Activation of the cardiac ciliary neurotrophic factor receptor reverses left ventricular hypertrophy in leptin-deficient and leptin-resistant obesity

Abstract

Disruption of the leptin signaling pathway within the heart causes left ventricular hypertrophy (LVH). Because human obesity is a syndrome of leptin resistance, which is not amenable to leptin treatment, the identification of parallel signal transduction pathways is of potential therapeutic value. Ciliary neurotrophic factor (CNTF), which acts parallel to leptin in the hypothalamus, is not previously recognized to have cardiac activity. We hypothesized that CNTF receptors are present on cardiomyocytes and their activation reverses LVH in both leptin-deficient ob/ob and leptin-resistant db/db mice. The localization of CNTF receptors (CNTFRα) to the sarcolemma in C57BL/6, ob/ob and db/db was confirmed in situ with immunohistochemistry, and immunoblotting (60 and 40 kDa) on isolated myocytes. ob/ob mice were randomly assigned to receive s.c. recombinant CNTF (CNTFAx15; 0.1 mg·kg⁻¹ per day; n = 11) calorie-restriction (n = 9), or feeding ad libitum (n = 11). db/db mice were allocated to three similar groups (n = 8, 7, and 8, respectively) plus a leptin group (1 mg·kg⁻¹ per day; n = 7). Echocardiography showed that CNTFAx15 reduced cardiac hypertrophy [posterior wall thickness decreased by 29 ± 8% (P < 0.01) in ob/ob and by 21 ± 3% in db/db mice (P < 0.01)], which was consistent with the reduction of myocyte width. Western blotting showed that leptin and CNTFAx15 activated Stat3 and ERK1/2 pathway in cultured adult mice cardiomyocytes and cardiac tissue from in ob/ob and db/db mice. Together, these findings support the role of a previously undescribed signaling pathway in obesity-associated cardiac hypertrophy and have therapeutic implications for patients with obesity-related cardiovascular disease and other causes of LVH.

• signal transduction
• cardiac remodeling
• mouse
• heart
• Axokine

Full text

Activation of the cardiac ciliary neurotrophic factor
receptor reverses left ventricular hypertrophy
in leptin-deficient and leptin-resistant obesity
Shubha V. Y. Raju*
†
, Meizi Zheng*
†
, Karl H. Schuleri*
†
, Alexander C. Phan*, Djahida Bedja*, Roberto M. Saraiva*,
Omer Yiginer*, Koenraad Vandegaer*, Kathleen L. Gabrielson*, Christopher P. O’Donnell
‡
, Dan E. Berkowitz
§
,
Lili A. Barouch*, and Joshua M. Hare*
¶
*Division of Cardiology and Institute for Cell Engineering, §Department of Anesthesiology and Critical Care Medicine, Johns Hopkins Medical Institutions,
Baltimore, MD 21205; and ‡Division of Pulmonary Medicine, Department of Medicine, University of Pittsburgh School of Medicine, Pittsburgh, PA 15213
Edited by Solomon H. Snyder, Johns Hopkins University School of Medicine, Baltimore, MD, and approved January 5, 2006 (received for review
December 5, 2005)
Disruption of the leptin signaling pathway within the heart causes
left ventricular hypertrophy (LVH). Because human obesity is a
syndrome of leptin resistance, which is not amenable to leptin
treatment, the identification of parallel signal transduction path-
ways is of potential therapeutic value. Ciliary neurotrophic factor
(CNTF), which acts parallel to leptin in the hypothalamus, is not
previously recognized to have cardiac activity. We hypothesized
that CNTF receptors are present on cardiomyocytes and their
activation reverses LVH in both leptin-deficient ob兾ob and leptin-
resistant db兾db mice. The localization of CNTF receptors (CNTFR
␣
)
to the sarcolemma in C57BL兾6, ob兾ob and db兾db was confirmed in
situ with immunohistochemistry, and immunoblotting (60 and 40
kDa) on isolated myocytes. ob兾ob mice were randomly assigned to
receive s.c. recombinant CNTF (CNTF
Ax15
; 0.1 mg䡠kg
ⴚ1
per day; nⴝ
11) calorie-restriction (nⴝ9), or feeding ad libitum (nⴝ11). db兾db
mice were allocated to three similar groups (nⴝ8, 7, and 8,
respectively) plus a leptin group (1 mg䡠kg
ⴚ1
per day; nⴝ7).
Echocardiography showed that CNTF
Ax15
reduced cardiac hyper-
trophy [posterior wall thickness decreased by 29 ⴞ8% (P<0.01)
in ob兾ob and by 21 ⴞ3% in db兾db mice (P<0.01)], which was
consistent with the reduction of myocyte width. Western blotting
showed that leptin and CNTF
Ax15
activated Stat3 and ERK1兾2
pathway in cultured adult mice cardiomyocytes and cardiac tissue
from in ob兾ob and db兾db mice. Together, these findings support
the role of a previously undescribed signaling pathway in obesity-
associated cardiac hypertrophy and have therapeutic implications
for patients with obesity-related cardiovascular disease and other
causes of LVH.
signal transduction 兩cardiac remodeling 兩mouse 兩heart 兩Axokine
Left ventricular hypertrophy (LVH) and its subsequent pro-
gression to congestive heart failure represents a major cause
of morbidity and mortality in the United States (1). Obesity, an
important mediator of LVH (2), results from either deficiency of
or receptor insensitivity to leptin (3–5), a hormone that regulates
appetite and energy metabolism (6). We have previously dem-
onstrated that both leptin deficiency and resistance contribute to
LVH in murine models (7). Additionally, we showed regression
of LVH with leptin repletion in leptin-deficient ob兾ob mice.
However, because the majority of human obesity is associated
with hyperleptinemia and leptin resistance (4, 8, 9) that is
unresponsive to leptin treatment, we sought to identify an
alternate signaling axis that regulates cardiac architecture. Here
we address this issue with ciliary neurotrophic factor (CNTF),
which activates a related signaling pathway to leptin and has
similar effects on body weight and metabolism (10 –12). The
CNTF receptor (CNTFR) complex closely resembles the leptin
receptor (ObR) structurally and has a similar distribution in the
hypothalamic nuclei associated with regulation of feeding and
body weight (13, 14). Both receptors are members of the gp130
cytokine family of receptors. The CNTF receptor is a trimeric
receptor complex with a CNTF binding component (CNTFR
␣
),
a leukemia inhibitory factor
␤
subunit, and the signal transducer
of IL-6 (gp130) (15). Support for parallels between leptin and
CNTF signaling pathways include structural homology between
ObR and the gp130 subunit of the CNTFR (16) as well as
activation of similar signal transduction pathways such as the
Janus kinase-signal transducer and activator of transcription
pathway (17–19). Thus, we tested whether CNTF receptors are
present and functional in the heart and whether their activation
would reverse established LVH in both ob兾ob and db兾db
mice (10).
Results
Presence of CNTFR
␣
Receptors in Cardiac Tissue. To determine
whether CNTFRs were present in the heart, we initially per-
formed in situ peroxidase staining. The CNTFR (CNTFR
␣
) was
visualized on myocytes in both longitudinal and transverse
sections of C57bl兾6 wt mice hearts (Fig. 1 aand c). Double
immunofluorescence staining for desmin and CNTFR
␣
local-
ized its presence to the sarcolemma (Fig. 1e), which was further
confirmed by immunoblots (60-kDa glycosylated and 40-kDa
Conflict of interest statement: No conflicts declared.
This paper was submitted directly (Track II) to the PNAS office.
Abbreviations: CNTF, ciliary neurotrophic factor; CNTFR, ciliary neurotrophic factor recep-
tor; LV, left ventricle; LVH, left ventricular hypertrophy; LVM, left ventricular mass.
†S.V.Y.R., M.Z., and K.H.S. contributed equally to this work.
¶To whom correspondence should be addressed. E-mail: jhare@mail.jhmi.edu.
© 2006 by The National Academy of Sciences of the USA
Table 1. Baseline body weights and echocardiographic
parameters in WT, ob兾ob, and db兾db mice prior to the start
of intervention
Parameter WT ob兾ob db兾db
No. of mice 15 31 30
Age, months 5 5–6 5–6
BW, g 22 ⫾070⫾1* 59 ⫾1*
†
IVSd, mm 0.76 ⫾0.03 1.10 ⫾0.02* 1.03 ⫾0.03*
PWTd, mm 0.85 ⫾0.02 1.03 ⫾0.02* 0.96 ⫾0.02*
LVEDd, mm 3.26 ⫾0.10 3.51 ⫾0.07 3.59 ⫾0.07
‡
LVEDs, mm 1.63 ⫾0.06 1.51 ⫾0.07 1.51 ⫾0.07
LVM, mg 87 ⫾6 143 ⫾4* 138 ⫾5*
BW, body weight; IVSd, interventricular septal thickness in diastole; PWTd,
posterior wall thickness in diastole; LVEDd, left ventricular end-diastolic di-
ameter; LVEDs, left ventricular end-systolic diameter. *, P⬍0.01 vs. WT; †, P⬍
0.01 vs. ob兾ob; and ‡, P⬍0.05 vs. WT by one-way ANOVA.
4222–4227
兩
PNAS
兩
March 14, 2006
兩
vol. 103
兩
no. 11 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0510460103

nonglycosylated bands) on isolated myocytes (Fig. 1g). Mouse
brain was used as positive control.
Regression of LVH in the
ob
兾
ob
Mice with CNTF
Ax15
.Based on our
earlier finding that leptin regresses established LVH in ob兾ob
mice (ref. 7; Table 1), we hypothesized that CNTF
Ax15
would
mimic these effects on cardiac structure. CNTF
Ax15
regressed
cardiac hypertrophy, decreasing interventricular septal thickness
by 20 ⫾3% (P⫽0.000035) posterior wall thickness by 29 ⫾8%,
(P⫽0.000022), (Fig. 2 band h), and calculated left ventricular
(LV) mass by 21 ⫾9%, (P⫽0.0077). In contrast and as
previously shown, calorie restriction did not reduce LVH in these
mice (Table 2). The degree of regression with CNTF
Ax15
re-
stored wall thickness and left ventricular mass (LVM) essentially
to normal. (Table 2) Interestingly, a reduction in LVM was noted
in the calorie restricted group, but this decrease was due entirely
to a smaller chamber size in these animals and, in the absence
of reduced wall thickness, is unlikely to represent true regression
of LVH.
Regression of Myocyte Width in the
ob
兾
ob
Mice with CNTF
Ax15
.To
determine whether the observed impact of CNTF
Ax15
on LVH
could be attributed to a regression of myocyte hypertrophy, we
performed histological studies with hematoxylin兾eosin staining
on ob兾ob mouse hearts (n⫽3–5) from each of the three groups
(controls fed ad libitum, calorie restricted, or CNTF
Ax15
treated),
at the end of their treatment period. Myoycte width with
CNTF
Ax15
in the ob兾ob mice was reduced to 9.0 ⫾0.3
␮
m
compared to 11.3 ⫾0.5
␮
m in those fed ad libitum (P⬍0.001),
and, in the calorie-restricted mice, 12.0 ⫾0.4
␮
m(P⬍0.001).
There was no significant difference in cell size between the
calorie restricted and ad libitum groups (Fig. 3 band h). Myocyte
width in age-matched C57bl兾6 wt mice is 8.2 ⫾0.2
␮
m (20).
Regression of LVH in
db
兾
db
Mice with CNTF
Ax15
.We next assessed the
impact of CNTF
Ax15
on LVH in leptin-receptor deficient db兾db
mice. CNTF
Ax15
reduced septal thickness by 27 ⫾6% (P⫽
0.0068), posterior wall thickness by 21 ⫾3% (P⫽0.00019) (Fig.
Table 2. Changes in echocardiographic parameters after intervention in ob兾ob mice
Parameter
Fed ad libitum (n⫽11) CNTF
A⫻15
-treated (n⫽11) Calorie-restricted (n⫽9)
Before After Before After Before After
BW, g 70 ⫾172⫾170⫾243⫾2* 71 ⫾246⫾2*
IVSd, mm 1.11 ⫾0.04 1.15 ⫾0.04 1.13 ⫾0.04 0.90 ⫾0.03* 1.05 ⫾0.04 0.95 ⫾0.04
PWTd, mm 1.02 ⫾0.03 1.12 ⫾0.03 1.10 ⫾0.03 0.86 ⫾0.03* 0.97 ⫾0.02 0.90 ⫾0.04
LVEDd, mm 3.60 ⫾0.14 3.29 ⫾0.10 3.38 ⫾0.09 3.20 ⫾0.09 3.56 ⫾0.16 3.22 ⫾0.08
LVEDs, mm 1.49 ⫾0.10 1.19 ⫾0.06
†
1.50 ⫾0.11 1.12 ⫾0.03
†
1.54 ⫾0.15 1.23 ⫾0.05
†
LVM, mg 149 ⫾7 144 ⫾6 147 ⫾10 106 ⫾14* 135 ⫾5 100 ⫾5*
*,P⬍0.01 vs. preintervention baseline and †, P⬍0.05 vs. preintervention baseline by paired ttests.
Abbreviations are as per Table 1.
Fig. 1. Demonstration of CNTFR
␣
in the heart. (a–f) Immunohistochemistry
with peroxidase staining performed on longitudinal (aand b) and transverse
(cand d) sections of the heart showing the presence of CNTFR
␣
in the heart.
Double immunofluorescence staining for desmin (green) and CNTFR
␣
(red)
localized its presence to the myocyte (e). b,d, and fare negative controls
omitting the primary antibody. (g) Western blot on isolated cardiomyocytes
from C57BL兾6 (wt), ob兾ob and db兾db mice confirming the presence of CNTFR
␣
in the myocyte. Bands corresponding to the 60-kDa glycosylated and 40-kDa
nonglycosylated portions of the receptor can be seen. Mouse brain (mb) was
used as a positive control.
Fig. 2. Regressionof LVH with CNTF in ob兾ob and db兾db mice. (a–g) M-mode
echocardiograms after 4 weeks after intervention of 5–6 months old ob兾ob
mice fed ad libitum (a), treated with CNTFAx15 (b), or calorie-restricted (c) and
db兾db mice fed ad libitum (d), treated with CNTFAx15 (e), calorie restricted ( f),
and treated with leptin (g). Notable findings include reduction in both septal
and posterior wall thickness in the CNTFAx15-treated animals compared with
other groups. (hand i) Graphical demonstration of significant decreases in
posterior wall thickness in diastole (PWTd) in ob兾ob mice (h;*,P⫽0.000022)
and db兾db mice (i;*,P⫽0.029; †, P⫽0.00019) with CNTFAx15 treatment. AL,
ad libitum; LEP, leptin; CR, calorie restriction. Arrows denote endo- and
epicardial borders of heart wall.
Raju et al. PNAS
兩
March 14, 2006
兩
vol. 103
兩
no. 11
兩
4223
MEDICAL SCIENCES

2eand i), and calculated LV mass by 41 ⫾6% (P⫽0.00064),
restoring cardiac architecture toward normal. CNTF
Ax15
also
significantly reduced LV end-systolic and -diastolic dimensions
(P⫽0.042 and 0.013, respectively) (Table 3). Importantly, as
predicted on the basis of an absent leptin receptor, leptin did not
change LV wall thickness or LVM in db兾db mice (Table 3).
However, LVM was decreased in the calorie-restricted group,
but, as discussed earlier, this reduction is most likely attributable
to a decrease in chamber size.
Regression of Myocyte Width in the
db
兾
db
Mice with CNTF
Ax15
.
Assessment of myocyte width in db兾db mice by hematoxylin兾
eosin staining showed that only CNTF
Ax15
regressed toward
normal (9.3 ⫾0.3 microns), in comparison to leptin-treated
(12.6 ⫾0.6 microns; P⬍0.001), calorie-restricted (10.8 ⫾0.4
microns; P⬍0.05) and ad libitum controls (11.4 ⫾0.5 microns;
P⬍0.01) (Fig. 3 eand i).
CNTF
Ax15
Signal Transduction. To determine whether CNTF
Ax15
and leptin activated similar signal transduction pathways in
isolated cardiac myocytes, the activation of STAT3, ERK1兾2 and
JNK pathways, and calcineurin (CaN) signaling was determined.
CNTF
Ax15
and leptin produced near-identical phosphorylation
of STAT3 (Fig. 4a). Similarly, CNTF
Ax15
and leptin both led to
the phosphorylation of ERK1兾2, although the potency of this
effect was greater with CNTF
Ax15
(Fig. 4b). Neither CNTF
Ax15
nor leptin affected JNK phosphorylation or CaN levels in
isolated myocytes (Fig. 5). In terms of chronic 4-week therapy,
the predominant effect was an increase in STAT3 phosphory-
lation and protein expression after CNTF
Ax15
treatment. Al-
though the p-STAT3兾STAT3 ratio was not increased with
CNTF
Ax15
, the overall STAT3 phosphorylation and total STAT3
Table 3. Changes in echocardiographic parameters after intervention in db兾db mice
Parameter
Fed ad libitum (n⫽7) CNTF
A⫻15
-treated (n⫽8) Calorie-restricted (n⫽8) Leptin-treated (n⫽7)
Before After Before After Before After Before After
BW, g 59 ⫾162⫾260⫾243⫾1* 60 ⫾241⫾1* 58 ⫾261⫾2
IVSd, mm 0.94 ⫾0.06 1.08 ⫾0.06 1.11 ⫾0.07 0.79 ⫾0.05* 1.02 ⫾0.02 0.93 ⫾0.06 1.05 ⫾0.06 1.11 ⫾0.02
PWTd, mm 0.90 ⫾0.04 1.04 ⫾0.05
†
1.02 ⫾0.02 0.80 ⫾0.04* 0.94 ⫾0.02 0.94 ⫾0.05 1.02 ⫾0.04 1.08 ⫾1.02
LVEDd, mm 3.29 ⫾0.17 3.33 ⫾0.15 3.65 ⫾0.13 3.31 ⫾0.08* 3.80 ⫾0.12 3.22 ⫾0.08* 3.56 ⫾0.12 3.31 ⫾0.08
LVEDs, mm 1.21 ⫾0.11 1.09 ⫾0.18 1.50 ⫾0.13 1.20 ⫾0.02
†
1.72 ⫾0.14 1.25 ⫾0.05
†
1.53 ⫾0.06 1.23 ⫾0.04*
LVM, mg 107 ⫾11 135 ⫾6 150 ⫾787⫾8* 148 ⫾7 103 ⫾8* 143 ⫾5 137 ⫾4
*,P⬍0.01 vs. preintervention baseline and †, P⬍0.05 vs. preintervention baseline by paired ttests. Abbreviations are as per Table 1.
Fig. 3. CNTFAx15 mediated reduction of myocyte width in ob兾ob and db兾db
mice. (a–g) Histology sections stained with hematoxylin兾eosin from ob兾ob
mice fed ad libitum (a), treated with CNTFAx15 (b), or calorie-restricted (c) and
db兾db mice fed ad libitum (d), treated with CNTFAx15 (e), calorie-restricted ( f),
and treated with leptin (g). (hand i) Graphical demonstration that myocytes
from CNTFAx15-treated ob兾ob (h) and db兾db (i) mice have decreased cell width
compared to the control groups (*,P⬍0.0001). AL, ad libitum; LEP, leptin; CR,
calorie restriction. (Scale bar: 5 microns.) Arrows denote borders of cardiac
myocytes.
Fig. 4. Leptin and CNTFAx15 activate STAT3 and ERK pathways. (a) Western
blot analysis demonstrates time-dependent STAT3 phosphorylation in iso-
lated adult mouse cardiomyocytes after exposure to leptin (
■
;50ng兾
␮
l) or
CNTFAx15. (
E
;50ng兾
␮
l). STAT3 phosphorylation occurred within 15 min and
subsided over time. (b) Similar activation occurred in the ERK1兾2 pathway,
with the effect of CNTFAx15 being more prolonged. [*,P⬍0.05; †, P⬍0.01; ‡,
P⬍0.001 vs. time 0; §, P⬍0.05 between groups (two-way ANOVA).] (cand d)
Immunoblots from ob兾ob mice (filled bars) and db兾db mice (hatched bars)
demonstrate increased abundance in both phosphorylated STAT3 and STAT3
expression with 4 weeks of CNTFAx15 (CNTF) treatment. Leptin (LEP) increased
STAT3 abundance in ob兾ob but not db兾db mice. CR, calorie restriction. [*,P⬍
0.05; †, P⬍0.01 vs. ad libitum (AL).]
4224
兩
www.pnas.org兾cgi兾doi兾10.1073兾pnas.0510460103 Raju et al.

abundance were higher in cardiac tissue from Axokine-treated
ob兾ob (Fig. 4c) and db兾db mice (Fig. 4d). CNTF
Ax15
treatment
resulted in a similar effect on the ERK1兾2 pathway (Fig. 6 aand
b). As expected, leptin did not produce this effect in db兾db mice
lacking the leptin receptor. Consistent with the isolated cardi-
omyocyte data, JNK phosphorylation and CaN protein level
were not increased in the cardiac tissue. However, total JNK
protein expression was augmented (Fig. 6 cand d).
Discussion
The major finding of this study is that the CNTF signaling pathway,
which is known to activate a parallel signal transduction pathway to
leptin in the hypothalamus, has important cardiac bioactivity. The
CNTF receptor is present in the cardiac sarcolemma, and its
activation, like leptin (7), regresses established LVH in leptin-
deficient mice. Importantly, CNTF
Ax15
signaling has similar cardiac
effects in mice with leptin receptor dysfunction. CNTF
Ax15
influ-
ences signal transduction in isolated cardiac myocytes supporting a
role for direct effects on the heart, activating both the phosphor-
ylation of STAT3 and ERK1兾2. These findings have therapeutic
implications given that the majority of human obesity is due to
leptin resistance rather than leptin deficiency (4, 8, 9).
The role of leptin or its deficiency in mediating cardiac
hypertrophy is controversial. We (7) have previously demon-
strated that leptin deficiency contributes to LVH in vivo in
intact ob兾ob mice with established LVH. On the other hand,
in normal neonatal myocytes, exposure to leptin in vitro
stimulates growth or hypertrophy (21, 22). However, these
findings are not incompatible with each other, as the obser ved
dichotomy could easily arise because of differences in milieu.
For example, neonatal cells physiologically undergo rapid
growth under the inf luence of normal trophic factors, unlike
adult myocytes. Syed et al. (23) recently provided proof of
principle for this concept and showed that although certain
genes produced changes in ventricular structure and function
in neonatal cells, they had no impact on the adult myocardium.
These findings suggest that physiologic differences in charac-
teristics between neonatal and adult myocytes play an impor-
tant role in their response to external stimuli and the net result
may even be the opposite (23).
CNTF
Ax15
and leptin stimulated the phosphorylation of
STAT3 acutely in cardiac myocytes and augmented total STAT3
levels in association with regression of hypertrophy in the ob兾ob
and db兾db mice. STAT3 phosphorylation is attributed with a
prohypertrophic effect in neonatal myocytes (24, 25) but also a
cardioprotective antiremodeling effect (26, 27). The present data
support the dominance of the cardioprotective effects in this
setting. Importantly, these results also demonstrate that CNT-
F
Ax15
and leptin directly activate similar signal transduction
cascades within cardiac myocytes.
Recent epidemiologic studies support our paradigm as well.
In a study conducted in healthy individuals free from cardio-
vascular disease, Pladevall et al. (28) showed that leptin
deficiency was associated with increased left ventricular mass
index, thus demonstrating that leptin has an antihypertrophic
effect in the presence of an intact signaling pathway. Although
some older studies showed a positive correlation between
leptin and LVH when adjusting for body mass index (BMI)
(29), others have highlighted that this positive correlation is
completely abrogated after adjusting for BMI (30). Nonethe-
less, it is well established that obesity in clinical populations is
associated with elevated leptin primarily due to leptin resis-
tance, implying depressed downstream signaling despite in-
creased leptin levels (4, 8, 9). Our current results further
substantiate that like leptin, the CNTF signaling pathways
regress established LVH, probably by restoring downstream
leptin or leptin-like signaling, and offer additional support to
the paradigm that leptin signaling deficiency contributes to the
development of LVH in obesity in vivo.
Several points warrant mention. The potential mechanism of
action of CNTF
Ax15
is complex. Although we have demonstrated
direct actions within the heart (such as STAT3 and ERK1兾2
activation), we cannot exclude the possibility of systemic effects.
For example, central nervous system effects of CNTF
Ax15
may
directly influence the heart. Indeed, the leptin兾leptin-receptor
axis is known to activate the central ner vous system (31, 32) and
to mediate metabolic actions by acting centrally. Additionally, it
is possible that the receptors are not present solely on cardiac
myocytes but also on other tissues, such as cardiac ganglia,
implying the possibility of paracrine signaling effects within the
Fig. 5. Impact of leptin and CNTFAx15 on pJNK and CaN signaling in cultured
cardiomyocytes, ob兾ob, and db兾db mice. (aand b) Western blot analysis shows
no induction of JNK phosphorylation (a) or altered calcineurin (CaN) protein
expression (b) in isolated adult mouse cardiomyocytes with leptin (Upper)or
CNTFAx15 (Lower) treatment.
Fig. 6. Impact of CNTFAx15 on ERK and JNK signal transduction pathways in
ob兾ob and db兾db mice. (aand b) Western blot analysis depicts increased ERK
phosphorylation and ERK protein expression in cardiac tissue of ob兾ob (filled
bars) and db兾db mice (hatched bars) with CNTFAx15 (CNTF) treatment but not
with leptin (LEP) in db兾db mice. (cand d) CNTFAx15 increased JNK protein
abundance in heart tissue in ob兾ob and db兾db mice. However, calorie restric-
tion (CR) does not affect JNK protein expression in cardiac tissue of ob兾ob and
db兾db mice. (*,P⬍0.05.)
Raju et al. PNAS
兩
March 14, 2006
兩
vol. 103
兩
no. 11
兩
4225
MEDICAL SCIENCES

heart. Nonetheless, our finding that CNTF
Ax15
causes a regres-
sion of established LVH not only in leptin-deficient animals but
also in those lacking a functional leptin receptor in the context
of stimulating signal transduction within cardiac myocytes es-
tablishes the existence of a previously unrecognized pathway in
the regulation of LVH.
Our findings identify a cardiac signal transduction pathway in
obesity related cardiac hypertrophy. Because the majority of
human obesity is associated with leptin resistance, further eval-
uation of CNTF
Ax15
for potential therapeutic applications in the
treatment of LVH in patients with leptin resistance may be
fruitful.
Materials and Methods
Animals. We used ob兾ob,db兾db, and C57BL兾6 WT mice obtained
from The Jackson Laboratory. Animal treatment and care was
provided in accordance with institutional guidelines. All animals
were housed under diurnal lighting conditions and allowed food
and tap water ad libitum except the calorie-restricted groups,
where each mouse received1goffood per day as described
below. The Animal Care and Use Committee of The Johns
Hopkins University approved all protocols.
Isolated Myocyte Preparation. Cardiac myocytes were isolated and
prepared from mouse hearts as described in detail by Khan et al.
(33). Myocytes were plated with 80% confluency into 35-mm
dishes (Falcon) containing a serum-free 500
␮
MCa
2⫹
contain-
ing isolation solution for 30 min. Myocytes were then treated
with 50 ng兾
␮
l leptin and 50 ng兾
␮
l CNTF
Ax15
for various periods
of time.
Treatment by Administration of Exogenous Leptin, CNTF
Ax15
, and
Calorie Restriction. Five- to 6-month-old ob兾ob and db兾db mice
were randomly assigned to three or four groups, respectively,
and treated for a period of 4 weeks. The first group received
CNTF
Ax15
, the second group was calorie-restricted so as to lose
similar weight as treated animals, and the third group was fed
ad libitum as controls. Additional db兾db mice were assigned to
a fourth group to receive leptin. Mice in the CNTF
Ax15
group
were injected daily with recombinant mouse CNTF
Ax15
(Re-
generon Pharmaceuticals, Tarrytown, NY, 0.1 mg䡠kg
⫺1
per
day), whereas the leptin group were injected with recombinant
mouse leptin (R & D Systems, 1 mg䡠kg
⫺1
per day). The calorie
restriction regimen consisted of1goffood per day, which was
previously determined to result in a similar rate of weight loss
to leptin or CNTF
Ax15
administration at these doses (7). The
calorie-restricted groups and the ad libitum controls were
injected with Dulbecco’s PBS.
Echocardiography. To examine the impact of CNTF
Ax15
and leptin
on established LVH (Table 1) in leptin-deficient ob兾ob mice (7),
we performed echocardiograms on mice randomized to one of
three treatment groups: CNTF
Ax15
(n⫽11, 5–6 months old,
weight: 70 ⫾2 g), calorie restriction (n⫽9, 5–6 months old,
weight: 71 ⫾2 g), and controls fed ad libitum (n⫽11, 5–6
months old, weight: 70 ⫾1 g). We also investigated the effects
of CNTF
Ax15
(n⫽8, 5–6 months old, weight: 60 ⫾2 g), calorie
restriction (n⫽8, 5–6 months old, weight: 60 ⫾2 g), leptin (n⫽
7, 5–6 months old, weight: 58 ⫾2 g), and controls fed ad libitum
(n⫽7, 5–6 months old, weight: 59 ⫾1 g) on LVH after a 4-week
treatment period in db兾db mice. Echocardiography was per-
formed on conscious unanesthetized mice at baseline and re-
peated at the end of the treatment period. The individual
performing the echocardiograms was blinded to the group
assignments. Mice were trained before each study until they were
relaxed for the procedure. Studies were performed by using a
Sequoia C256 (Siemens, Mountain View, CA) echocardiogram
with a 15 MHz linear array transducer. Interventricular septal
and posterior wall thicknesses, as well as diastolic and systolic LV
dimensions were recorded from M-mode images by using aver-
aged measurements from three to five consecutive cardiac cycles.
LVM was calculated by using LVM ⫽1.055 ⫻(interventricular
septal thickness in diastole thickness ⫹posterior wall thickness
⫹left ventricular end diastolic diameter)
3
⫺(left ventricular end
diastolic diameter)
3
.
Histology. Mice hearts were harvested, washed with ice-cold
PBS, sectioned, and either fixed in 10% formalin or in Streck’s
tissue fixative (Streck Laboratories, Omaha, NE) overnight
and were paraffin embedded the next morning. Histochemical
and immunohistochemical studies were carried out on 4-
␮
m-
thick sections.
Histochemical Staining. Hematoxylin兾eosin staining were per-
formed to evaluate cell sizes. Myocyte diameters were measured
in regions of myocardium with parallel myocyte fascicles in
longitudinal sections by using IMAGEJ (National Institutes of
Health). Representative high-powered-fields distributed around
the myocardium were used to measure 7–29 cells from each
heart. The individual analyzing histology data were blinded to
the group assignments.
Immunohistochemistry. For antigen retrieval, the sections were
steamed with Target Retrieval System (DAKO) for 20 min,
before blocking with avidin and biotin blocking solutions
(DAKO) and 10% normal goat serum. The receptor was stained
with a monoclonal mouse anti-CNTFR-
␣
(BD Pharmingen, no.
558783; 1:100 dilution). After washing, the secondary antibody
(biotinylated goat anti-mouse Ig; E0433, DAKO; 1:400 dilution)
and streptavidin peroxidase (Vectastain, Vector Laboratories)
were applied.
The peroxidase activity was developed 5 min with 0.066%
diaminobenzidine兾0.01% H
2
O
2
兾2.5% NiSO
4
. Slides were
cleared in xylene, coverslipped, and viewed on a Zeiss Axiovert
200 microscope equipped with 510-Meta confocal laser scanning
module.
Fluorescence. Tissue section were rinsed with ice-cold PBS twice
for 2 min and were blocked with 10% goat serum for 30 min. The
slides were incubated with rabbit anti-human desmin antibody
(Accurate Chemical & Scientific, no. YMPS31) and monoclonal
mouse anti-CNTFR-
␣
antibody (BD Pharmingen, no. 558783) at
4°C overnight. After washing, FITC and rhodamine coupled
secondary antibodies (Vector Laboratories) were applied for 45
min at room temperature. Tissue slide were washed and cover-
slipped with mounting medium and viewed.
Western Blots. Hearts from ob兾ob,db兾db, and WT mice were
harvested after cer vical dislocation and rinsed in isolation
buffer to remove excess blood. Cells were isolated as described
by Minhas et al. (34). Isolated myocytes were homogenized in
cold cell lysis buffer (Cell Signaling Technology, Beverly, MA)
with one tablet of complete protease inhibitors (Roche Diag-
nostics) per 10 ml of buffer. The protein lysate was quantitated
with the bicinchonic assay (Pierce), and 30 –50
␮
g of total
protein was separated in 4–12% Bis-Tris䡠HCl polyacrylamide
gels (Invitrogen). The proteins were transferred onto nitro-
cellulose membranes and blocked for 1 h with 5% nonfat milk.
After blocking, the membranes were incubated with mono-
clonal anti-CNTFR
␣
antibody (1:1,000, BD Pharmingen),
polyclonal anti-phospho-STAT3 antibody, polyclonal anti-
phospho-ERK1兾2 antibodies, polyclonal anti-phospho-JNK
antibodies (all 1:1,000, Cell Signaling Technology), or mono-
clonal anti-calcineurin antibodies (1:2,000, Chemicon Inter-
national) overnight at 4°C. Immunoblots were detected by
using enhanced chemiluminescent kits (SuperSignal, Pierce)
4226
兩
www.pnas.org兾cgi兾doi兾10.1073兾pnas.0510460103 Raju et al.

and analyzed with a densitometer (Bio-Rad). The membranes
were then stripped and reprobed with polyclonal anti-STAT3
antibody, polyclonal anti-ERK1兾2 antibodies, polyclonal anti-
JNK antibodies (all 1:1,000, Cell Signaling Technology), or
monoclonal anti-GAPDH antibodies (1:10,000, Research Di-
agnostics). Mouse brain was used as a positive control for
calcineurin.
Sample Size and Statistical Analysis. Sample size was estimated as
five to seven animals in each group for statistical power ⬎0.80
to determine regression of hypertrophy and LVM. Data are
reported as mean ⫾SEM. Statistical significance was deter-
mined by paired ttests (GraphPad (San Diego) INSTAT statistical
software), or by two-way ANOVA with Bonferroni correction
[STATA (StataCorp LP, College Station, TX) statistical soft-
ware]. Pvalues ⬍0.05 were considered significant.
This work was supported in part by the Donald W. Reynolds Foun-
dation (L.A.B. and J.M.H.), National Institutes of Health Grants
R01 HL-65455 and National Institute on Aging Grant R01AG025017
(to J.M.H.), National Institutes of Health Grant K08 Hl 076220-01 (to
L.A.B.), and the Talles Family Fund for Cardiomyopathy Research (to
L.A.B.). CNTF
Ax15
(Axokine) was generously provided by Regeneron
Pharmaceuticals, Tarrytown, NY, through a materials transfer agree-
ment with The Johns Hopkins University.
1. Kaplinsky, E. (1994) Cardiovasc. Drugs Ther. 8, Suppl. 3, 549–556.
2. Eckel, R. H., Barouch, W. W. & Ershow, A. G. (2002) Circulation 105,
2923–2928.
3. Bray, G. A. & York, D. A. (1997) J. Clin. Endocrinol. Metab. 82, 2771–2776.
4. Considine, R. V., Sinha, M. K., Heiman, M. L., Kriauciunas, A., Stephens,
T. W., Nyce, M. R., Ohannesian, J. P., Marco, C. C., McKee, L. J., Bauer, T. L.,
et al. (1996) N. Engl. J. Med. 334, 292–295.
5. Montague, C. T., Farooqi, I. S., Whitehead, J. P., Soos, M. A., Rau, H.,
Wareham, N. J., Sewter, C. P., Digby, J. E., Mohammed, S. N., Hurst, J. A., et
al. (1997) Nature 387, 903–908.
6. Sweeney, G. (2002) Cell. Signalling 14, 655–663.
7. Barouch, L. A., Berkowitz, D. E., Harrison, R. W., O’Donnell, C. P. & Hare,
J. M. (2003) Circulation 108, 754–759.
8. Scarpace, P. J., Matheny, M. & Tumer, N. (2001) Neuroscience 104, 1111–1117.
9. Scarpace, P. J. & Tumer N. (2001) Physiol. Behav. 74, 721–727.
10. Sleeman, M. W., Garcia, K., Liu, R., Murray, J. D., Malinova, L., Moncrieffe,
M., Yancopoulos, G. D. & Wiegand, S. J. (2003) Proc. Natl. Acad . Sci. USA 100,
14297–14302.
11. Gloaguen, I., Costa, P., Demartis, A., Lazzaro, D., Di Marco, A., Graziani, R.,
Paonessa, G., Chen, F., Rosenblum, C. I., Van der Ploeg, L. H., et al. (1997)
Proc. Natl. Acad. Sci. USA 94, 6456– 6461.
12. Lambert, P. D., Anderson, K. D., Sleeman, M. W., Wong, V., Tan, J.,
Hijarunguru, A., Corcoran, T. L., Murray, J. D., Thabet, K. E., Yancopoulos,
G. D., et al. (2001) Proc. Natl. Acad. Sci. USA 98, 4652–4657.
13. Elias, C. F., Aschkenasi, C., Lee, C., Kelly, J., Ahima, R. S., Bjorbaek, C., Flier,
J. S., Saper, C. B. & Elmquist, J. K. (1999) Neuron 23, 775–786.
14. Elmquist, J. K., Elias, C. F. & Saper, C. B. (1999) Neuron 22, 221–232.
15. Sleeman, M. W., Anderson, K. D., Lambert, P. D., Yancopoulos, G. D. &
Wiegand, S. J. (2000) Pharm. Acta Helv. 74, 265–272.
16. Davis, S., Aldrich, T. H., Stahl, N., Pan, L., Taga, T., Kishimoto, T., Ip, N. Y.
& Yancopoulos, G. D. (1993) Science 260, 1805–1808.
17. Boulton, T. G., Stahl, N. & Yancopoulos, G. D. (1994) J. Biol. Chem. 269,
11648–11655.
18. Bjorbaek, C., Elmquist, J. K., Frantz, J. D., Shoelson, S. E. & Flier, J. S. (1998)
Mol. Cell 1, 619–625.
19. Ziotopoulou, M., Erani, D. M., Hileman, S. M., Bjorbaek, C. & Mantzoros,
C. S. (2000) Diabetes 49, 1890–1896.
20. Barouch, L. A., Cappola, T. P., Harrison, R. W., Crone, J. K., Rodriguez, E. R.,
Burnett, A. L. & Hare, J. M. (2003) J. Mol. Cell. Cardiol. 35, 637– 644.
21. Rajapurohitam, V., Gan, X. T., Kirshenbaum, L. A. & Karmazyn, M. (2003)
Circ. Res. 93, 277–279.
22. Xu, F. P., Chen, M. S., Wang, Y. Z., Yi, Q., Lin, S. B., Chen, A. F. & Luo, J. D.
(2004) Circulation 110, 1269–1275.
23. Syed, F., Odley, A., Hahn, H. S., Brunskill, E. W., Lynch, R. A., Marreez, Y.,
Sanbe, A., Robbins, J. & Dorn, G. W. (2004) Circ. Res. 95, 1200–1206.
24. Kunisada, K., Tone, E., Fujio, Y., Matsui, H., Yamauchi-Takihara, K. &
Kishimoto, T. (1998) Circulation 98, 346–352.
25. Pan, J., Fukuda, K., Kodama, H., Sano, M., Tak ahashi, T., Makino, S., Kato,
T., Manabe, T., Hori, S. & Ogawa, S. (1998) Heart Vessels 13, 199–208.
26. Harada, M., Qin, Y. J., Takano, H., Minamino, T., Zou, Y. Z., Toko, H.,
Ohtsuka, M., Matsuura, K., Sano, M., Nishi, J., et al. (2005) Nat. Med. 11,
305–311.
27. Oshima, Y., Fujio, Y., Nakanishi, T., Itoh, N., Yamamoto, Y., Negoro, S.,
Tanaka, K., Kishimoto, T., Kawase, I. & Azuma, J. (2005) Cardiovasc. Res. 65,
428– 435.
28. Pladevall, M., Williams, K., Guyer, H., Sadurni, J., Falces, C., Ribes, A., Pare,
C., Brotons, C., Gabriel, R., Serrano-Rios, M., et al. (2003) J. Hypertens. 21,
1467–1473.
29. Paolisso, G., Tagliamonte, M. R., Galderisi, M., Zito, G. A., Petrocelli, A.,
Carella, C., de Divitiis, O. & Varricchio, M. (1999) Hypertension 34, 1047–1052.
30. Tritos, N. A., Kissinger, K. V., Mann ing, W. J. & Danias, P. G. (2004) Clin.
Endocrinol. 60, 60 –66.
31. Schwartz, M. W., Woods, S. C., Porte, D., Jr., Seeley, R. J. & Baskin, D. G.
(2000) Nature 404, 661–671.
32. Haynes, W. G., Sivitz, W. I., Morgan, D. A., Walsh, S. A. & Mark, A. L. (1997)
Hypertension 30, 619–623.
33. Khan, S. A., Skaf, M. W., Harrison, R. W., Lee, K., Minhas, K. M., Kumar, A.,
Fradley, M., Shoukas, A. A., Berkowitz, D. E. & Hare, J. M. (2003) Circ. Res.
92, 1322–1329.
34. Minhas, K. M., Khan, S. A., Raju, S. V., Phan, A. C., Gonzalez, D. R., Skaf,
M. W., Lee, K., Tejani, A. D., Barouch, L. A., O’Donnell, C. P., et al. (2005)
J. Physiol. 565, 463–474, and erratum (2005) 566, 999.
Raju et al. PNAS
兩
March 14, 2006
兩
vol. 103
兩
no. 11
兩
4227
MEDICAL SCIENCES