Cardiosulfa, a small molecule that induces abnormal heart development in zebrafish, and its biological implications

Article abstract of DOI:10.1002/anie.200902370

A change of heart: Cardiosulfa, a small molecule that induces heart deformation during zebrafish development, has been identified by using a forward chemicalgenetic approach. Zebrafish embryos exposed to cardiosulfa have a narrow and elongated heart within an enlarged pericardial sac (see picture; heart marked with green fluorescent protein).

Full text of DOI:10.1002/anie.200902370

Angewandte
Chemie
DOI: 10.1002/anie.200902370
Chemical Genetics
Cardiosulfa, a Small Molecule that Induces Abnormal Heart
Development in Zebrafish, and Its Biological Implications**
Sung-Kyun Ko, Hui Juan Jin, Da-Woon Jung, Xizhe Tian, and Injae Shin*
Heart disease is one of the most common causes of human
death. As a result, a major goal in understanding this disease
is the elucidation of mechanisms and pathways that are
directly related to the human heart function. The zebrafish
has emerged as a powerful model vertebrate for the assess-
ment of heart development owing to the fact that its heart is
nearly identical to that of a human embryo at three weeks of
gestation.[1] The hearts of vertebrate embryo, including
zebrafish, consist of two chambers, an anterior ventricle and
a posterior atrium, each having a characteristic shape, size,
and function that contribute to efficient circulation. Owing to
the transparency of embryonic zebrafish, heart function
during developmental stages can be visually detected by
using microscopy. Furthermore, significant defects in cardiac
form or function are easily observed in live embryos, because
embryonic zebrafish can survive in the absence of circulating
blood during early developmental stages.
identify small molecules that perturb heart development in
zebrafish. Herein, we describe the development of a novel,
sulfonamide-based small molecule that induces abnormal
heart morphology during zebrafish development.
Sulfonamides have great potential for use as bioactive
molecules because they display a diverse range of biological
activities, including antihypertensive, antidiabetic, and anti-
bacterial activities.[5] As a result, by using solution-phase
chemistry, we have constructed a novel sulfonamide library
comprising members with diverse substitution patterns. The
goal was to identify compounds that induce distinctive
phenotypic changes of the zebrafish heart during develop-
ment. The library was prepared by reacting excess sulfonyl
chlorides with primary and secondary amines in the presence
of either poly(4-vinylpyridine) or amberite IRA 960 as
polymer-supported bases (Scheme 1a, pathway a). After the
coupling reactions are complete, excess sulfonyl chlorides
Implementation of a conventional genetic approach, in
combination with phenotypic screening, serves as a common
approach to dissecting developmental processes and charac-
terizing protein functions.[2] However, there are drawbacks
associated with this methodology, including relatively long
experimental times and premature organism death resulting
from gene mutations or deletions. The forward chemical
genetic approach that employs small molecules to tempora-
rily regulate protein function in zebrafish has certain advan-
tages over genetic methods owing to its simplicity and ability
to conditionally regulate the activities of gene products.[3]
To date, small molecules that affect developmental
processes in zebrafish have been identified by employing
the forward chemical genetic approach and used to gain an
understanding of the functions of gene products involved in
vertebrate developmental events.[4] However, a more intense
effort is required to uncover new chemical modulators. As
part of a recent investigation aimed at this goal, we have
prepared and screened a sulfonamide library designed to
Scheme 1. a) Preparation of a sulfonamide library by using solution-
phase chemistry (substituents R1, R2, and R3 are given in the Support-
ing Information, Scheme S1). b) Structure of cardiosulfa.
were removed by adding aminomethylated polystyrene resin
to the reaction mixture.[6] Amines that have poor solubilities
in CH2Cl2 or poor reactivities were reacted with excess
sulfonyl chlorides in dry pyridine, and the target products
were isolated by column chromatography (Scheme 1a, path-
way b).[7] These synthetic pathways produced about 300
sulfonamides in good yields and high purities.
The sulfonamide library was screened for phenotypic
changes of the zebrafish heart during development by using a
96-well plate format.[8] Zebrafish embryos (three embryos per
well) were exposed to 10 mm sulfonamides at 288C, and their
phenotype was visually observed by using a dissecting micro-
scope over a five-day post-fertilization (dpf) period. One
[*] S.-K. Ko, Dr. H. J. Jin, Dr. D.-W. Jung, X. Tian, Prof. Dr. I. Shin
Department of Chemistry, Yonsei University
Seoul 120-749 (Korea)
Fax: (+82)2-364-7050
E-mail: injae@yonsei.ac.kr
[**] We thank Dr. T.-L. Huh (The Korea Zebrafish Organogenesis Mutant
Bank) for technical assistance with the zebrafish studies and for
providing Tg(cmlc2:GFP) transgenic zebrafish and cmlc2- and vmhc-
containing plasmids. This work was supported by grants of the NRL
and WCU (R32-2008-000-10217-0) programs (KOSEF/MEST). S.-
K.K. thanks the Seoul Science Fellowship program. S.-K.K. and X.Z.T.
also thank the BK21 program (KRF).
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200902370.
Angew. Chem. Int. Ed. 2009, 48, 7809 –7812
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
7809

Communications
compound, named “cardiosulfa” (Scheme 1b), was found to
Temporal control experiments with cardiosulfa were
performed to identify the developmental stage at which this
sulfonamide induces abnormal heart formation in zebrafish.
In these experiments, cardiosulfa was added or washed away
at various time points during development. When cardiosulfa
was added at initiation and washed away before 12 hpf,
normal heart morphology was observed (Figure 1e). Sim-
ilarly, embryos exposed to cardiosulfa after 24 hpf did not
show any apparent effects on heart development. However,
embryos exposed to this sulfonamide between 12 and 24 hpf
exhibited heart deformation. The results, obtained from
temporal control experiments, indicate that a critical stage
for heart development occurs between 12 and 24 hpf.[4c]
Several cardiosulfa analogues that contain a nitro and/or
trifluoromethyl group at the ortho, meta, and para positions of
the benzene ring were prepared using solution-phase chemis-
try to explore structure–activity relationships (Supporting
Information, Table S1). Each analogue (concentrations of 10,
20, and 30 mm) was added to zebrafish embryos at the
initiation of the experiments. Analogues with a single
substituent at the ortho and para position (2, 5, 7; Supporting
Information, Table S1) induced abnormal heart development
although they were slightly less effective than cardiosulfa. In
contrast, meta-substituted analogues (3, 4, 6; Supporting
Information, Table S1) did not perturb heart development.
Based on this limited analysis, it seems that substitution at the
meta position abolishes the abnormal heart formation effect
during zebrafish development.
induce impaired heart morphology, severe pericardial edema,
and severe yolk edema (Figure 1a,b). Cardiosulfa showed
pronounced effects on cardiac development at concentrations
of 5–30 mm. Zebrafish embryos exposed to these concentra-
tions of the substance displayed narrow and elongated
ventricle and atrium within an enlarged pericardial sac. In
addition, peripheral blood flow was reduced in the exposed
embryos. Although serious heart deformation is observed in
the exposed embryos, they remain alive up to 7 dpf.
The status of sarcomeric proteins in the heart were
examined by staining cardiosulfa-exposed zebrafish embryos
with two monoclonal antibodies, MF-20 and S46, that are
commonly employed to identify cardiovascular cells and to
screen mutated hearts (whole-mount immunostaining).
Whereas MF-20 recognizes a sarcomeric myosin heavy
chain present in both the ventricle and atrium, S46 binds to
an atrium-specific sarcomeric myosin heavy chain in zebra-
fish.[11] Both antibodies were found to stain the atrium and
ventricle in cardiosulfa-treated embryos, suggesting that
sarcomeric proteins are present at normal levels throughout
heart development in the treated zebrafish (Supporting
Information, Figure S1).
The expression pattern of heart-related genes in whole
embryos was examined by using digoxigenin-labeled anti-
sense RNA probes (whole-mount in situ hybridization). This
technique enables the detection of specific genes in morpho-
logically preserved embryos. Two zebrafish cardiac myosin
genes, cmlc2 and ventricular myosin heavy chain (vmhc),
were used to distinguish two populations of myocardial
precursors at an early stage. Whereas cmlc2 is expressed
throughout both chambers, vmhc is expressed throughout the
ventricle but not in the atrium.[9] The results obtained from
these experiments show that by 48 hpf, expression patterns of
two genes are nearly identical in both the untreated and
treated embryos (Figure 2a), suggesting that cardiosulfa does
not affect early heart development in zebrafish.
Figure 1. Effects of cardiosulfa on zebrafish heart development.
a,b) Three-day-old zebrafish embryo untreated (a) and treated (b) with
20 mm cardiosulfa. Arrows in the insets indicate normal and abnormal
heart morphology; V=ventricle, A=atrium. c,d) Tg(cmlc2:GFP) trans-
genic zebrafish embryo (three days old) untreated (c) and treated (d)
with 20 mm cardiosulfa. Heart marked with green fluorescent protein
(GFP). e) Effect of heart development upon exposure of embryos to
20 mm cardiosulfa during the times indicated by bars; thereafter, the
embryos were transferred to the fresh media. Embryos were observed
at 96 hours post-fertilization (hpf).
To further examine cardiovascular malformations,
Tg(cmlc2:GFP) transgenic zebrafish embryos were treated
with 20 mm of cardiosulfa. Because cmlc2 (cardiac myosin
light chain 2) is expressed throughout the heart tube,[9] the
heart morphology of transgenic zebrafish can be clearly
visualized by using fluorescence microscopy.[10] As shown in
Figure 1c,d, serious heart deformation with a narrow and
elongated ventricle and atrium was observed in the exposed
embryos at 72 hpf (hours post-fertilization). The heart rates of
treated embryos were similar to those of untreated ones at
48 hpf (untreated embryo/treated embryo = 102:96 beats per
minute (bpm)), but it lowers as development proceeds (at
72 hpf, untreated/treated = 128:80 bpm; at 96 hpf, untreated/
treated = 140:40 bpm).
As experiments of whole-mount in situ hybridization are
difficult to perform at late developmental stages, expression
patterns of heart-related proteins after 48 hpf were analyzed
by using Tg(cmlc2:GFP) transgenic embryos. By 54 hpf, the
7810
www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 7809 –7812

Angewandte
Chemie
Figure 3. Gene expression profiling of 24 h-, 48 h-, 60 h-, and 72 h-old
zebrafish embryos. a) Data represent increased (red) or decreased
(green) gene expression that has changed more than twofold at
various time points after exposure to 20 mm cardiosulfa at the initiation
of experiments. Relative gene expression was obtained using zebrafish
44 K DNA chips. b) RT-PCR data for untreated and treated embryos
exposed to 20 mm cardiosulfa at the initiation of experiments.
Figure 2. Expression of heart-specific genes and proteins in whole
embryos. a) cmcl2 and vmhc expression analyzed by whole-mount
in situ hybridization at 48 hpf. b) Views of Tg(cmlc2:GFP) transgenic
zebrafish embryos at 54 hpf and c) 60 hpf. Embryos were exposed to
20 mm cardiosulfa at the initiation of experiments.
arising from microarray analysis (Figure 3b). As it is known
that activation of AhR-mediated signaling pathways that
leads to expression of genes such as cyp1a and cyp1b1
produces cardiovascular malformations in fish, birds, and
mice,[14] we surmised that the effects of cardiosulfa on heart
development might occur by an AhR-mediated mechanism.
To gain more insights into the mode of action of
cardiosulfa, two experiments using (À)-epigallocatechin-3-
gallate (EGCG) and AhR2-morpholino (MO) antisense were
carried out. Because Hsp90 is essential for the activation of
AhR-mediated transcription of cyp1a and cyp1b1 genes, an
inhibitor of Hsp90, such as EGCG, suppresses AhR-mediated
signaling pathways.[15] Consequently, it is expected that
administration of an inhibitor for Hsp90 would cause
normal heart development in cardiosulfa-exposed embryos.
The effect of EGCG on heart development in zebrafish
embryos was determined. The results show that 150 mm of this
compound does not affect zebrafish development (Figure 4c).
Next, embryos were exposed to both 150 mm EGCG and
15 mm cardiosulfa at the initiation of the experiments;
impaired heart development was not observed in embryos
exposed to both compounds (Figure 4d). Changes of the
expression of AhR target genes in embryos treated with
cardiosulfa and EGCG were then examined by RT-PCR.
Expression level of these genes in embryos treated with both
compounds was significantly reduced in comparison with that
in embryos treated only with cardiosulfa (Supporting Infor-
mation, Figure S2).
hearts of embryos exposed to cardiosulfa are phenotypically
normal (Figure 2b). However, at 60 hpf, the exposed embryo
is a narrow and elongated tube even though no significant
change takes place in the expression patterns of heart-related
proteins. These results demonstrate that cardiosulfa does not
induce changes in the expression patterns of heart-related
genes, such as cmlc2 and vmhc, by 54 hpf, but that it does
cause abnormalities in the heart at later stages of develop-
ment.
Understanding the mode of action of cardiosulfa should
lead to important insights into the mechanisms involved in
heart development and disease states. We carried out gene
expression profiling by using zebrafish 44 K DNA chips to
assess the effects of cardiosulfa on heart development. Gene
expression profiles were obtained on 24 h-, 48 h-, 60 h-, and
72 h-old zebrafish after early exposure of the embryos to
20 mm of this compound. It was observed that known house-
keeping genes, such as b-actin, were not affected by cardio-
sulfa treatment. Importantly, this sulfonamide induces a high
expression level of genes involved in aryl hydrocarbon
receptor (AhR)-mediated signaling pathways. For example,
cardiosulfa causes the remarkable expression of cyp1a
(cytochrome P450 1A), cyp1b1 (cytochrome P450 1B1) and
cyp1c1 (cytochrome P450 1C1), members of the aryl hydro-
carbon receptor (AhR) gene battery (Figure 3a and Support-
ing Information, Table S2).[12] Furthermore, aryl hydrocarbon
receptor 2 gene (ahr2) displays a subtle but significant
induction in the exposed embryos. However, gene expression
of Hsp90, which is associated with AhR, is not affected by
cardiosulfa.[13]
The observed rescue of heart phenotype by EGCG may
provide some evidence for heart deformation induced by
cardiosulfa possibly by an AhR-mediated mechanism. How-
ever, as EGCG is known to have other functions than
inhibition of Hsp90,[16] we also performed knockdown experi-
ments using AhR2-MO antisense to more directly examine
the effect of cardiosulfa on an AhR2-mediated signaling
pathways. AhR2-MO was microinjected into the zebrafish
embryos at the initiation of experiments; embryos treated
with AhR2-MO did not show any effect on heart develop-
ment (Figure 5c). Importantly, abnormal heart morphology
To obtain additional evidence for the cardiosulfa-pro-
moted changes in expression levels of these genes, reverse-
transcription–polymerase chain reactions (RT-PCR) using
treated and untreated embryos were carried out. The results
of RT-PCR measurements of the genes cyp1a, cyp1b1, cyp1c1,
ahr2, and hsp90 were generally consistent with the results
Angew. Chem. Int. Ed. 2009, 48, 7809 –7812
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7811

Communications
findings that illustrate the value of chemical biology
approaches as tools for the elucidation of differentiation
and development processes.[17]
Received: May 4, 2009
Revised: July 25, 2009
Published online: September 15, 2009
Keywords: chemical genetics · combinatorial chemistry ·
.
screening · sulfonamides · zebrafish
[1] a) K. S. Warren, M. C. Fishman, Am. J. Physiol. 1998, 275, H1 –
H7; b) J. T. Shin, M. C. Fishman, Annu. Rev. Genomics Hum.
Genet. 2002, 3, 311 – 340.
Figure 4. Rescue of heart phenotype by EGCG. Three-day-old embry-
[2] a) J. S. Eisen, Cell 1996, 87, 969 – 977; b) P. W. Ingham, Hum.
Mol. Genet. 1997, 6, 1755 – 1760.
[3] a) B. R. Stockwell, Nat. Genet. 2000, 1, 116 – 125; b) J. Yeh, C. M.
Crews, Dev. Cell 2003, 5, 11 – 19; c) L. I. Zon, R. T. Peterson,
Nat. Rev. Drug Discovery 2005, 4, 35 – 44.
onic zebrafish a) untreated, b) exposed to 15 mm cardiosulfa,
c) exposed to 150 mm EGCG, and d) exposed to both 150 mm EGCG
and 15 mm cardiosulfa.
[4] a) R. T. Peterson, B. A. Link, J. E. Dowling, S. L. Schreiber,
Proc. Natl. Acad. Sci. USA 2000, 97, 12965 – 12969; b) S. M.
Khersonsky, D. W. Jung, T. W. Kang, D. P. Walsh, H. S. Moon, H.
Jo, E. M. Jacobson, V. Shetty, T. A. Neubert, Y. T. Chang, J. Am.
Chem. Soc. 2003, 125, 11804 – 11805; c) R. T. Peterson, J. D.
Mably, J.-N. Chen, M. C. Fishman, Curr. Biol. 2001, 11, 1481 –
1491.
[5] a) K. J. Engel, Pharmaceutical Substances: Synthesis patents,
applications, 4th ed., Thieme, Stuttgart, 2000; b) D. Vullo, M.
Franchi, E. Gallori, J. Antel, A. Scozzafava, C. T. Supuran, J.
Med. Chem. 2004, 47, 1272 – 1279.
[6] D. L. Flynn, J. Z. Crich, R. V. Devraj, S. L. Hockerman, J. J.
Parlow, M. S. South, S. Woodard, J. Am. Chem. Soc. 1997, 119,
4874 – 4881.
[7] P. Fernandez-Ferri, A. Ubeda, I. Guillꢀn, J. Lasri, M. E.
Gonzꢁlez-Rosende, M. Akssira, J. Sepffllveda-Arques, Eur. J.
Med. Chem. 2003, 38, 289 – 296.
[8] Y.-K. Yang, S.-K. Ko, I. Shin, J. Tae, Nat. Protoc. 2007, 2, 1740 –
1745.
Figure 5. Effect of AhR2-MO on heart development of embryos treated
with cardiosulfa. Three-day-old embryonic zebrafish a) after exposure
to vehicle, b) exposed to 15 mm cardiosulfa, c) after exposure to AhR2-
MO, and d) after exposure to both AhR2-MO and 15 mm cardiosulfa.
[9] D. Yelon, S. A. Horne, D. Y. R. Stainier, Dev. Biol. 1999, 214,
23 – 37.
[10] C. G. Burns, D. J. Milan, E. J. Grande, W. Rottbauer, C. A.
MacRae, M. Fishman, Nat. Chem. Biol. 2005, 1, 263 – 264.
[11] D. Y. R. Stainier, M. C. Fishman, Dev. Biol. 1992, 153, 91 – 101.
[12] M. E. Hahn, Comp. Biochem. Physiol. C 1998, 121, 23 – 53.
[13] D. R. Bell, A. Poland, J. Biol. Chem. 2000, 275, 36407 – 36414.
[14] a) H. M. Handley-Goldstone, M. W. Grow, J. J. Stegeman,
Toxicol. Sci. 2005, 85, 683 – 693; b) H. M. Handley-Goldstone,
J. J. Stegeman, Drug Metab. Rev. 2006, 38, 261 – 289; c) S. A.
Carney, J. Chen, C. G. Burns, K. M. Xiong, R. E. Peterson, W.
Heideman, Mol. Pharmacol. 2006, 70, 549 – 561.
[15] a) Z. Yin, E. C. Henry, T. A. Gasiewicz, Biochemistry 2009, 48,
336 – 345; b) D. Hughes, J. B. Guttenplan, C. B. Marcus, K.
Subbaramaiah, A. J. Dannenberg, Cancer Prev. Res. 2008, 1,
485 – 493.
[16] H.-K. Na, Y.-J. Surh, Food Chem. Toxicol. 2008, 46, 1271 – 1278.
[17] a) D. R. Williams, S.-K. Ko, S. Park, M.-R. Lee, I. Shin, Angew.
Chem. 2008, 120, 7576 – 7579; Angew. Chem. Int. Ed. 2008, 47,
7466 – 7469; b) D. R. Williams, M.-R. Lee, Y.-A. Song, S.-K. Ko,
G.-H. Kim, I. Shin, J. Am. Chem. Soc. 2007, 129, 9258 – 9259;
c) D. R. Williams, G.-H. Kim, M.-R. Lee, I. Shin, Nat. Protoc.
2008, 3, 835 – 839; d) D. P. Walsh, Y. T. Chang, Chem. Rev. 2006,
106, 2476 – 2530; e) S. Ding, P. G. Schultz, Nat. Biotechnol. 2004,
22, 833 – 840.
was not observed in embryos exposed to both substances
(Figure 5d). The results obtained from RT-PCR show that
expression of ahr2 and cyp1a in embryos treated with AhR2-
MO or both substances is suppressed in comparison with that
in embryos treated only with cardiosulfa (Supporting Infor-
mation, Figure S3). The results obtained from experiments
using EGCG and AhR2-MO suggest that the effect of
cardiosulfa on abnormal heart development is likely to be a
consequence of its activation of AhR2. Alternatively, it is
possible that cardiosulfa activates xenobiotic metabolizing
enzymes owing to its toxicity, and this may be responsible for
heart deformation. Further work will focus on understanding
detailed mechanism of abnormal heart development elicited
by cardiosulfa.
In conclusion, we have identified a novel small molecule
that impairs zebrafish cardiovascular development and func-
tion, and evokes a striking edema response in the pericardial
and yolk sac. This compound can be used to unveil
mechanism and pathways relevant to heart disease. The
observations made in this investigation build upon recent
7812
www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 7809 –7812