstrated drastically

increased fat incorporation compared to Sax2+ ⁄ + fed

a standard chow (Fig. 3C,K), fat incorporation in

BAT of Sax) ⁄ ) remained unchanged (Fig. 3D,L).

In addition to adipose tissue, we also examined liver

tissue by H&E as well as periodic acid-Schiff (PAS)

staining for morphological and glycogen storage differ-

ences, respectively. Liver tissue obtained from

Sax2+ ⁄ + and Sax2) ⁄ ) animals fed standard chow did

not show any structural differences with little or no fat

incorporation (Fig. 3E,F). In addition PAS staining

revealed less glycogen storage in the mutant animal

(Fig. 3G,H), which corresponds to the results obtained

in postnatal animals [20]. Examining the tissues of ani-

mals fed a high-fat diet, again we found a dramatic

increase of fat incorporation in the wild-type tissue

and to a very small extent in the mutant (Fig. 3M,N).

Furthermore H&E and PAS staining revealed high gly-

cogen and fat storage in the tissues of Sax2+ ⁄ +

animals, whereas Sax2) ⁄ ) liver tissues showed only

slightly elevated glycogen storage levels comparable to

the levels of Sax2+ ⁄ + animals fed standard chow

(Fig. 3O,P). Taken together, these data confirm that

the postnatal phenotype maintains through adulthood.

Furthermore, these data strongly suggest that the high-

fat diet cannot rescue the phenotype (e.g. increasing

incorporation of lipid and glycogen into respective tis-

sues). In addition, it is demonstrated that the necessary

specialized cells and molecular pathways required for

lipid and glycogen storage are present and functional

in Sax2) ⁄ ) mice.

Determination of blood glucose levels and serum

hormone assays

Deregulation of glucose metabolism could be one

explanation for resistance to diet-induced obesity. To

explore this possibility, we examined the glucose

metabolism of Sax2+ ⁄ + and Sax2) ⁄ ) animals by

determining fasting blood glucose levels, as well as by

performing glucose tolerance tests. Unlike their male

WAT

BAT

LiverH&E

LiverPAS

+/+ –/– +/+ –/–Standard diet High fat diet

A B

C D

E F

G H

I J

K L

M N

O P

Fig. 3. Histological analysis of WAT, BAT and liver tissues of female Sax2+ ⁄ + and Sax2) ⁄ ) animals. WAT, BAT and liver tissues of Sax2+ ⁄ +

(A, C, E, G, I, K, M, O) and Sax2) ⁄ ) animals (B, D, F, H, J, L, N, P) fed standard chow (A–H) or a high-fat diet (I–P), respectively. The tissues

were stained with H&E (A–F, I–N). Liver tissue was also stained with PAS reagent for glycogen incorporation (G, H, O, P). HFD, high-fat diet;

STD, standard diet. Size bar = 100 lm, WAT (A, B, I, J); 50 lm, BAT and liver (C–H, K to P).



counterparts, female Sax2) ⁄ ) animals fed a standard

chow showed small but significant higher fasting blood

glucose levels compared to Sax+ ⁄ + (n = 4 for all

groups; females, P < 0.05; Fig. 4A; data not shown).

However, neither female, nor male Sax2) ⁄ ) exhibited

any significant differences in the glucose tolerance

tests, suggesting that glucose metabolism per se is not

directly affected by lack of Sax2 gene expression

A B

C D

E F

Fig. 4. Analysis of blood glucose, insulin and leptin levels of female Sax2+ ⁄ + and Sax2) ⁄ ) animals. (A) Blood glucose levels of Sax2+ ⁄ + and

Sax2) ⁄ ) animals fed standard (STD) chow or a high-fat diet (HFD) after a 16 h fast. (B–D) Glucose tolerance tests of Sax2+ ⁄ + and Sax2) ⁄ )

animals fed standard chow (B) or a high-fat diet for 7 weeks (C) and 11 weeks (D). (E, F) Determination of blood insulin (E) and leptin levels

(F) of Sax2+ ⁄ + and Sax2) ⁄ ) animals fed standard chow and high-fat diet, respectively. The inset in (F) represents an enlargement of the STD

data to better demonstrate the ratio between Sax2+ ⁄ + and Sax2) ⁄ ) leptin levels. HFD, high-fat diet; STD, standard diet. All symbols with

error bars are the mean ± SEM and asterisks indicate the statistical significance: *P < 0.05; **P < 0.01; ***P < 0.005.



(n = 4 for all groups; Fig. 4B; data not shown). This

is in contrast to the data obtained from animals fed a

high-fat diet. Because of the smaller number of male

Sax2) ⁄ ) animals and the more severe reaction to the

high-fat diet, this analysis involved only female ani-

mals. Mutant animals fed a high-fat diet exhibited

lower blood glucose levels, although there was only a

significant difference at 7 weeks on the diet (n = 5 for

both groups) and not at 11 weeks (Sax2+ ⁄ +, n = 4;

Sax2) ⁄ ), n = 5; Fig. 4A). Furthermore, glucose toler-

ance tests performed on animals fed a high-fat diet for

7 weeks did not show a difference between control and

mutant animals (n = 5 for both groups; Fig. 4C). This

changed when glucose tolerance tests were performed

on animals fed a high-fat diet for 11 weeks. Although

the data obtained from Sax2) ⁄ ) animals on a high-fat

diet are comparable to animals on standard chow,

Sax2+ ⁄ + animals developed insulin resistance

(Sax2+ ⁄ +, n = 4; Sax2) ⁄ ), n = 5; Fig. 4D). Exami-

nation of serum insulin levels in Sax2+ ⁄ + and

Sax2) ⁄ ) animals, both on a standard as well as a high-

fat diet, further confirmed these data, as indicated by

significantly elevated insulin levels only in the

Sax2+ ⁄ + animals fed a high-fat diet (standard chow:

Sax2+ ⁄ +, n = 4; Sax2) ⁄ ), n = 3; P > 0.05; high-fat

diet: Sax2+ ⁄ +, n = 3; Sax2) ⁄ ), n = 4; P < 0.05;

Fig. 4E). Serum insulin levels of Sax2) ⁄ ) animals

remained the same on both diets.

To further establish the cause for resistance to diet-

induced obesity of Sax2) ⁄ ) mice, we determined serum

leptin levels, an additional major player in the regula-

tion of energy homeostasis. Leptin, an adipokine factor,

is expressed predominantly in WAT, and the secretion

of leptin occurs proportionally to the size of adipose tis-

sue [22]. As shown in Fig. 4F, serum leptin levels in

Sax2) ⁄ ) animals were significantly lower for both die-

tary groups compared to Sax2+ ⁄ + animals (standard

chow: Sax2+ ⁄ +, n = 3; Sax2) ⁄ ), n = 4; P < 0.01;

high-fat diet: Sax2+ ⁄ + and Sax2) ⁄ ), n = 4;

P < 0.005; Fig. 4F). Although serum leptin levels in

animals fed a high-fat diet increased 16-fold compared

to animals fed standard chow, the ratio of serum leptin

levels between Sax2+ ⁄ + and Sax2) ⁄ ) animals remained

the same under the different diets (Fig. 4F).

Analysis of Sax2 expression in the adult brain

To ensure that Sax2 expression of adult animals

occurs in the same pattern as during postnatal devel-

opment, we examined the brains of Sax2 heterozygous

and mutant animals by b-galactosidase staining. As

shown in Fig. 5A,B,D,E, Sax2 expression in the adult

brain was comparable to the expression pattern during

postnatal development [20,21]. Sax2 expression

occured in the hindbrain in the vicinity of the parame-

dian raphe (Fig. 5D) and the B3 raphe (Fig. 5E)

nuclei. In addition, we also found b-galactosidasestaining in the midbrain in Sax2 mutants, which is

absent from Sax2 heterozygous brains (Fig. 5B). Dur-

ing postnatal development, NPY and POMC expres-

sion, two critical factors in energy homeostasis, are

affected by the loss of Sax2 expression. To further

determine how Sax2 is involved in the regulation of

energy homeostasis, we performed co-localization

assays using an antibody recognizing b-galactosidaseas a marker for Sax2 expression and antibodies for

POMC, NPY and serotonin. The immunofluorescence

assays were performed on cryostat sections corre-

sponding to the hindbrain region, indicated by (e) in

Fig. 5C, representing the area of the nucleus of the sol-

itary tract (NTS), as well as the B3 raphe and Raphe

oralis regions. In the area of the NTS, POMC showed

co-localization with b-galactosidase (Fig. 5F–H),

whereas NPY was present in the vicinity of Sax2

expressing cells but did not co-localize (Fig. 5I–K).

Serotonin positive cells also were present in the vicinity

of Sax2 expressing cells but did not overlap, as shown

for the B3 raphe region (Fig. 5L–N). These data sug-

gest that Sax2 might be involved in the regulation of

energy homeostasis via the melanocortin pathway.

During postnatal development, we found an increase

of serotonin levels in the hindbrain of Sax2 mutants,

which most likely contributes to the phenotype. These

elevated serotonin levels were not found in the adult

hindbrain, as shown for the raphe oralis (Fig. 5O–P).

Determination of NPY and POMC mRNA

expression by real-time RT-PCR

It is well established that leptin is regulating energy

homeostasis in the brain by interacting with the leptin

receptor (ObRb), particularly receptors located on

NPY and POMC neurons in the arcuate nucleus of the

hypothalamus, as well as on nuclei of the brainstem

such as the NTS [12,14,23]. Under obese conditions,

humans and mice develop leptin resistance, resulting in

the loss of the inhibitory effect of leptin on NPY

expression [24]. Previously, we reported the deregula-

tion of NPY and POMC mRNA expression in

Sax2) ⁄ ) mice during postnatal development [20]. The

expression levels of NPY and POMC mRNA in the

mutant hindbrain imply a fasting status compared to

wild-type, whereas forebrain NPY mRNA levels sug-

gest satiation [20] (data not shown), indicating that

Sax2 expression might be required for the coordinated

crosstalk between factors involved in energy homeostasis.



To determine whether NPY and POMC mRNA

expression might be involved in the resistance of

Sax2) ⁄ ) mice to diet-induced obesity, we performed

real-time RT-PCR assays. RNA was isolated from the

hind- and forebrain of Sax2+ ⁄ + and Sax2) ⁄ ) animals

either fed standard chow or a high-fat diet. RT-PCR

was performed employing specific primers for NPY

and POMC mRNAs, as well as primers for GAPDH

mRNA as an internal standard. As shown in Fig. 6,

there was no significant difference in both NPY and

POMC mRNA levels in the fore- and hindbrain for

Sax2+ ⁄ + and Sax2) ⁄ ) animals on a standard chow

diet. Although POMC levels in the mutant were

reduced compared to the wild-type, this was not

statistically significant. Unlike during postnatal devel-

opment, NPY and POMC expression levels no longer

indicated fasting conditions for Sax2) ⁄ ) animals. In

addition, there was also no significant difference of

NPY and POMC expression in the hindbrain of ani-

mals on a high-fat diet, with the exception of signifi-

cantly lower POMC levels in the Sax) ⁄ )compared to

Sax2+ ⁄ + on standard chow. This differs from the

forebrain where the expression of NPY mRNA was

significantly lower in the Sax) ⁄ ) brain on a high-fat

diet compared to those from Sax+ ⁄ + on a high-fat

diet. In addition, there was a significant difference of

the NPY expression levels of Sax2 mutants on

standard and high-fat diet. The decrease of NPY

A B C

D

F G H

I J K

L M

O P

N

E

Fig. 5. Analysis of Sax2 expression pattern

in adult animals. (A) Lateral view of b-galac-

tosidase stained adult brains. Top: Sax2

heterozygous brain. Bottom: Sax2 mutant

brain. (B, D, E) Coronal sections of b-galac-

tosidase stained brains from the regions

indicated in (C) (b, d, e). (F–N) Co-expression

analysis by immunofluorescence of cryostat

sections of the hindbrain region indicated in

(C) (e) using an antibody recognizing b-galac-

tosidase as marker for Sax2 expression,

as well as antibodies for POMC, NPY and

serotonin; ·63 magnification; scale

bar = 7.5 lm. (O, P) Comparison of

serotonin concentration in control and Sax2

mutant animals; ·63 magnification; scale

bar = 25 lm. lacZ, green; POMC, NPY and

serotonin, red; *b-galactosidase staining;

B3, B3 raphe nuclei; NTS, nucleus of the

solitary tract; RPM, raphe paramedian;

RO, raphe oralis.



mRNA levels further indicates the requirement of

Sax2 expression for diet-induced obesity. Obese ani-

mals develop leptin resistance, which is manifested in

the loss of the inhibitory effect of leptin on NPY

mRNA expression [24]. We further conclude from

these data, unlike during postnatal development, NPY

and POMC mRNA expression in the adult hindbrain

is no longer as strongly affected by the loss of Sax2

expression.

Discussion

Sax2, also called Nkx1.1, is a homeobox gene of the

Nkx1 gene family located on chromosome 5 of the

mouse genome. The human homolog is located on

chromosome 4 in the vicinity of the Wolf–Hirschhorn

syndrome. To date, no involvement in this disorder has

been determined for Sax2 [25]. In the mouse, loss of

Sax2 gene expression causes postnatal lethality and a

dramatic metabolic phenotype [20,21]. Few of the Sax2

mutants survive to adulthood further exhibiting a lean

phenotype. We have shown previously that serotonin

levels in the mutant are increased in the postnatal hind-

brain [20]. The data obtained in the present study sug-

gest that, during postnatal development, Sax2 might be

involved in the regulation of serotonin synthesis but

loses this function later in development (Fig. 5O–P).

Serotonin plays an important role during pre- and post-

natal development of the brain. It is possible that seroto-

nin levels of the surviving mutants are more moderately

increased, thereby allowing a closer to normal develop-

ment of the brain and survival to adulthood.

In the present study, we demonstrate that the abla-

tion of Sax2 gene expression prevents diet-induced

obesity in adult mice. There are several possibilities

that might cause the metabolic phenotype of Sax2 (i.e.

either deregulation of food uptake, food absorption

and ⁄or a defect in the metabolic pathways to store

energy). Glucose tolerance tests, as well as serum insu-

lin levels, suggest that the glucose metabolism per se is

not affected by Sax2 deficiency. Furthermore, our his-

tological analysis of adipose and liver tissues demon-

strates that Sax2) ⁄ ) mice are able to incorporate fat as

well as glycogen also to a lesser extent compared to

Sax2+ ⁄ + mice. These data strongly suggest that the

pathways required for energy storage (e.g. storage of

lipids and glycogens) are not directly affected by lack

of Sax2 gene expression. In addition, food uptake by

Sax2) ⁄ ) animals is comparable to Sax2+ ⁄ + and

Sax2) ⁄ + animals, although female mutants take up

slightly less than their counterparts. Overall, the differ-

ence in food uptake does not account for the size dif-

ferences, particularly not in the case of male animals.

Loss of Sax2 expression could also affect energy

expenditure. Although we did not observe hyperactive

behavior of the mutant animals, it is possible that

increased energy expenditure is responsible for their

lean phenotype. Both male and female mutants on a

high-fat diet exhibit wet fur starting in the neck, which

is an area close to BAT. It is possible that this specific

diet causes a rise in surface temperature, as was shown

for the DGAT1 mice. However, unlike the DGAT1

mutants, Sax2 mutants do not exhibit increased UCP1

mRNA expression, which is an important factor in the

regulation of body temperature and energy expenditure

[26] (data not shown). Indeed, we have demonstrated

that Sax2 mutants under normal feeding condi-

tions exhibit a lower body temperature compared to

A

B

Fig. 6. Determination of mRNA expression levels of NPY and

POMC by real-time RT-PCR. Determination of relative NPY (A) and

POMC (B) mRNA levels in the fore- and hindbrain of female

Sax2+ ⁄ + and Sax2) ⁄ ) animals fed standard chow (STD) or a high-

fat diet (HFD), respectively, by real-time RT-PCR. Statistical analysis

by the 2)DDCT method with Sax2+ ⁄ + on STD as reference; FB, fore-

brain; HB, hindbrain; *P < 0.05.



wild-type animals, which suggests that the lean pheno-

type is not the result of energy expenditure and rather

indicates a fasting status of the animals.

It is possible that the loss of Sax2 expression

impairs the absorption of nutrients. Several studies

link leptin to the regulation of intestinal absorption of

nutrients in addition to its regulatory role in the brain

[27–30]. These reports demonstrate that leptin levels

correlate with absorption efficiency [27,30]. We have

shown that leptin levels are considerably lower in

Sax2) ⁄ ) mice, potentially accounting for the reduction

in weight through low absorption efficiencies. How-

ever, the question remains as to whether low leptin lev-

els are the cause or effect of the lean phenotype.

The data of the present study comparing male and

female Sax2) ⁄ ) animals suggest a gender-specific role

for Sax2 in energy homeostasis. The difference in body

weight is much more severe in male than in female

Sax2) ⁄ ) mice, particularly in animals fed a high-fat

diet. Unexpectedly, male Sax2) ⁄ ) mice on a high-fat

diet lose weight, falling below the starting weight after

an initial weight gain. It is well known that fat storage

occurs differently in males and females [31,32]. While

males accumulate fat preferentially in abdominal and

visceral tissues, females store fat subcutaneously [31].

One major factor in the gender-specific distribution of

fat accumulation is estrogen, as shown in rats under-

going an ovariectomy, which led to an increase in

visceral fat and loss of subcutaneous fat; it was fur-

ther demonstrated that estrogen treatment was able to

restore fat distribution [33].

Sax2 is a transcription factor and, although we have

not determined a function for its role during early

development and ⁄or in energy homeostasis, it is possi-

ble that Sax2 deficiency prevents the crosstalk of fac-

tors involved in the maintenance of energy

homeostasis. During postnatal development, the loss

of Sax2 expression causes an increase in serotonin lev-

els in the brainstem and a deregulation of NPY and

POMC expression in the hind- and forebrain [20], sug-

gesting that the crosstalk between the different regions

of the brain involved in energy balance is affected.

This could occur either through an involvement of

Sax2 in the development of morphological structures

(e.g. brain circuits) or the regulation of the expression

of factors required for the regulation of energy homeo-

stasis. Further studies are required to determine the

pathway(s) through which Sax2 is regulating energy

homeostasis. In particular, the identification of target

genes will be an important step forward in defining a

role for Sax2 in energy homeostasis. Altogether, Sax2

provides an excellent model for studying the regulation

of energy homeostasis by neurons of the brainstem.

Materials and methods

Animals

The generation of Sax2) ⁄ ) has been described elsewhere

[21]. All experiments were performed on animals with a

mixed genetic background of S129 ⁄C57BL ⁄ 6J. Mutant,

wild-type and heterozygous animals were all taken from the

same litter. Food uptake, high-fat diet and glucose toler-

ance tests were performed on age-matched adult male and

female animals starting at 4 months of age. Body tempera-

ture was measured on 6 days in succession at the same time

of day using a veterinary thermometer for rodents (Micro-

life, Widnau, Switzerland). Experiments were carried out in

accordance with the guidelines of the Mount Sinai School

of Medicine Institutional Animal Care and Use Committee

(USA).

Analysis of food uptake and high-fat diet

Four-month-old male and female Sax2+ ⁄ +, Sax2) ⁄ + and

Sax2) ⁄ ) animals were single housed 1 week before the start

of the experiment. To determine the daily food uptake, the

animals were fed a Nutra-Gel diet (BioServ, Frenchtown,

NJ, USA; catalog number S4798) for 10 days in succession

and the amount of food consumed was measured daily at

the same time of day. Daily food intake was determined by

averaging the amount of food consumed for the last 5 days

in succession of the experiment. After 1 week on standard

chow, half the animals were exposed to a high-fat diet (Bio-

Serv; catalog number F2685; 35.5% fat, 35% carbohydrate,

20% protein, 0.1% fiber and 3.7% ash; caloric intake

amounts to 5.4 kcalÆg)1) for 6–11 weeks, whereas the con-

trol group was fed a standard chow (Purina Mills, LLC,

Gray Summit, MO, USA; catalog number 5053). The body

weight of all animals was determined weekly.

Glucose tolerance test

Female Sax2+ ⁄ + and Sax2) ⁄ ) animals fed a standard

chow or a high-fat diet were starved for 12–16 h before the

experiment. Fasting blood glucose levels were determined

using a One Touch glucose meter (Lifescan, Johnson and

Johnson, Milpitas, CA, USA), followed by an intraperito-

neal injection of a 10% glucose solution (2 mg glucoseÆgbody weight)1). Blood glucose levels were determined at 5,

15, 30, 60 and 120 min after injection.

Blood serum analysis

Blood was collected from Sax2+ ⁄ + and Sax2) ⁄ ) animals

fed a standard chow as well as a high-fat diet. Blood insulin

and leptin levels were determined using ELISA kits (Crystal

Chem Inc., Downers Grove, IL, USA) in accordance with

the manufacturer’s instructions.



Histological analysis

Brains, WAT, BAT and liver tissues were collected and

fixed in 4% paraformaldehyde overnight, washed in

NaCl ⁄Pi, dehydrated through graded ethanol, followed by

two changes in Americlear (Fisher Scientific Co., Pitts-

burgh, PA, USA) and embedded in Paraplast (Fisher Scien-

tific Co.). Both H&E and PAS staining methods were

performed as described previously [20].

Sax2 expression studies

b-Galactosidase (Carl Roth GmbH, Karlsruhe, Germany)

stained adult Sax2 heterozygous and homozygous brains

were embedded in 1% sucrose and 50 lm vibratom sections

were prepared. Co-expression studies were performed on

14 lm cryostat sections of adult Sax2 heterozygous and

homozygous brains employing monoclonal anti-b-galactosi-dase serum (Sigma-Aldrich, St Louis, MO, USA) (dilution

1 : 2000) as marker for Sax2 expression, as well as anti-sero-

tonin (Immunostar, Inc., Hudson, WI, USA) (dilution

1 : 3000), anti-NPY (dilution 1 : 5000) and anti-POMC

(dilution 1 : 1000) (both Sigma-Aldrich) polyclonal sera.

Secondary antibodies, CyTM2-conjugated donkey anti-mouse

(b-galactosidase), CyTM3-conjugated donkey anti-rabbit

(serotonin, NPY) and CyTM3-conjugated donkey ant-chicken

(POMC) (Jackson ImmunoResearch, Newmarket, UK) were

used at a dilution of 1 : 500. Images were obtained at a Leica

TCS SP5-II (Leica Microsystems, Wetzlar, Germany) confo-

cal laser scanning microscope.

Quantitative real-time RT-PCR

Total RNA was prepared from fore- and hindbrains of

Sax2+ ⁄ + and Sax2) ⁄ ) fed a standard chow or a high-fat

diet, respectively using RNeasy Minikit (Qiagen, Hilden,

Germany). Quantitative real-time RT-PCR was performed

as described previously [20] using the oligonucleotides:

NPY, 5¢-GCTTGAAGACCCTTCCATTGG-3¢ and 5¢-GG

CGGAGTCCAGCCTAGTGG-3¢; POMC, 5¢-CATTAGG

CTTGGAGCAGGTC-3¢ and 5¢-GAATGAGAAGACCCC

TGCAC-3¢; and GAPDH, 5¢-CCAGAGCTGAACGGGAA

G-3¢ and 5¢-TGCTGTTGAAGTCGCAGG-3¢.

Statistical analysis

Results are expressed as the mean ± SEM. Comparisons

between groups were made by an unpaired two-tailed Stu-

dent’s t-test or analyzed using the 2)DDCT method, as

described previously [34] (quantitative real-time RT-PCR).

P < 0.05 was considered statistically significant.

Acknowledgements

We would like to thank Ms Jacqueline Andratschke

for her excellent technical assistance. This work was

supported in part by the Irma T. Hirschl Charitable

Trust to A.B.

References

1 Shin AC, Zheng H & Berthoud HR (2009) An

expanded view of energy homeostasis: neural integra-

tion of metabolic, cognitive, and emotional drives to

eat. Physiol Behav 97, 572–580.

2 Schwartz MW, Woods SC, Porte D Jr, Seeley RJ &

Baskin DG (2000) Central nervous system control of

food intake. Nature 404, 661–671.

3 Friedman JM (2004) Modern science versus the stigma

of obesity. Nat Med 10, 563–569.

4 Cone RD (2005) Anatomy and regulation of the central

melanocortin system. Nat Neurosci 8, 571–578.

5 Park AJ & Bloom SR (2005) Neuroendocrine control of

food intake. Curr Opin Gastroenterol 21, 228–233.

6 Myers MG Jr, Munzberg H, Leinninger GM & Leshan

RL (2009) The geometry of leptin action in the brain:

more complicated than a simple ARC. Cell Metab 9,

117–123.

7 Grill HJ & Kaplan JM (2002) The neuroanatomical axis

for control of energy balance. Front Neuroendocrinol 23,

2–40.

8 Grill HJ, Schwartz MW, Kaplan JM, Foxhall JS, Brein-

inger J & Baskin DG (2002) Evidence that the caudal

brainstem is a target for the inhibitory effect of leptin

on food intake. Endocrinology 143, 239–246.

9 Berthoud HR (2008) The vagus nerve, food intake and

obesity. Regul Pept 149, 15–25.

10 Skibicka KP & Grill HJ (2009) Hindbrain leptin stimu-

lation induces anorexia and hyperthermia mediated by

hindbrain melanocortin receptors. Endocrinology 150,

1705–1711.

11 Ahima RS & Antwi DA (2008) Brain regulation of

appetite and satiety. Endocrinol Metab Clin North Am

37, 811–823.

12 Huo L, Maeng L, Bjorbaek C & Grill HJ (2007) Leptin

and the control of food intake: neurons in the nucleus

of the solitary tract are activated by both gastric disten-

sion and leptin. Endocrinology 148, 2189–2197.

13 Zheng H, Patterson LM, Morrison C, Banfield BW, Ran-

dall JA, Browning KN, Travagli RA & Berthoud HR

(2005) Melanin concentrating hormone innervation of

caudal brainstem areas involved in gastrointestinal func-

tions and energy balance. Neuroscience 135, 611–625.

14 Schwartz GJ & Moran TH (2002) Leptin and neuropep-

tide y have opposing modulatory effects on nucleus of

the solitary tract neurophysiological responses to gastric

loads: implications for the control of food intake.

Endocrinology 143, 3779–3784.

15 Heisler LK, Jobst EE, Sutton GM, Zhou L, Borok E,

Thornton-Jones Z, Liu HY, Zigman JM, Balthasar N,

Kishi T et al. (2006) Serotonin reciprocally regulates



melanocortin neurons to modulate food intake. Neuron

51, 239–249.

16 Lam DD, Przydzial MJ, Ridley SH, Yeo GS, Rochford

JJ, O’Rahilly S & Heisler LK (2008) Serotonin 5-HT2C

receptor agonist promotes hypophagia via downstream

activation of melanocortin 4 receptors. Endocrinology

149, 1323–1328.

17 Xu Y, Jones JE, Kohno D, Williams KW, Lee CE,

Choi MJ, Anderson JG, Heisler LK, Zigman JM, Low-

ell BB et al. (2008) 5-HT2CRs expressed by pro-opio-

melanocortin neurons regulate energy homeostasis.

Neuron 60, 582–589.

18 Kishi T, Aschkenasi CJ, Choi BJ, Lopez ME, Lee CE,

Liu H, Hollenberg AN, Friedman JM & Elmquist JK

(2005) Neuropeptide Y Y1 receptor mRNA in rodent

brain: distribution and colocalization with melanocor-

tin-4 receptor. J Comp Neurol 482, 217–243.

19 Kishi T, Aschkenasi CJ, Lee CE, Mountjoy KG, Saper

CB & Elmquist JK (2003) Expression of melanocortin 4

receptor mRNA in the central nervous system of the

rat. J Comp Neurol 457, 213–235.

20 Simon R, Lufkin T & Bergemann AD (2007) Homeo-

box gene Sax2 deficiency causes an imbalance in energy

homeostasis. Dev Dyn 236, 2792–2799.

21 Simon R & Lufkin T (2003) Postnatal lethality in mice

lacking the Sax2 homeobox gene homologous to

Drosophila S59 ⁄ slouch: evidence for positive and

negative autoregulation. Mol Cell Biol 23, 9046–9060.

22 Considine RV, Sinha MK, Heiman ML, Kriauciunas

A, Stephens TW, Nyce MR, Ohannesian JP, Marco

CC, McKee LJ, Bauer TL et al. (1996) Serum immuno-

reactive-leptin concentrations in normal-weight and

obese humans. N Engl J Med 334, 292–295.

23 Huo L, Gamber KM, Grill HJ & Bjorbaek C (2008)

Divergent leptin signaling in proglucagon neurons

of the nucleus of the solitary tract in mice and rats.

Endocrinology 149, 492–497.

24 Enriori PJ, Evans AE, Sinnayah P, Jobst EE, Tonelli-

Lemos L, Billes SK, Glavas MM, Grayson BE, Perello

M, Nillni EA et al. (2007) Diet-induced obesity causes

severe but reversible leptin resistance in arcuate melano-

cortin neurons. Cell Metab 5, 181–194.

25 Simon R & Bergemann A (2008) Mouse models of

Wolf–Hirschhorn syndrome. Am J Med Genet C Semin

Med Genet 148C, 275–280.

26 Chen HC, Ladha Z, Smith SJ & Farese RV Jr (2003)

Analysis of energy expenditure at different ambient

temperatures in mice lacking DGAT1. Am J Physiol

Endocrinol Metab 284, E213–218.

27 Hindlet P, Bado A, Kamenicky P, Delomenie C, Bouras-

set F, Nazaret C, Farinotti R & Buyse M (2009) Reduced

intestinal absorption of dipeptides via PepT1 in mice with

diet-induced obesity is associated with leptin receptor

down-regulation. J Biol Chem 284, 6801–6808.

28 Inigo C, Patel N, Kellett GL, Barber A & Lostao MP

(2007) Luminal leptin inhibits intestinal sugar absorp-

tion in vivo. Acta Physiol (Oxf) 190, 303–310.

29 Miyawaki K, Yamada Y, Ban N, Ihara Y, Tsukiyama

K, Zhou H, Fujimoto S, Oku A, Tsuda K, Toyokuni S

et al. (2002) Inhibition of gastric inhibitory polypeptide

signaling prevents obesity. Nat Med 8, 738–742.

30 Buyse M, Berlioz F, Guilmeau S, Tsocas A, Voisin T,

Peranzi G, Merlin D, Laburthe M, Lewin MJ, Roze C

et al. (2001) PepT1-mediated epithelial transport of

dipeptides and cephalexin is enhanced by luminal

leptin in the small intestine. J Clin Invest 108, 1483–

1494.

31 Lovejoy JC & Sainsbury A (2009) Sex differences in

obesity and the regulation of energy homeostasis. Obes

Rev 10, 154–167.

32 Farhang B, Diaz S, Tang SL & Wagner EJ (2009) Sex

differences in the cannabinoid regulation of energy

homeostasis. Psychoneuroendocrinology 3, S237–S246.

33 Clegg DJ, Brown LM, Woods SC & Benoit SC (2006)

Gonadal hormones determine sensitivity to central

leptin and insulin. Diabetes 55, 978–987.

34 Livak KJ & Schmittgen TD (2001) Analysis of relative

gene expression data using real-time quantitative PCR

and the 2(-delta delta C(T)) method. Methods 25, 402–

408.



Ablation of Sax2 gene expression prevents diet-induced obesity

Documents

Transcript of Ablation of Sax2 gene expression prevents diet-induced obesity