FGF23 is a novel regulator of intracellular calcium and cardiac contractility in addition to cardiac hypertrophy

Article (PDF Available)inAJP Endocrinology and Metabolism 304(8) · February 2013with177 Reads
DOI: 10.1152/ajpendo.00596.2012 · Source: PubMed
Abstract
Fibroblast growth factor 23 (FGF23) is a hormone primarily released by osteocytes that regulates phosphate and vitamin D metabolism. Recent observational studies in humans suggest that circulating FGF23 is independently associated with cardiac hypertrophy and increased mortality, but it is unknown if FGF23 can directly alter cardiac function. We found that FGF23 significantly increased cardiomyocyte cell size in vitro, the expression of gene markers of cardiac hypertrophy, and total protein content of cardiac muscle. In addition, FGFR1 and FGFR3 mRNA were the most abundantly expressed FGF receptors in cardiomyocytes and the co-receptor α-Klotho was expressed at very low levels. We tested an animal model of chronic kidney disease (Col4a3(-/-) mice) which has elevated serum FGF23. We found elevations in common hypertrophy gene markers in Col4a3(-/-) hearts compared to wild type, but did not observe changes in wall thickness or cell size by week 10. However, the Col4a3(-/-) hearts did show reduced fractional shortening (-17%) and ejection fraction (-11%). Acute exposure of primary cardiomyocytes to FGF23 resulted in elevated intracellular Ca(2+) ([Ca2+]i) (F/Fo +86%) which was blocked by verapamil pretreatment. FGF23 also increased ventricular muscle strip cardiac contractility (67%) which was inhibited by FGF receptor antagonism. We hypothesize that while FGF23 can acutely increase [Ca(2+)]i , chronically this may lead to decreases in contractile function or stimulate cardiac hypertrophy as observed with other stress hormones. In conclusion, FGF23 is a novel bone-heart endocrine factor and may be an important mediator of cardiac Ca2+ regulation and contractile function during chronic kidney disease.
FGF23 is a novel regulator of intracellular calcium and cardiac contractility in
addition to cardiac hypertrophy
Chad D. Touchberry,
1
Troy M. Green,
1
Vladimir Tchikrizov,
1
Jaimee E. Mannix,
1
Tiffany F. Mao,
1
Brandon W. Carney,
1
Magdy Girgis,
5
Robert J. Vincent,
5
Lori A. Wetmore,
3
Buddhadeb Dawn,
5
Lynda F. Bonewald,
2
Jason R. Stubbs,
4
and Michael J. Wacker
1
1
Muscle Biology Group, School of Medicine, University of Missouri-Kansas City, Kansas City, Missouri;
2
Bone Biology
Group, School of Dentistry, University of Missouri-Kansas City, Kansas City, Missouri;
3
Department of Chemistry, William
Jewell College, Liberty, Missouri;
4
The Kidney Institute, University of Kansas Medical Center, Kansas City, Kansas;
and
5
Division of Cardiovascular Diseases, University of Kansas Medical Center, Kansas City, Kansas
Submitted 28 November 2012; accepted in final form 25 February 2013
Touchberry CD, Green TM, Tchikrizov V, Mannix JE, Mao TF,
Carney BW, Girgis M, Vincent RJ, Wetmore LA, Dawn B, Bone-
wald LF, Stubbs JR, Wacker MJ. FGF23 is a novel regulator of
intracellular calcium and cardiac contractility in addition to cardiac
hypertrophy. Am J Physiol Endocrinol Metab 304: E863–E873, 2013.
First published February 26, 2013; doi:10.1152/ajpendo.00596.2012.—
Fibroblast growth factor 23 (FGF23) is a hormone released primarily
by osteocytes that regulates phosphate and vitamin D metabolism.
Recent observational studies in humans suggest that circulating
FGF23 is independently associated with cardiac hypertrophy and
increased mortality, but it is unknown whether FGF23 can directly
alter cardiac function. We found that FGF23 significantly increased
cardiomyocyte cell size in vitro, the expression of gene markers of
cardiac hypertrophy, and total protein content of cardiac muscle. In
addition, FGFR1 and FGFR3 mRNA were the most abundantly
expressed FGF receptors in cardiomyocytes, and the coreceptor
-klotho was expressed at very low levels. We tested an animal model
of chronic kidney disease (Col4a3
/
mice) that has elevated serum
FGF23. We found elevations in common hypertrophy gene markers in
Col4a3
/
hearts compared with wild type but did not observe
changes in wall thickness or cell size by week 10. However, the
Col4a3
/
hearts did show reduced fractional shortening (17%) and
ejection fraction (11%). Acute exposure of primary cardiomyocytes
to FGF23 resulted in elevated intracellular Ca
2
([Ca
2
]
i
; F/F
o
86%) which was blocked by verapamil pretreatment. FGF23 also
increased ventricular muscle strip contractility (67%), which was
inhibited by FGF receptor antagonism. We hypothesize that although
FGF23 can acutely increase [Ca
2
]
i
, chronically this may lead to
decreases in contractile function or stimulate cardiac hypertrophy, as
observed with other stress hormones. In conclusion, FGF23 is a novel
bone/heart endocrine factor and may be an important mediator of
cardiac Ca
2
regulation and contractile function during chronic kid-
ney disease.
fibroblast growth factor 23; pathological cardiac hypertrophy; chronic
kidney disease; Col4a3; cardiac function; -klotho
FIBROBLAST GROWTH FACTOR 23 (FGF23) is a hormone released
primarily by osteocytes (2, 4, 14, 38) that functions to regulate
phosphate and vitamin D homeostasis through direct actions on
the kidney and parathyroid (2). Although an endocrine axis has
been established between bone and kidney, a new paradigm is
emerging in which FGF23 could be important in establishing
an endocrine axis between bone and heart. Circulating levels of
FGF23 are markedly elevated 100- to 1,000-fold in patients
with chronic kidney disease (CKD) (24, 31) and are indepen-
dently associated with cardiovascular morbidity and mortality
(8, 22, 26, 34, 35, 48, 52). Specifically, an association between
left ventricular (LV) hypertrophy and serum FGF23 levels has
been established in CKD patients (20, 36, 52).
Nevertheless, despite strong associations between FGF23
and adverse outcomes, it remains relatively unknown whether
FGF23 is simply a marker of cardiac disease risk or a direct
mediator of cardiac pathology and cardiac performance. Only
one study to date has analyzed the direct effects of FGF23 on
the heart both in vitro and in vivo (13). This important work by
Faul et al. (13) shows that FGF23 can directly induce hyper-
trophy in isolated neonatal cardiomyocytes as well as with
intramyocardial FGF23 injections. These authors also demon-
strated that a FGF receptor (FGFR) antagonist reduced the LV
hypertrophy in a 5/6 nephrectomy rat model of CKD. These
findings are significant in that FGF23 may be an important
player in directly inducing cardiac hypertrophy during CKD.
Moving forward, our laboratory has explored and addressed
crucial questions that require answers, such as whether the
hypertrophic effects of FGF23 can also be replicated in adult
cardiomyocytes in vitro and observed in another animal model
of CKD (Col4a3
/
) that has elevated serum FGF23. Also, can
FGF23 directly alter intracellular Ca
2
([Ca
2
]
i
) and cardiac
contractility? Previous studies have found a clinical association
between FGF23 and LV mass (22, 35, 52) as well as declines
in cardiac performance as measured by reduction in ejection
fraction (19). However, to date no investigation has determined
whether FGF23 can alter cardiac function independent of
changes in cardiac hypertrophy. Determining what direct ef-
fects FGF23 may have on the heart will not only reveal how
FGF23 may alter cardiac function on a beat-to-beat basis but
also yield insights into how this hormone may induce chronic
pathologies such as hypertrophy or heart failure.
MATERIALS AND METHODS
Materials. Recombinant mouse FGF23 was purchased fromR&D
Systems (Minneapolis, MN). The FGFR1 inhibitor PD-166866 was
purchased from EMD biosciences (San Diego, CA). Organ baths,
stimulating electrodes, and LabChart 6 software were obtained from
AD Instruments (Colorado Springs, CO). The stimulation unit (SD9)
was purchased from Grass Technologies (Quincy, MA). Hanks’
balanced salt solution (HBSS) and Flou-4 AM were obtained from
Invitrogen (Carlsbad, CA). Enzymes for cardiomyocyte isolation were
obtained from Worthington (Lakewood, NJ). Total RNA isolation kits
Address for reprint requests and other correspondence: M. J. Wacker,
School of Medicine, Univ. of Missouri-Kansas City, 2464 Charlotte St.,
Kansas City, MO 64108 (e-mail: wackerm@umkc.edu).
Am J Physiol Endocrinol Metab 304: E863–E873, 2013.
First published February 26, 2013; doi:10.1152/ajpendo.00596.2012.
0193-1849/13 Copyright
©
2013 the American Physiological Societyhttp://www.ajpendo.org E863
at UMKC Libraries on April 29, 2013http://ajpendo.physiology.org/Downloaded from
were purchased from IBI scientific (Peosta, IA), and the real-time
reverse-transcriptase polymerase chain reaction (RT-PCR) was per-
formed using a TaqMan RNA-to-CT 1 step kit from ABI (Carlsbad,
CA). -Tubulin and phospho-ERK1/2 antibodies were purchased
from Cell Signaling Technology (Danvers, MA). NCX1 primary
antibodies were purchased from Swant (Marly, Switzerland). Clay-
comb’s medium and fetal bovine serum were purchased from Sigma-
Aldrich (St. Louis, MO). All remaining reagents were purchased from
Fisher Scientific (Pittsburgh, PA).
Experimental animals. Twelve-week-old wild-type (WT) male
CD1 mice (Harlan Laboratories, Madison, WI) were used in experi-
ments using exogenous FGF23. In addition, male and female 10-wk-
old Col4a3
/
mice (background SV129) and age/liter-matched WT
mice were also used in this study. The Col4a3
/
mice are a model of
human autosomal-recessive Alport syndrome and can have serum
levels of FGF23 5,000 pg/ml by 12 wk (46). The Col4a3
/
mice
develop nonhypertensive progressive renal fibrosis (17, 18). By 12
wk, Col4a3
/
mice present with significant elevations in serum
phosphorus, parathyroid hormone, blood urea nitrogen, and creatine
compared with WT mice (46). In addition, at week 12, Col4a3
/
mice have significantly reduced serum calcium and 1,25-dihydroxyvi-
tamin D (46). All mice were housed in a temperature-controlled (22
2°C) room with a 12:12-h light-dark cycle. Animals were fed ad
libitum. All protocols were approved by the Animal Care and Use
Committee of the University Missouri-Kansas City School of Medi-
cine and the University of Kansas Medical Center.
HL-1 cell culture. HL-1 cardiomyocytes were plated (5,000/cm
2
)in
flasks precoated with 0.00125% fibronectin and 0.02% gelatin. Cells
were cultured for 24 h in Claycomb’s media (supplemented with 10%
FBS, 2 mM
L-glutamine, 0.1 mM norepinephrine, 0.3 mM ascorbic
acid, 100 U/ml penicillin, and 100 mg/ml streptomycin), as described
previously (50, 51). Prior to experiments, cells were rendered quies-
cent in a minimal media (0.5% FBS, 2 mM
L-glutamine, and penicil-
lin-streptomycin, without norepinephrine) for 48 h prior to treatment.
Cells were treated with vehicle, FGF23, and FGF23 PD-166866 (50
nM) for 48 h prior to analysis. Cells were collected and analyzed for
changes in cells size by flow cytometry using FACSCalibur (FSC).
Isolation of primary cardiac myocytes. Following cervical disloca-
tion the heart was rapidly excised, extraneous tissue was removed, and
the aorta was cannulated under a dissection scope. Cardiomyocytes
were isolated in a standard manner, utilizing retrograde perfusion via
a proprietary procedure developed in our laboratory with Worthington
Biochemical (Lakewood, NJ). Briefly, hearts were retrograde perfused
through the aorta using a Langendorff perfusion apparatus with
Ca
2
-free perfusion buffer (3 ml/min) for 4 min and then switched to
a digestion buffer containing collagenase II (18,000 U), papain (20 U),
and DNase (2,000 U) for 8 –10 min at 37°C. The heart was removed
from perfusion, cut into pieces, and pipetted gently to disperse cells in
suspension. Calcium-tolerant myocytes were then plated into dishes
previously coated with 10 g/ml laminin and kept in L-15 medium
with blebbistatin (25 M) (27).
Gene expression. Total RNA from isolated cardiomyocytes, FGF23,
or vehicle-treated cardiac tissue cultures and Col4a3
/
and WT mice
were extracted, and real-time RT-PCR was performed. Gene expression
from tissue culture was normalized (2
⫺⌬⌬C
T
analysis) to -actin (54).
Gene expression for isolated cardiac myocytes was performed using
2
⫺⌬⌬C
T
analysis against -actin and then normalized to -klotho expres
-
sion. For WT and Col4a3
/
mouse gene expression, 2
⫺⌬C
T
analysis
was performed, using -actin as a standard, and then averages of each
gene of the WT mice were calculated. Each individual animal was then
normalized to the average of the WTs. -Actin was chosen as a house-
keeping gene because FGF23 treatment and Col4a3
/
mouse models
had minimal changes in gene expression compared with vehicle-treated
and WT mice, respectively. GAPDH comparisons also yielded similar
results as that with -actin.
Tissue culture. CD1 mouse ventricular tissue strips were isolated
from mouse hearts and used for the tissue culture experiments. Hearts
were quickly excised and placed into an ice-cold cardioprotective
medium that included the addition of 2,3-butanedione monoxime (30
mM), as described previously (50). Tissue cultures were treated with
vehicle, FGF23, or FGF23 PD-166866 (50 nM).
Total protein and Western blot. Clamp-frozen ventricular tissues
were weighed and homogenized in a 12:1 (vol/wt) ratio of ice-cold
cell extraction buffer (Invitrogen), as described previously (56). Total
protein concentration of the samples was determined by use of the
microbicinchoninic acid protein assay (Pierce Chemical) and then
normalized to tissue weight (g/mg). Protein samples (20 –50 g)
were run on 4 –20% SDS-PAGE gels, and Western blots were per-
formed using standard techniques. Because downstream changes in
the signaling pathway are reliant on a net increase in ERK phosphor-
ylation and total ERK protein expression is unlikely to change over 15
min, p-ERK blots were normalized to -tubulin as a loading control.
Whole heart lysates from WT and Col4a3
/
mice were used for
analysis of NCX1 via Western blotting and normalized to -tubulin.
Echocardiogragphy on Col4a3
/
mice. Mice were weighed and
anesthetized with isoflurane inhalation (3% for induction, 1% for
maintenance). The anterior chest was shaved, and the mice were
placed on a heating pad in the left lateral decubitus position. A rectal
temperature probe was placed to ensure that the body temperature
remained at 37.0°C during the study. Left ventricle structure and
function were assessed by previously validated two-dimensional M-
mode and Doppler echocardiographic techniques (9, 61). Echocardio-
graphic images were obtained using a Philips HDI 5000 SonoCT
ultrasound system equipped with a 12–5 MHz phased-array probe
fitted with a 0.3-cm standoff and a 15–7 MHz broadband linear probe.
Digital images were analyzed offline according to modified American
Society for Echocardiography standards (42) using the ProSolv image
analysis software (version 3.5; Problem Solving Concepts). LV end-
diastolic and end-systolic diameters and anterior and posterior wall
thickness in diastole were measured from M-mode tracings obtained
at the midpapillary level (41). LV ejection fraction was derived from
M-mode parameters. LV mass was estimated from the M-mode data,
and LV end-diastolic and end-systolic volumes were calculated using
the formula of Teichholz et al. (49). Analysis of data was performed
by an investigator blinded to the treatment assigned.
Histology. Hearts were removed and fixed in 4% paraformaldehyde
for 24 h, embedded in paraffin, and sectioned (5 m). Sections were
deparaffinized and stained with wheat germ agglutinin or picrosirius
red. Measurements of the cross-sectional area (n 250 cells from
both male and female animals) were obtained from images using
Slidebook software (Intelligent Imaging Innovations, Denver, CO).
Ca
2
imaging. Imaging during FGF23 perfusion was conducted as
reported previously (50, 51, 56). Briefly, cells were loaded at room
temperature with the fluorescent Ca
2
indicator Fluo-4-AM (2 M) for
20 min. Cells were washed three times in HBSS and allowed to deesterify
for 10 min at room temperature. Intracellular Ca
2
levels were measured
with an inverted microscope with fluorescent imaging capabilities [Olym-
pus IX51 (Olympus, Melville, NY) and Hamamatsu Orca-ERGA charge-
coupled device cameras (Hamamatsu, Bridgewater, NJ), Semrock Bright
Line filter set (Semrock, Rochester, NY), EXFO X-cite metal halide light
source (EXFO, Mississauga, ON, Canada), and Slidebook ratiometric
software (Intelligent Imaging Innovations)]. Diluted FGF23 (18,000 pg/
ml) was carefully perfused to the plates at a rate of 0.3 ml/min. In these
experiments, the five treatment conditions [vehicle, FGF23 alone, PD-
166866 (50 nM) FGF23, FGF23 in the presence of 0 mM extracellular
Ca
2
, and verapamil (10 M FGF23)] were tested five times, which
totaled six to 10 cardiomyocytes from a given animal, which was
repeated in three to five different animals. The fluorescent changes from
each cell were averaged, and the data were grouped by animal and then
used for data analysis. All data are presented as the peak increase in
fluorescence (F) after FGF23 application divided by the initial fluores-
cence before FGF23 application (F
0
). A F/F
0
of 1 indicates no change in
fluorescence from the baseline. Cells were tested for viability with KCl
(80 mM) at the end of each experiment and included in the data set only
E864 FGF23 ALTERS CARDIAC CONTRACTILITY AND INDUCES HYPERTROPHY
AJP-Endocrinol Metab doi:10.1152/ajpendo.00596.2012 www.ajpendo.org
at UMKC Libraries on April 29, 2013http://ajpendo.physiology.org/Downloaded from
if the response to KCl was greater than a 50% increase from baseline
fluorescence.
Cardiac contractility measurements. CD1 mice were euthanized by
cervical dislocation. The mouse hearts used for the muscle strip
experiments were quickly excised and placed into an ice-cold cardio-
protective Ringer’s solution (with Ca
2
) that included the addition of
2,3-butanedione monoxime (30 mM) for 30 min, as described previ-
ously (50). Briefly, LV muscle strips were prepared (1–2 mm wide by
6 8 mm long) in the cardioprotective solution. The strips were tied on
the proximal and distal ends with a silk thread. The muscle strips were
then rinsed three times (5 min each) in Ringer’s solution (with Ca
2
,
pH 7.4) to remove the 2,3-butanedione monoxime. The muscle strips
were hung vertically and attached to a force transducer between
bipolar platinum-stimulating electrodes suspended in 25-ml glass
tissue chambers and bubbled under 100% O
2
. Heart muscles were
stretched to the length of maximum force development in Ringer’s
solution (pH 7.4, without 2,3-butanedione monoxime) and stimulated
with pulses of 1 Hz for 5 ms. The stimulation voltage was set 20%
above threshold, and the muscles in the chamber were superfused with
Ringer’s solution (with Ca
2
, pH 7.4). Muscles were allowed to
stabilize for 90 min prior to experimentation and provided with fresh
media changes every 30 min. Muscles were paced at 1 Hz to obtain a
stable baseline and treated with either vehicle or FGF23. The con-
tractile data were recorded and analyzed on the LabChart 6 software.
Waveform changes were analyzed in the segments corresponding to
peak isometric tension (mN). Slope (mN/s) was analyzed by taking
the average slope from 10 to 20 ms after the start of the peak. Area
(mN s) was calculated using the region from 10 to 90% of the peak.
(s) was fitted at the baseline using data from 95 to 0% of peak. Strip
experiments were normalized within each condition to baseline levels
of contractility and presented as a relative change from baseline
contraction data.
Statistical analysis. All graphs were made and statistical proce-
dures performed using GraphPad Prism 5.0. Data are presented as
means SE. Data were compared using either a paired t-test or a
one-way analysis of variance, and significance was set at the P 0.05
level. When necessary, the one-way analysis of variance was followed
with appropriate post hoc tests. A Bonferroni post hoc adjustment was
used to correct for two to three comparisons to avoid type I error. In
cases where we made at least three comparisons, we utilized a Tukey
Post hoc adjustment to avoid type II error. FSC data was analyzed
using FlowJo Version 8.8.6 probability binning population compari-
son software (Tree Star) using a modified Cox Chi Squared Test
[T(X)]. A value of T(X) 4 implies that the two distributions are
different with a P value of 0.01 (99% confidence).
RESULTS
Markers of cardiac hypertrophy with exogenous FGF23. We
began this series of studies by testing the hypothesis that FGF23
directly induces hypertrophy in cardiomyocytes. Flow cytometry
revealed a concentration-dependent increase in HL-1 cell surface
area of cardiomyocytes exposed to FGF23 [T(X) 63, P 0.05;
Fig. 1, A and B]. Twenty-four-hour exposure of ventricular muscle
strips to FGF23 (900 pg/ml) resulted in increased expression of
early growth response 1 (EGR1), atrial natriuretic peptide (ANP),
and brain natriuretic peptide (BNP) (P 0.05; Fig. 2A). In
addition, 48-h exposure to FGF23 increased gene expression of
-myosin heavy chain (-MHC) and skeletal muscle -actin
(SkAct) (P 0.05; Fig. 2A). No statistically significant changes
were noted for c-Myc, c-Fos, or c-Jun following FGF23 treat-
ment. It is well known that FGFR signaling in the kidney involves
the activation of the MAPK cascade, particularly ERK (58). In
cardiomyocytes, ERK phosphorylation is known to induce the
EGR1 transcription factor as well as increase fetal gene expres-
sion associated with pathological hypertrophy (33). ERK phos-
phorylation in isolated cardiac muscle strips was increased signif-
icantly 15 min after treatment with FGF23 when compared with
vehicle-treated strips (p 0.05; Fig. 2B). In addition, FGF23
increased protein synthesis 8% (P 0.05; Fig. 2C), and this
increase was inhibited by the preaddition of PD-166866. These
data suggest that changes in cell size are FGFR mediated and are
not due simply to cell swelling.
FGFR and
-klotho gene expression. We quantified the
expression levels of FGFRs and -klotho in isolated cardio-
myocytes (Fig. 3A). The -cycle threshold (C
T
) values were
calculated using
-actin as the reference gene, and 2
⫺⌬C
T
calculations were performed. The order of expression from
highest to lowest was FGFR3, FGFR1, FGFR4, FGFR2, and
a-klotho. For ease and clarity of data presentation, we calcu-
lated the relative expression of each FGFR by comparing it
with -klotho, as shown in Fig. 3B. Statistical analysis was not
conducted on these data since it was transformed and normal-
ized to the lowest-expressed gene. Statistical analysis was
conducted on the raw C
T
values. FGFR3, FGFR1, and
FGFR4 were significantly higher compared with -klotho
(P 0.05). FGFR3 and FGFR1 did not differ statistically from
one another (P 0.05), but both were statistically higher than
FGFR4 and FGFR2 (P 0.05). FGFR4 and FGFR2 did not
differ from one another (P 0.05). Similar expression results
were confirmed using GAPDH as the housekeeping gene.
Col4a3
/
mice. Since we observed hypertrophic signaling
occurring with exposure to FGF23, we analyzed ventricular heart
tissue from 10-wk-old Col4a3
/
mice for markers of patholog
-
ical hypertrophy. The Col4a3
/
mouse is a model of human
Fig. 1. Fibroblast growth factor 23 (FGF23) increases cell size
in a dose-dependent manner. A: representative forward-scatter
histograms (FSC-H) of HL-1 cardiomyocytes treated with ve-
hicle or FGF23 (900 pg/ml) for 48 h; T(X) 63. B: summary
of forward-scatter (FSC-H) data on cardiomyocytes treated
with increasing doses of FGF23 (9 –900 pg/ml) using flow
cyctometry [T(X) 63, P 0.01]. FSC-H analysis of 10,000
live gated cells/sample (n 5 experiments). Results from
independent experiments were normalized to vehicle controls
and averaged. *Statistical difference from vehicle; †statistical
difference from FGF23 treatment (9 pg/ml).
E865FGF23 ALTERS CARDIAC CONTRACTILITY AND INDUCES HYPERTROPHY
AJP-Endocrinol Metab doi:10.1152/ajpendo.00596.2012 www.ajpendo.org
at UMKC Libraries on April 29, 2013http://ajpendo.physiology.org/Downloaded from
Alport syndrome, in which there is progression of CKD, and
therefore, it has has elevated levels of FGF23 starting at week 6
(46). We observed increases in ANP, -MHC, and SkAct in
Col4a3
/
hearts compared with hearts of their WT littermates
(P 0.05; Fig. 4A). Interestingly, we did not see gross morpho-
logical evidence of hypertrophy in the Col4a3
/
mice. There
were no changes in anterior or posterior wall thickness compared
with WT controls as measured by echocardiography (P 0.05;
Fig. 4B). We did not detect an increase in heart size by estimates
in LV mass from the echocardiogram or by comparing heart
weights with tibia length (P 0.05; Fig. 4C). In addition, the
average cardiomyocyte cross-sectional area based on the histolog-
ical analysis from both male and female mice were similar
in Col4a3
/
and WT hearts (Fig. 4, D and E). A randomly
selected subsample of cross-sections was stained with picrosirius
red, and we did not observe any differences in fibrosis between
WT and Col4a3
/
mice (mean% fibrosis of heart sections
0.46 vs. 0.49%, respectively; n 2). Representative M-mode
tracings from echocardiography are shown in Fig. 5A. Col4a3
/
mice did exhibit declines in LV function, as determined by
reductions in both fractional shortening and ejection fraction (17
and 11%, respectively, P 0.05; Fig. 5B). Because there
appeared to be a decrease in contractile function without hyper-
trophy, we hypothesized that there may be Ca
2
-handling issues,
and therefore, we tested three major Ca
2
-handling genes. Inter
-
estingly, Col4a3
/
mouse hearts showed a significant upregula
-
tion of the Na
/Ca
2
exchanger 1 (NCX1) mRNA (1.76
0.33-fold, P 0.05; n 5), however, there was not a significant
increase in NCX1 protein expression (1.15 0.09 for WT vs.
1.40 0.15 for Col4a3
/
, P 0.05; n 6 –7). In addition,
there were no significant increases in calsequestrin (Cal) (1.39
0.28-fold, P 0.05; n 5) or sarcoplasmic/endoplasmic reticu-
lum Ca
2
/ATPase (SERCA) (1.02 0.04-fold, P 0.05; n 5)
mRNA expression in the Col4a3
/
mouse hearts.
Ca
2
imaging of primary cardiomyocytes. To determine
whether FGF23 can modulate the levels of [Ca
2
]
i
,were
-
Fig. 2. FGF23 increases markers of hyper-
trophy in mouse ventricular tissue via FGFR.
A: FGF23 exposure increased the early
growth response 1 (EGR1) gene and the hy-
pertrophy-associated genes atrial natriuretic
peptide (ANP) and brain natriuretic peptide
(BNP) after 24 h. Forty-eight-hour treatment
with FGF23 increased the expression of other
hypertrophy genes -myosin heavy chain (-
MHC) and skeletal muscle -actin (SkAct)
(n 7). B: FGF23 treatment increased ERK
phosphorylation after 15 min (n 4). C: total
protein content of ventricular muscle strips
increased following 48-h treatment with
FGF23 and was blocked by pretreatment with
PD-166866 (n 5–7). *Statistical signifi-
cance from vehicle; †PD166 FGF23 treat-
ment is statistically different than FGF23
treatment alone.
Fig. 3. FGF receptors (FGFRs) and -klotho are expressed
in isolated adult ventricular cardiomyocytes. A: real-time
RT-PCR reaction (run in triplicates) showing the average
fluorescence values at each cycle number for FGFR1–4 and
-klotho from isolated adult ventricular cardiomyocytes.
B: summary of the relative expression data from cardiomy-
ocytes for FGFRs when normalized to the lowest-expressed
gene (-klotho; n 6). Statistics were conducted on the
nonnormalized gene expression data and are reported in the
results. -Act, -actin.
E866 FGF23 ALTERS CARDIAC CONTRACTILITY AND INDUCES HYPERTROPHY
AJP-Endocrinol Metab doi:10.1152/ajpendo.00596.2012 www.ajpendo.org
at UMKC Libraries on April 29, 2013http://ajpendo.physiology.org/Downloaded from
corded the Ca
2
responses in primary cardiomyocytes with the
fluorescent Ca
2
indicator Fluo-4 AM. Figure 6, A and B,
displays a representative response of a cardiomyocyte to
FGF23 (18,000 pg/ml). The myocyte displayed a spontaneous
Ca
2
oscillation prior to treatment, which after FGF23 had a
large and transient increase in [Ca
2
]
i
. FGF23 increased F/F
o
by 86% on average compared with vehicle (P 0.05; Fig. 6C).
We were able to prevent these increases in [Ca
2
]
i
by pretreat
-
ing cells with PD-166866 (P 0.05; Fig. 6C). In addition, we
were able to eliminate the increases in [Ca
2
]
i
by eliminating
extracellular Ca
2
upon pretreatment of the cells with the
L-type voltage-gated Ca
2
channel antagonist verapamil (10
M, P 0.05; Fig. 6D). The average time for a peak response
to perfused FGF23 was 186.6 6.6 s. The average time for a
peak response to perfused KCl was 68.8 5.6 s, and the
average F/F
o
response to KCl was 3.17 0.41. The dead-space
time of our perfusion system in these experiments was 45 s.
To test whether prolonged exposure to FGF23 induced an
increase in resting levels of Ca
2
, we loaded primary cardio
-
myocytes with the ratiometric fluorescent Ca
2
indicator Fura-
2-AM. Cardiomyocytes treated with FGF23 for 2 h increased
[Ca
2
]
i
by 25% (P 0.05; data not shown).
Contractility. Since FGF23 acutely increased [Ca
2
]
i
,we
explored the effect of exogenous FGF23 in cardiac muscle con-
tractility. We compared the contractile responses elicited by in-
creasing concentrations of FGF23 compared with vehicle. Peak
changes in contractility were noted between 15 and 20 min
following FGF23 treatment. Figure 7A displays raw tracings of
paced ventricular muscle strips following treatment with vehicle
and FGF23 (9,000 pg/ml). Increasing concentrations of FGF23
(900 and 9,000 pg/ml) increased isometric force when compared
with vehicle (P 0.05; Fig. 7B). We analyzed the effects of
FGF23 on specific characteristics of each contractile waveform.
FGF23 (900 and 9,000 pg/ml) increased the slope of contraction
(P 0.05; Fig. 7C) and the overall area (P 0.05; Fig. 7C)ofthe
contractile waveform (i.e., the integral/impulse) when compared
with vehicle. However, FGF23 had no effect on
, the time
constant of decay (rate of relaxation), compared with vehicle (Fig.
7C). To test whether the changes in contractility were receptor
mediated, we repeated a series of experiments with PD-166866
(50 nM) that eliminated the increases in isometric tension, slope,
and overall area induced by FGF23 (P 0.05; Fig. 7D). In
addition, FGF23 PD-166866 had no effect on
.
DISCUSSION
There have been several recent clinical reports suggesting
that FGF23 may alter heart function, particularly during CKD
(15, 21, 22, 35). However, there have been very few studies
Fig. 4. Col4a3
/
mouse hearts express genetic
markers of hypertrophy but do not display in-
creases in gross or cellular measures of hyper-
trophy. A: hearts from 10-wk-old Col4a3
/
mice had increased expression of the hypertro-
phy-associated genes ANP, -MHC, and SkAct
(n 6). B Col4a3
/
mice did not show any
changes in anterior (Ant.) or posterior (Pst.) wall
thickness during diastole obtained by echocar-
diography (n 12–13). C: no changes were
observed either in the left ventricular (LV)
mass/body weight (BW) ratio or when heart
weight (HW) was compared with tibia length
(TL) in the Col4a3
/
mice (n 12–13).
D: representative images showing wheat germ
agglutinin-stained cardiomyocytes sectioned
from Col4a3
/
and wild-type (WT) mouse
hearts. E: summary of cardiomyocyte cross-
sectional area data from WT and Col4a3
/
cardiomyocytes (n 3 animals/group). *Statis-
tical significance from WT.
Fig. 5. Col4a3
/
mouse hearts display reduced LV function. A: representative
M-mode images showing reduced systolic excursion of LV walls in Col4a3
/
mice (bottom), indicating reduced systolic function compared with their WT
littermates (top). B: quantitative echocardiographic data show reduced LV
function, as evidenced by lower fractional shortening (FS) and ejection fraction
(EF) in Col4a3
/
mice compared with WT littermates. *Statistical signifi
-
cance from WT.
E867FGF23 ALTERS CARDIAC CONTRACTILITY AND INDUCES HYPERTROPHY
AJP-Endocrinol Metab doi:10.1152/ajpendo.00596.2012 www.ajpendo.org
at UMKC Libraries on April 29, 2013http://ajpendo.physiology.org/Downloaded from
that have attempted to address the direct effects of FGF23 on
the myocardium. Therefore, we sought to determine what
alterations in cardiac function would occur during exposure to
FGF23. The major findings of this study are as follows. 1) Ex-
posure to FGF23 causes a dose-dependent increase in cell size
as well as increased protein synthesis and expression of com-
mon hypertrophy markers; 2) the Col4a3
/
mouse model of
CKD that is known to have elevated FGF23 demonstrates
increased gene expression of markers of pathological hyper-
trophy but does not show increases in cardiomyocyte size or
ventricular wall thickness; 3) Col4a3
/
mice have alterations
in contractile function that appear to precede the potential
development of pathological hypertrophy; 4) acute exposure to
FGF23 increases [Ca
2
]
i
in adult ventricular myocytes; and
5) acute FGF23 exposure alters cardiac contractility by increas-
ing force, rate of force development, and the area under the
curve (integral).
Hypertrophy. The normal level of FGF23 in the plasma in
healthy patients is 13.3 19.0 pg/ml; however, in CKD
plasma, FGF23 levels can rise 100- to 1,000-fold higher than
patients with normal renal function (25). In patients with CKD,
these elevated serum levels of FGF23 have been clinically
associated with increased LV mass and increased risk of LV
hypertrophy (20, 35, 52). Moreover, FGF23 levels have been
shown to predict outcomes in patients with systolic heart
failure (39). However, despite these strong clinical associa-
tions, there has been only a single basic science study to date
to show that FGF23 directly induces a hypertrophic phenotype
in the heart (13). Therefore, we determined whether the hyper-
trophic effects of FGF23 could also be replicated in adult
cardiomyocytes in vitro and whether hypertrophy is present in
another animal model of chronic kidney disease (Col4a3
/
)
that has elevated serum FGF23.
To investigate the effects of FGF23 on hypertrophy, we first
performed concentration response experiments analyzing
changes in cell size with HL-1 cardiomyocytes. We utilized
concentrations of 90, 900, and 9,000 pg/ml, as that represents
an approximate baseline level (in WT mice), and then 10- and
100-fold higher concentrations, which would be expected dur-
ing CKD. HL-1 cardiomyocytes were utilized since they main-
tain phenotypic characteristics of adult myocytes (7), have
been used previously in models of cardiac hypertrophy (5, 6,
30), do not contain fibroblasts, and can be used in large
population numbers over extended periods of time to more
accurately detect changes in cell size. Similarly to Faul et al.
(13), who used neonatal cardiomyocytes, our data show that
FGF23 induced HL-1 cell growth up to 24% in a concentra-
tion-dependent manner. To validate that FGF23 could have a
direct effect on cardiac tissue, we analyzed early growth
response genes and fetal genes associated with pathological
cardiac hypertrophy in ventricular muscle strips. Similarly to
previous reports (13, 53), we were unable to induce EGR1
expression 1 h following FGF23 exposure (data not shown);
however, we report that FGF23 can induce EGR1 expression in
cardiac tissue 24 h following administration. The expression of
c-Myc (a mediator of growth signaling in cardiomyocytes) did
not reach statistical significance despite increasing more than
twofold following treatment with FGF23. In addition, FGF23
treatment also resulted in the elevated expression of BNP,
ANP, -MHC, and SkAct, which are well-known markers of
cardiac hypertrophy. Specifically, there have been significant
correlations between elevated FGF23 and elevated BNP
plasma levels in patients with left ventricular hypertrophy (19,
43). BNP in particular is used as a diagnostic indicator for heart
failure (12), suggesting that FGF23 may directly promote the
progression of heart failure.
Previous research in noncardiac tissue has shown that
FGF23 is a potent inducer of ERK phosphorylation and sub-
sequent EGR1 expression (1, 53, 58). In cardiac muscle strips
treated with FGF23, we noted a p-ERK response within 15 min
Fig. 6. FGF23 treatment increases intracellular Ca
2
([Ca
2
]
i
) in primary cardiomyocytes. A: acute fluo-4
changes in [Ca
2
]
i
[in relative fluorescent units (RFU)]
in primary cardiomyocytes immediately following treat-
ment with FGF23 in the presence of extracellular Ca
2
.
For the fluorescent images, warmer colors (yellow, red)
indicated increased fluorescence (increase in [Ca
2
]
i
). B:
fluo-4 image of a primary cardiomyocyte at baseline and
at the peak fluorometric response following acute FGF23
treatment. C: summary data showing the average acute
changes in fluorescence mediated by FGF23 and during
receptor antagonism (PD-166866; n 10 –30 cells, 3–5
animals). Measurements are indicated as a change in
fluorescence after treatment divided by the initial fluo-
rescence (F/F
O
). D: summary data showing the average
changes in fluorescence mediated by FGF23 treatment in
the absence of extracellular Ca
2
(0 mM, 0Ca
2
) and
following pretreatment with verapamil (Verap; n
10 –30 cells, 3–5 animals). *Statistical difference from
vehicle; †statistical difference from FGF23 treatment.
E868 FGF23 ALTERS CARDIAC CONTRACTILITY AND INDUCES HYPERTROPHY
AJP-Endocrinol Metab doi:10.1152/ajpendo.00596.2012 www.ajpendo.org
at UMKC Libraries on April 29, 2013http://ajpendo.physiology.org/Downloaded from
of treatment. Interestingly, it has been reported previously that
FGF23 did not increase cardiac expression of p-ERK at 30 and
60 min following exposure in neonatal cardiomyocytes (13).
However, these authors did show that the ERK inhibitor
U-0126 was able to attenuate increases in FGF23-induced cell
size (13). These differences with our findings for p-ERK may
be due to differences in age (neonatal vs. adult), the tissue type
(isolated myocytes vs. whole tissue), or the timing of the
measurements in these studies.
Finally, we applied exogenous FGF23 to cultured adult
ventricular muscle strips and measured changes in total protein
content. FGF23 increased protein synthesis following 48 h of
treatment, and this effect was also dependent on FGFR acti-
vation, as it was eliminated by pretreatment with PD-166866.
Taken together, our data lend strength to the hypothesis that
FGF23 can directly induce cardiac hypertrophy.
FGFRs and
-klotho. It is currently unknown which recep-
tors are necessary for FGF23 to exert its effects on the heart.
FGF23 is known to bind FGFR1– 4 with varying degrees of
affinity (59, 60), and previous reports using end point RT-PCR
and immunohistochemistry in neonatal cardiomyocytes and
adult hearts have shown that FGFR1– 4 are present (13, 23).
Our data now extend these findings by using real-time RT-PCR
to quantitate relative expression levels in isolated primary
cardiomyocytes, which showed that FGFR3 and FGFR1 are
the most abundantly expressed. Since we were able to block
the acute and chronic effects FGF23 with PD-166866, a selec-
tive inhibitor of FGFR1 at 50 nM (37), and FGFR1 is abun-
dantly expressed, it seems likely that FGFR1 is an important
mediator of FGF23 cell signaling in the myocycardium.
FGF23 is thought to have a high binding affinity for FGFR1-
-klotho complexes in other tissues, like the kidney and
parathyroid. In contrast to previous studies (13, 29, 47, 53), we
have detected -klotho expression in the heart. Our ability to
detect -klotho may be attributed to the increased sensitivity of
the one-step real-time RT-PCR procedure (54). However,
given that we found -klotho to be 3,750-fold lower in
expression than our highest-expressed gene, FGFR3, we sup-
port the hypothesis that -klotho likely plays a limited role in
FGF23 signaling in cardiomyocytes (13). Moreover, it is pos-
sible that other FGFRs in addition to FGFR1 may be involved
in FGF23-mediated actions on cardiomyocytes. Previous stud-
ies have suggested that FGFR1, FGFR3, and FGFR4 can act in
concert to mediate FGF23 effects in the kidney (32). It has also
been suggested that, in the absence of -klotho, FGF23 has a
high affinity for FGFR4 (23, 60) and may mediate effects on
the heart (13). However, we would hypothesize that FGFR4 as
well as FGFR2 may play limited roles due to their lower
expression levels. Taken together, our data lend support to the
hypothesis that FGF23 signaling in the heart may be indepen-
dent of -klotho expression and suggest that FGFR1 and
FGFR3 may be critical to signaling in the heart. Nevertheless,
a more thorough inquiry into exact mechanisms responsible for
FGFR cardiac signaling awaits further investigation.
Evidence of cardiac dysfunction in Col4a3
/
mice. It has
been shown recently that the 5/6-nephrectomized rat model of
Fig. 7. FGF23 acutely increases contractile
force in ex vivo ventricular muscle strips via
FGFR. A: tracings of LV muscle contractions
at baseline and following treatment with either
FGF23 or vehicle. B: mean changes in isomet-
ric tension data induced by FGF23 or vehicle
normalized to baseline contractions. C: FGF23
treatment increased the slope and area of the
contractile waveforms but not the rate of relx-
ation,
(n 6 –12 animals). D: PD-166866
pretreatment prevents FGF23-induced in-
creases in tension, slope, and area. No changes
were noted in
as a result of FGF23, PD-
166866, or vehicle treatments (n 6 –12).
*Statistical difference from vehicle; †statistical
difference from FGF23 treatment.
E869FGF23 ALTERS CARDIAC CONTRACTILITY AND INDUCES HYPERTROPHY
AJP-Endocrinol Metab doi:10.1152/ajpendo.00596.2012 www.ajpendo.org
at UMKC Libraries on April 29, 2013http://ajpendo.physiology.org/Downloaded from
CKD develops left ventricular hypertrophy that was signifi-
cantly attenuated with the FGFR antagonist PD-173074 (13).
Interestingly, another study found that a FGF23-neutralizing
antibody did not reduce the hypertrophy in this animal model
(44). Therefore, we were interested in exploring whether a
different model of CKD with elevated FGF23 could develop
cardiac hypertrophy. The Col4a3
/
mouse is a model of
autosomal-recessive Alport syndrome. This mouse model is
nonhypertensive and has a progressive increase in the serum
levels of FGF23 that precedes elevations of traditional markers
of renal dysfunction (17, 18). From weeks 4 to 6, FGF23 levels
increase from 130 to 260 pg/ml, respectively. However,
serum FGF23 levels increase exponentially from weeks 8 to 12
(440 to 5,400 pg/ml, respectively) (46). Therefore, we used
Col4a3
/
mice to explore the connection between elevated
serum FGF23 and cardiac hypertrophy. Similarly to the 5/6-
nephrectomized mice (13, 44), Col4a3
/
mice presented with
increased gene expression markers of hypertrophy (ANP,
SkAct, and MHC); however, there were no significant in-
creases in anterior/posterior wall thickness, nor was there
increased cardiac mass by echocardiography or necropsy. Fur-
thermore, there was no increase in average cardiomyocyte size
based on histological analysis. The Col4a3
/
mouse model
has significantly increased rates of mortality beginning at 10
wk, which prevented further characterization of these animals.
We postulate that, given additional time, these mice may also
have demonstrated significant increases in cell size given that
the hypertrophic gene markers were increased at 10 wk.
Although we did not observe hypertrophy, there were sig-
nificant decreases in the contractile parameters (fractional
shortening and ejection fraction). In the 5/6-nephrectomized
animals, Faul et al. (13) reported a decline in ejection fraction
(although it did not reach statistical significance), which was
observed concurrent with increases in hypertrophy. This de-
crease in ejection fraction was eliminated by a FGFR antago-
nist (13). Interestingly, the changes in contractility in the
Col4a3
/
mice in our study occurred without significant
hypertrophy or fibrosis and in a CKD animal model that does
not demonstrate chronic hypertension. Thus the Col4a3
/
mouse model may provide an interesting tool for studying
cardiac effects of FGF23. Differences in cardiac function
between animal models with high FGF23 need to continue to
be explored to fully elucidate cardiac effects of FGF23 in vivo
during CKD.
The contractile deficits we observed in the animal model led
us to explore whether FGF23 is altering Ca
2
-handling genes
or directly altering Ca
2
levels. We first explored Ca
2
-
handling genes that are known to increase during heart failure.
The Col4a3
/
mouse had a significant upregulation of NCX1
mRNA but not SERCA or Cal. Interestingly, we did not
observe a significant increase in NCX1 protein expression in
Col4a3
/
mice compared with WT. Because the NCX1 is an
important regulator of [Ca
2
]
i
in cardiomyocytes and is in
-
creased during cardiac hypertrophy (28, 40), future studies
concerning the effects of FGF23 on the myocardium may be
warranted.
[Ca
2
]
i
. Next, we wanted to explore whether FGF23 could
be directly altering [Ca
2
]
i
. Acute exposure of primary cardio
-
myocytes to FGF23 increased [Ca
2
]
i
significantly, and we
were able to eliminate this increase via pretreatment with
PD-166866. Removing extracellular Ca
2
also abolished the
FGF23-evoked increase in [Ca
2
]
i
, suggesting that FGF23
likely opens a Ca
2
channel on the cellular membrane of
cardiomyoyctes to augment contraction rather than altering
release of Ca
2
from internal stores. To test this hypothesis, we
pretreated cardiomyoctyes with the L-type Ca
2
channel
blocker verapamil. Pretreatment with verapamil also com-
pletely inhibited Ca
2
entry, suggesting that FGF23 can affect
L-type gating. Furthermore, prolonged exposure to FGF23
resulted in a 25% increase in basal levels of Ca
2
, suggesting
that over time FGF23 may lead to Ca
2
overload. Increased
basal levels of Ca
2
have been linked to remodeling and
hypertrophy of the heart (3, 11, 16). Thus [Ca
2
]
i
may be a
critical link between elevated FGF23, acute alterations in
cardiac function, long-term remodeling, hypertrophy, and ulti-
mately heart failure.
Cardiac contractility. Finally, to determine the acute effects
of this elevated [Ca
2
]
i
, we explored the effects of exogenous
FGF23 on isolated ventricular muscle contractility. We hy-
pothesized that the increases in [Ca
2
]
i
were large enough to
acutely improve cardiac contractility. We have shown for the
first time that FGF23 significantly increases isometric tension,
slope, and the area of contraction in isolated cardiac muscle.
During cardiac excitation contraction coupling, force is gener-
ated on a beat-to-beat basis by a 10-fold increase in cytosolic
Ca
2
by a process known as Ca
2
-induced Ca
2
release
(CICR). As a cardiac myocyte depolarizes, [Ca
2
]
i
begins to
accumulate principally from the opening of L-type voltage-
gated Ca
2
channels, and this triggers CICR from RyR2 to
drive muscle contraction. Therefore, [Ca
2
]
i
increases are
tightly coupled to increased cardiac contractility. Our data
showing an increase not only in the magnitude but also in the
slope and area suggest that FGF23 may alter the CICR mech-
anism, allowing for greater [Ca
2
]
i
, thus promoting faster and
more powerful contractions. Our changes in tension, slope, and
area lend strength to our calcium imaging data showing that
FGF23 promotes Ca
2
entry via L-type Ca
2
channels. During
relaxation, a return to Ca
2
homeostasis is controlled princi
-
pally by SERCA and the NCX, with minor contribution from
the plasma membrane Ca
2
-ATPase. If FGF23 is increasing
[Ca
2
]
i
and the contraction by slowing Ca
2
removal, then
there should be a corresponding increase in
(rate of relaxation
following contraction). Our data show that FGF23 does not
affect
, demonstrating that it is unlikely to have a major acute
effect on SERCA, NCX, or Ca
2
-ATPase transporter function.
Receptor antagonism with PD-166866 was able to eliminate
the changes in the contractile waveforms, demonstrating that
FGFRs mediate the effects of FGF23 in the myocardium.
Significance. Although an axis of signaling has been re-
ported between bone and brain, gut, kidney, parathyroid, and
adipose, potential endocrine cross-talk between bone and heart
has not been well explored. A major question that arises from
the current investigations on FGF23 and the heart is whether
FGF23 is having both physiological and pathological effects on
the myocardium. One possible physiological benefit of FGF23
altering cardiac contractility may be increasing renal phosphate
clearance. Acutely increasing cardiac contractility may be a
mechanism to increase cardiac output and renal blood flow. In
addition, FGF23 may act on the heart to promote the expres-
sion of ANP/BNP, which would increase vasodilation, natri-
uresis, and diuresis to clear excess phosphate. However, in
CKD the increases in ANP/BNP are unable to improve renal
E870 FGF23 ALTERS CARDIAC CONTRACTILITY AND INDUCES HYPERTROPHY
AJP-Endocrinol Metab doi:10.1152/ajpendo.00596.2012 www.ajpendo.org
at UMKC Libraries on April 29, 2013http://ajpendo.physiology.org/Downloaded from
clearance of phosphate. This accumulation of serum phosphate
then elevates serum FGF23, which at high concentrations may
promote cardiac dysfunction. This cycle appears to continue in
CKD since increased ANP/BNP and FGF23 are predictive of
CKD progression (10, 45) and cardiac pathologies (15, 21, 22,
26, 35, 43, 52).
From our findings, we propose that FGF23 is behaving like
other well-characterized stress hormones (i.e., norepinephrine,
epinephrine, and angiotensin II). Acutely, these hormones are
inotropic and act to restore homeostasis; however, over the
long term, chronic elevations in these hormones and [Ca
2
]
i
can cause contractile dysfunction, remodeling of the heart, and
progression to cardiac hypertrophy. Similarly, our findings
have led us to propose that endogenous FGF23 increases
[Ca
2
]
i
in cardiomyocytes. This finding is important because
increases in [Ca
2
]
i
can initially be diverted to the excitation-
contraction coupling process and increased stimulation of
CICR to improve cardiac contractility. However, long-term
exposure to FGF23 may create disruptions in Ca
2
homeosta
-
sis that then activate transcriptional remodeling mechanisms
that contribute to long-term impairments in contractile function
and ultimately cardiac hypertrophy. For example, it has been
shown that FGF23 can activate the Ca
2
-sensitive calcineurin-
NFAT signaling pathway (13), suggesting that [Ca
2
]
i
may be
an important trigger for hypertrophic signaling in car-
diomyoyctes. Thus, FGF23’s induction of Ca
2
signals ap
-
pears to be important for controlling both transcriptional regula-
tion and contractility. Importantly, our data in the Col4a3
/
mice
suggest that left ventricular dysfunction may precede the devel-
opment of cardiac hypertrophy. This may have important impli-
cations on a patient’s quality of life and may also serve as an
important clinical diagnostic marker for severity of disease.
In summary, our data show that FGF23 may have additional
effects on the heart, in addition to hypertrophy, specifically
related to calcium handling and cardiac contractility. There-
fore, our studies provide an important rationale to further
investigate the mechanisms for direct effects of this important
bone endocrine factor on the heart.
ACKNOWLEDGMENTS
We thank Yixia Xie for assistance in collecting histological data.
GRANTS
This work was supported by National Institutes of Health Grants 1-RC2-
AR-058962 (to L. F. Bonewald and M. J. Wacker), R01-HL-089939 (to B.
Dawn), and K08-DK-087949 (to J. R. Stubbs) and an American Heart
Association Scientist Development Grant (11SDG5330016) to M. J. Wacker.
DISCLOSURES
All authors state that they have no conflicts of interest, financial or
otherwise.
AUTHOR CONTRIBUTIONS
C.D.T., T.M.G., V.T., L.A.W., B.D., L.B., J.R.S., and M.J.W. contributed
to the conception and design of the research; C.D.T., T.M.G., V.T., J.E.M.,
T.F.M., B.W.C., M.G., R.J.V., L.A.W., and M.J.W. performed the experi-
ments; C.D.T., T.M.G., V.T., J.E.M., T.F.M., B.W.C., M.G., R.J.V., L.A.W.,
B.D., L.B., J.R.S., and M.J.W. analyzed the data; C.D.T., T.M.G., V.T.,
L.A.W., B.D., L.B., J.R.S., and M.J.W. interpreted the results of the experi-
ments; C.D.T., B.D., and M.J.W. prepared the figures; C.D.T. and M.J.W.
drafted the manuscript; C.D.T., T.M.G., V.T., T.F.M., B.W.C., L.A.W., B.D.,
L.B., J.R.S., and M.J.W. edited and revised the manuscript; C.D.T., T.M.G.,
V.T., J.E.M., T.F.M., B.W.C., M.G., R.J.V., L.A.W., B.D., L.B., J.R.S., and
M.J.W. approved the final version of the manuscript.
REFERENCES
1. Ben-Dov IZ, Galitzer H, Lavi-Moshayoff V, Goetz R, Kuro-o M,
Mohammadi M, Sirkis R, Naveh-Many T, Silver J. The parathyroid is
a target organ for FGF23 in rats. J Clin Invest 117: 4003–4008, 2007.
2. Bergwitz C, Juppner H. Regulation of phosphate homeostasis by PTH,
vitamin D, and FGF23. Annu Rev Med 61: 91–104, 2010.
3. Bers DM. Calcium cycling and signaling in cardiac myocytes. Annu Rev
Physiol 70: 23–49, 2008.
4. Bonewald LF, Wacker MJ. FGF23 production by osteocytes. Pediatr
Nephrol 28: 563–568, 2013.
5. Brunt KR, Tsuji MR, Lai JH, Kinobe RT, Durante W, Claycomb WC,
Ward CA, Melo LG. Heme oxygenase-1 inhibits pro-oxidant induced
hypertrophy in HL-1 cardiomyocytes. Exp Biol Med (Maywood) 234:
582–594, 2009.
6. Chandrasekar B, Mummidi S, Claycomb WC, Mestril R, Nemer M.
Interleukin-18 is a pro-hypertrophic cytokine that acts through a phospha-
tidylinositol 3-kinase-phosphoinositide-dependent kinase-1-Akt-GATA4
signaling pathway in cardiomyocytes. J Biol Chem 280: 4553–4567, 2005.
7. Claycomb WC, Lanson NA Jr, Stallworth BS, Egeland DB, Delcarpio
JB, Bahinski A, Izzo NJ Jr. HL-1 cells: a cardiac muscle cell line that
contracts and retains phenotypic characteristics of the adult cardiomyo-
cyte. Proc Natl Acad Sci USA 95: 2979 –2984, 1998.
8. Dalal M, Sun K, Cappola AR, Ferrucci L, Crasto C, Fried LP, Semba
RD. Relationship of serum fibroblast growth factor 23 with cardiovascular
disease in older community-dwelling women. Eur J Endocrinol 165:
797–803, 2011.
9. Dawn B, Guo Y, Rezazadeh A, Huang Y, Stein AB, Hunt G, Tiwari S,
Varma J, Gu Y, Prabhu SD, Kajstura J, Anversa P, Ildstad ST, Bolli
R. Postinfarct cytokine therapy regenerates cardiac tissue and improves
left ventricular function. Circ Res 98: 1098 –1105, 2006.
10. Dieplinger B, Mueller T, Kollerits B, Struck J, Ritz E, von Eckard-
stein A, Haltmayer M, Kronenberg F; MMKD Study Group. Pro-A-
type natriuretic peptide and pro-adrenomedullin predict progression of
chronic kidney disease: the MMKD Study. Kidney Int 75: 408 –414, 2009.
11. Eder P, Molkentin JD. TRPC channels as effectors of cardiac hypertro-
phy. Circ Res 108: 265–272, 2011.
12. Ewald B, Ewald D, Thakkinstian A, Attia J. Meta-analysis of B type
natriuretic peptide and N-terminal pro B natriuretic peptide in the diag-
nosis of clinical heart failure and population screening for left ventricular
systolic dysfunction. Intern Med J 38: 101–113, 2008.
13. Faul C, Amaral AP, Oskouei B, Hu MC, Sloan A, Isakova T, Gutiér-
rez OM, Aguillon-Prada R, Lincoln J, Hare JM, Mundel P, Morales
A, Scialla J, Fischer M, Soliman EZ, Chen J, Go AS, Rosas SE, Nessel
L, Townsend RR, Feldman HI, St John Sutton M, Ojo A, Gadegbeku
C, Di Marco GS, Reuter S, Kentrup D, Tiemann K, Brand M, Hill JA,
Moe OW, Kuro-O M, Kusek JW, Keane MG, Wolf M. FGF23 induces
left ventricular hypertrophy. J Clin Invest 121: 4393–4408, 2011.
14. Feng JQ, Ward LM, Liu S, Lu Y, Xie Y, Yuan B, Yu X, Rauch F,
Davis SI, Zhang S, Rios H, Drezner MK, Quarles LD, Bonewald LF,
White KE. Loss of DMP1 causes rickets and osteomalacia and identifies
a role for osteocytes in mineral metabolism. Nat Genet 38: 1310 –1315,
2006.
15. Ford ML, Smith ER, Tomlinson LA, Chatterjee PK, Rajkumar C,
Holt SG. FGF-23 and osteoprotegerin are independently associated with
myocardial damage in chronic kidney disease stages 3 and 4. Another link
between chronic kidney disease-mineral bone disorder and the heart.
Nephrol Dial Transplant 27: 727–733, 2012.
16. Goonasekera SA, Molkentin JD. Unraveling the secrets of a double life:
contractile versus signaling Ca2 in a cardiac myocyte. J Mol Cell
Cardiol 52: 317–322, 2012.
17. Gross O, Girgert R, Rubel D, Temme J, Theissen S, Müller GA. Renal
protective effects of aliskiren beyond its antihypertensive property in a
mouse model of progressive fibrosis. Am J Hypertens 24: 355–361, 2011.
18. Gross O, Schulze-Lohoff E, Koepke ML, Beirowski B, Addicks K,
Bloch W, Smyth N, Weber M. Antifibrotic, nephroprotective potential of
ACE inhibitor vs AT1 antagonist in a murine model of renal fibrosis.
Nephrol Dial Transplant 19: 1716 –1723, 2004.
19. Gruson D, Lepoutre T, Ketelslegers JM, Cumps J, Ahn SA, Rousseau
MF. C-terminal FGF23 is a strong predictor of survival in systolic heart
failure. Peptides 37: 258 –262, 2012.
20. Gutiérrez OM, Januzzi JL, Isakova T, Laliberte K, Smith K, Col-
lerone G, Sarwar A, Hoffmann U, Coglianese E, Christenson R, Wang
E871FGF23 ALTERS CARDIAC CONTRACTILITY AND INDUCES HYPERTROPHY
AJP-Endocrinol Metab doi:10.1152/ajpendo.00596.2012 www.ajpendo.org
at UMKC Libraries on April 29, 2013http://ajpendo.physiology.org/Downloaded from
TJ, deFilippi C, Wolf M. Fibroblast growth factor 23 and left ventricular
hypertrophy in chronic kidney disease. Circulation 119: 2545–2552, 2009.
21. Holden RM, Beseau D, Booth SL, Adams MA, Garland JS, Morton
RA, Collier CP, Foley RN. FGF-23 is associated with cardiac troponin T
and mortality in hemodialysis patients. Hemodial Int 16: 53–58, 2012.
22. Hsu HJ, Wu MS. Fibroblast growth factor 23: a possible cause of left
ventricular hypertrophy in hemodialysis patients. Am J Med Sci 337:
116 –122, 2009.
23. Hughes SE. Differential expression of the fibroblast growth factor recep-
tor (FGFR) multigene family in normal human adult tissues. J Histochem
Cytochem 45: 1005–1019, 1997.
24. Imanishi Y, Inaba M, Nakatsuka K, Nagasue K, Okuno S, Yoshihara
A, Miura M, Miyauchi A, Kobayashi K, Miki T, Shoji T, Ishimura E,
Nishizawa Y. FGF-23 in patients with end-stage renal disease on hemo-
dialysis. Kidney Int 65: 1943–1946, 2004.
25. Imel EA, Peacock M, Pitukcheewanont P, Heller HJ, Ward LM,
Shulman D, Kassem M, Rackoff P, Zimering M, Dalkin A, Drobny E,
Colussi G, Shaker JL, Hoogendoorn EH, Hui SL, Econs MJ. Sensi-
tivity of fibroblast growth factor 23 measurements in tumor-induced
osteomalacia. J Clin Endocrinol Metab 91: 2055–2061, 2006.
26. Ix JH, Katz R, Kestenbaum BR, de Boer IH, Chonchol M, Mukamal
KJ, Rifkin D, Siscovick DS, Sarnak MJ, Shlipak MG. Fibroblast
growth factor-23 and death, heart failure, and cardiovascular events in
community-living individuals: CHS (Cardiovascular Health Study). JAm
Coll Cardiol 60: 200 –207, 2012.
27. Kabaeva Z, Zhao M, Michele DE. Blebbistatin extends culture life of
adult mouse cardiac myocytes and allows efficient and stable transgene
expression. Am J Physiol Heart Circ Physiol 294: H1667–H1674, 2008.
28. Kent RL, Rozich JD, McCollam PL, McDermott DE, Thacker UF,
Menick DR, McDermott PJ, Cooper G 4th. Rapid expression of the
Na()-Ca2 exchanger in response to cardiac pressure overload. Am J
Physiol Heart Circ Physiol 265: H1024 –H1029, 1993.
29. Kuro-o M, Matsumura Y, Aizawa H, Kawaguchi H, Suga T, Utsugi T,
Ohyama Y, Kurabayashi M, Kaname T, Kume E, Iwasaki H, Iida A,
Shiraki-Iida T, Nishikawa S, Nagai R, Nabeshima YI. Mutation of the
mouse klotho gene leads to a syndrome resembling ageing. Nature 390:
45–51, 1997.
30. Landstrom AP, Kellen CA, Dixit SS, van Oort RJ, Garbino A,
Weisleder N, Ma J, Wehrens XH, Ackerman MJ. Junctophilin-2
expression silencing causes cardiocyte hypertrophy and abnormal intra-
cellular calcium-handling. Circ Heart Fail 4: 214 –223, 2011.
31. Larsson T, Nisbeth U, Ljunggren O, Jüppner H, Jonsson KB. Circu-
lating concentration of FGF-23 increases as renal function declines in
patients with chronic kidney disease, but does not change in response to
variation in phosphate intake in healthy volunteers. Kidney Int 64: 2272–
2279, 2003.
32. Li H, Martin A, David V, Quarles LD. Compound deletion of Fgfr3 and
Fgfr4 partially rescues the Hyp mouse phenotype. Am J Physiol Endocri-
nol Metab 300: E508 –E517, 2011.
33. Lorenz K, Schmitt JP, Vidal M, Lohse MJ. Cardiac hypertrophy:
targeting Raf/MEK/ERK1/2-signaling. Int J Biochem Cell Biol 41: 2351–
2355, 2009.
34. Mirza MA, Hansen T, Johansson L, Ahlström H, Larsson A, Lind L,
Larsson TE. Relationship between circulating FGF23 and total body
atherosclerosis in the community. Nephrol Dial Transplant 24: 3125–
3131, 2009.
35. Mirza MA, Larsson A, Melhus H, Lind L, Larsson TE. Serum intact
FGF23 associate with left ventricular mass, hypertrophy and geometry in
an elderly population. Atherosclerosis 207: 546 –551, 2009.
36. Negishi K, Kobayashi M, Ochiai I, Yamazaki Y, Hasegawa H, Ya-
mashita T, Shimizu T, Kasama S, Kurabayashi M. Association be-
tween fibroblast growth factor 23 and left ventricular hypertrophy in
maintenance hemodialysis patients. Comparison with B-type natriuretic
peptide and cardiac troponin T. Circ J 74: 2734 –2740, 2010.
37. Panek RL, Lu GH, Dahring TK, Batley BL, Connolly C, Hamby JM,
Brown KJ. In vitro biological characterization and antiangiogenic effects
of PD 166866, a selective inhibitor of the FGF-1 receptor tyrosine kinase.
J Pharmacol Exp Ther 286: 569 –577, 1998.
38. Pereira RC, Juppner H, Azucena-Serrano CE, Yadin O, Salusky IB,
Wesseling-Perry K. Patterns of FGF-23, DMP1, and MEPE expression in
patients with chronic kidney disease. Bone 45: 1161–1168, 2009.
39. Plischke M, Neuhold S, Adlbrecht C, Bielesz B, Shayganfar S, Biegl-
mayer C, Szekeres T, Horl WH, Strunk G, Vavken P, Pacher R,
Hulsmann M. Inorganic phosphate and FGF-23 predict outcome in stable
systolic heart failure. Eur J Clin Invest 42: 649 –656, 2012.
40. Roos KP, Jordan MC, Fishbein MC, Ritter MR, Friedlander M,
Chang HC, Rahgozar P, Han T, Garcia AJ, Maclellan WR, Ross RS,
Philipson KD. Hypertrophy and heart failure in mice overexpressing the
cardiac sodium-calcium exchanger. J Card Fail 13: 318 –329, 2007.
41. Sahn DJ, DeMaria A, Kisslo J, Weyman A. Recommendations regard-
ing quantitation in M-mode echocardiography: results of a survey of
echocardiographic measurements. Circulation 58: 1072–1083, 1978.
42. Schiller NB, Shah PM, Crawford M, DeMaria A, Devereux R, Feigen-
baum H, Gutgesell H, Reichek N, Sahn D, Schnittger I, Silverman NH,
Tajik AJ. Recommendations for quantitation of the left ventricle by
two-dimensional echocardiography. American Society of Echocardiogra-
phy Committee on Standards, Subcommittee on Quantitation of Two-
Dimensional Echocardiograms. J Am Soc Echocardiogr 2: 358 –367,
1989.
43. Seiler S, Cremers B, Rebling NM, Hornof F, Jeken J, Kersting S, Steimle
C, Ege P, Fehrenz M, Rogacev KS, Scheller B, Böhm M, Fliser D, Heine
GH. The phosphatonin fibroblast growth factor 23 links calcium-
phosphate metabolism with left-ventricular dysfunction and atrial fibrilla-
tion. Eur Heart J 32: 2688 –2696, 2011.
44. Shalhoub V, Shatzen EM, Ward SC, Davis J, Stevens J, Bi V,
Renshaw L, Hawkins N, Wang W, Chen C, Tsai MM, Cattley RC,
Wronski TJ, Xia X, Li X, Henley C, Eschenberg M, Richards WG.
FGF23 neutralization improves chronic kidney disease-associated hyper-
parathyroidism yet increases mortality. J Clin Invest 122: 2543–2553,
2012.
45. Spanaus KS, Kronenberg F, Ritz E, Schlapbach R, Fliser D, Hers-
berger M, Kollerits B, König P, von Eckardstein A; Mild-to-Moderate
Kidney Disease Study Group. B-type natriuretic peptide concentrations
predict the progression of nondiabetic chronic kidney disease: the Mild-
to-Moderate Kidney Disease Study. Clin Chem 53: 1264 –1272, 2007.
46. Stubbs JR, He N, Idiculla A, Gillihan R, Liu S, David V, Hong Y,
Quarles LD. Longitudinal evaluation of FGF23 changes and mineral
metabolism abnormalities in a mouse model of chronic kidney disease. J
Bone Miner Res 27: 38 –46, 2012.
47. Takeshita K, Fujimori T, Kurotaki Y, Honjo H, Tsujikawa H, Yasui
K, Lee JK, Kamiya K, Kitaichi K, Yamamoto K, Ito M, Kondo T, Iino
S, Inden Y, Hirai M, Murohara T, Kodama I, Nabeshima Y. Sinoatrial
node dysfunction and early unexpected death of mice with a defect of
klotho gene expression. Circulation 109: 1776 –1782, 2004.
48. Taylor EN, Rimm EB, Stampfer MJ, Curhan GC. Plasma fibroblast
growth factor 23, parathyroid hormone, phosphorus, and risk of coronary
heart disease. Am Heart J 161: 956 –962, 2011.
49. Teichholz LE, Kreulen T, Herman MV, Gorlin R. Problems in echo-
cardiographic volume determinations: echocardiographic-angiographic
correlations in the presence of absence of asynergy. Am J Cardiol 37:
7–11, 1976.
50. Touchberry CD, Bales IK, Stone JK, Rohrberg TJ, Parelkar NK,
Nguyen T, Fuentes O, Liu X, Qu CK, Andresen JJ, Valdivia HH, Brotto
M, Wacker MJ. Phosphatidylinositol 3,5-bisphosphate [PI(3,5)P2] poten-
tiates cardiac contractility via activation of the ryanodine receptor. J Biol
Chem 285: 40312–40321, 2010.
51. Touchberry CD, Elmore CJ, Nguyen TM, Andresen JJ, Zhao X,
Orange M, Weisleder N, Brotto M, Claycomb WC, Wacker MJ.
Store-operated calcium entry is present in HL-1 cardiomyocytes and
contributes to resting calcium. Biochem Biophys Res Commun 416: 45–50,
2011.
52. Unsal A, Kose Budak S, Koc Y, Basturk T, Sakaci T, Ahbap E,
Sinangil A. Relationship of Fibroblast Growth Factor 23 with Left
Ventricle Mass Index and Coronary Calcificaton in Chronic Renal Dis-
ease. Kidney Blood Press Res 36: 55–64, 2012.
53. Urakawa I, Yamazaki Y, Shimada T, Iijima K, Hasegawa H,
Okawa K, Fujita T, Fukumoto S, Yamashita T. Klotho converts
canonical FGF receptor