SEARCH

SEARCH BY CITATION

Keywords:

  • epicardium;
  • myocardium;
  • endocardium;
  • bulbus arteriosus;
  • ventricle;
  • atrium;
  • valves;
  • fast scanning confocal imaging;
  • image rendering;
  • fhla

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Using the transposon-mediated enhancer trap (ET), we generated 18 cardiac enhancer trap (CET) transgenic zebrafish lines. They exhibit EGFP expression in defined cell types—the endocardium, myocardium, and epicardium—or in anatomical regions of the heart—the atrium, ventricle, valves, or bulbus arteriosus. Most of these expression domains are maintained into adulthood. The genomic locations of the transposon insertions were determined by thermal asymmetric interlaced polymerase chain reaction (TAIL-PCR). The expression pattern of EGFP in some CETs is unique and recapitulates expression of genes flanking the transposon insertion site. The CETs enabled us to capture the dynamics of the embryonic heart beating in vivo using fast scanning confocal microscopy coupled with image reconstruction, producing three-dimensional movies in time (4D) illustrating region-specific features of heart contraction. This collection of CET lines represents a toolbox of markers for in vivo studies of heart development, physiology, and drug screening. Developmental Dynamics 239:914–926, 2010. © 2010 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

The hearts of fish and amphibians have often been used as models for understanding the development of the mammalian heart. Therefore, it is important to understand their design as a way to understand the organization of the more complex mammalian heart (reviewed in Fishman and Stainier, 1994; Weinstein and Fishman, 1996; Chen and Fishman, 2000; Yelon, 2001; Moorman and Christoffels, 2003).

Zebrafish is a well-established model animal for studies in developmental biology and molecular mechanisms of human diseases. The early zebrafish heart tube consists of two cell layers, an outer myocardial layer responsible for contraction enveloping an inner endocardial layer, which connects to the vasculature (Fishman and Chien, 1997). As the two-chambered heart, which is comprised of a single atrium and ventricle, matures, the epicardial layer surrounds the myocardium (Serluca, 2008). Cardiac valves in the form of leaflets form between the chambers to prevent blood from flowing backwards (Stainier et al., 2002; Scherz et al., 2008). These cardiac valves together with the specialized myocardium form the cardiac conduction system propagating electrical signals through the heart. As the development of the zebrafish heart depends on genetic pathways similar to that of the mammalian heart, several human cardiovascular diseases such as human long QT syndrome (Arnaout et al., 2007; Hassel et al., 2008) have been modeled in zebrafish.

Several cardiac transgenic lines of zebrafish with expression of fluorescent proteins under the control of various promoters of known genes have been established using the traditional method of direct transgenesis. Tg(cmlc2:GFP), which specifically marks the myocardium, was the earliest to be generated (Huang et al., 2003). The core promoter of only 244 bp is sufficient to drive myocardium-specific expression, which allowed the generation of useful transgenic lines such as Tg(cmlc2:EGFP-ras), enabling visualization of myocardium morphogenesis as well as a functional calcium reporter line Tg(cmlc2:gCaMP) useful for monitoring the cardiac conduction system (Chi et al., 2008). Transgenic lines used to study the endocardium include Tg(tie2:GFP), Tg(fli1a:EGFP) with expression in endothelial cells of the entire vasculature (Motoike et al., 2000; Lawson and Weinstein, 2002), and Tg(gata1:dsRed) with GFP expression in blood cells, endocardium and myocardium (Traver et al., 2003). These studies emphasized the importance of tissue-specific transgenic lines revealing various anatomical elements and/or physiological processes taking place in the vertebrate heart for high-resolution imaging of heart development.

And yet the toolbox of transgenics for studies of the heart remains incomplete. In particular, it lacks transgenics expressing GFP in certain cell types and regions. Furthermore, in direct transgenics a transgene is usually expressed in an ectopic position or abnormal conformation and this could result in non-specific changes of the expression pattern later in life (Caldovic et al., 1999; Thummel et al., 2006).

Using the insertional Tol2 transposon-mediated transgenesis, we generated a variety of enhancer trap (ET) transgenic lines (Parinov et al., 2004; Choo et al., 2006). The transgenics expressing EGFP in a tissue-specific manner were selected in an unbiased forward genetic screen. After the transposon had been integrated into the genome, it was remobilized in later generations by injection of transposase mRNA. This resulted in a large number of novel ET lines (Garcia-Lecea et al., 2008; Ke et al., 2008; Kondrychyn et al., 2009; Vasilyev et al., 2009; Winata et al., 2009; reviewed in Korzh, 2007). Some of the transgenics expressed EGFP in the heart and their analysis demonstrated various tissue- and region-specific expression patterns. These research tools are particularly useful for studying events taking place in adult stages of life. Unlike direct transgenics, the ET lines faithfully report gene regulation within a specific region throughout a lifetime provided that the insertion remains unaffected by chromosomal changes that normally take place in old age (Laun et al., 2007). Therefore, ET transgenics are more reliable as a source of specific and labeled cell types during the lifetime of the animal. Given the fact that the zebrafish as a model system has been largely used in the fields of developmental biology and toxicology, the availability of ET transgenics opens the possibility to use the larval and adult zebrafish as a source of defined cell types, for example, to monitor changes taking place in specific cell populations during aging.

The acquisition, visualization, and analysis of three-dimensional (3D) time-series (4D data) collected from living biological samples makes it possible to understand complex dynamic movements at a microscopic scale. This requires the technical capability to successively acquire time series of two-dimensional slices at increasing depths over one or several time periods, which are rearranged to recover a 3D+time (4D) sequence (Liebling et al., 2006). This approach is very useful in combination with tissue-specific expression of fluorescent reporters in ET transgenics of semitransparent larvae of zebrafish. Here we report several novel transgenics with tissue- and region-specific distribution of EGFP in the developing zebrafish heart that represent a toolbox for the developmental biologist and/or cell biologist by providing an opportunity to study many if not all aspects of heart development in vivo and to analyze molecular events taking place in specific cell populations.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Overview of the Cardiac Enhancer Trap Lines

Within the framework of our efforts to explore the potential of the Tol2 transposon as a tool for transgenesis and large-scale regulatory genome analysis (Kondrychyn et al., 2009), we generated a subset of nearly 20 cardiac enhancer trap (CET) lines that exhibit EGFP expression in the whole embryonic heart or in a cell-type or region-specific manner (Table 1).

Table 1. Characterization of CET Linesa
GFP expressionLine no.Chr.LocationTagged gene
Embryonic (3 dpf)Adult
HeartOther sites of expression
  • a

    CET lines marked in bold are discussed in the text. The transposon insertion sites have been mapped on the chromosomes, with the nearest gene indicated. Some transposons are inserted into exons.

  • *

    , ET lines that are homozygous viable.

A. Cell type-specific transgenics
Endocardium and myocardiumOlfactory bulb, notochord, ear, skinEndocardium and myocardiumET33-mi103N.A.N.A.Repetitive element
EndocardiumKidney by 5dpfEndocardiumET33-1A22Exon*si:dkey-42i9.10 (steroid receptor-interacting SNF2 domain protein, srisnf21)
EndocardiumNeuromasts, eyes, ear, fins, vessels, sclerotome, ionocytesN.A.ET33-mi60A3Upstream (24.9 kb)lfng (lunatic fringe)
EndocardiumRecapitulates gbx2 expression in endocardium but absent from rest of vasculatureN.A.ET33-mi74A63′UTR (0.5 kb)gbx2 (gastrulation brain homeo box 2)
EndocardiumEyes, brain, vasculature, branchial arches, somitesN.A.ET33-mi842Intronpard3 (par-3 partitioning defective 3 homolog)
MyocardiumRecapitulates fhla expressionMyocardiumET33-mi3A14Promoter (2.2 kb)*fhla (four and a half LIM domains a)
MyocardiumHatching gland, branchial arch, skin, fins, intestineN.A.ET33-mi32B16Upstream (58.1 kb)LOC795124 (similar to disintegrin and metalloproteinase with thrombospondin motifs)
MyocardiumHatching glandN.A.ET33-mi36B12Downstream (0.6 kb)LOC562569 (similar to phospholipase C delta 3)
MyocardiumVasculature, eyes, ear, lateral line, somitesN.A.ET33-mi89B14IntronENSDARG00000061603 (similar to sorbin and SH3 domain containing 2 isoform 2)
MyocardiumIonocytes, eyes, ear, branchial arches, kidney, pronephros, somitesN.A.ET33-mi93A14Exon/intronzgc:63527 (similar to DDX26B)
MyocardiumNotochord, optic tectum, earN.A.GW60A16Upstream (45.0 kb)ctnnb1 (catenin (cadherin-associated protein), beta 1)
Epicardium, BA, ventral aortaBrain, branchial arch, enteric neuronsBA subendothelium, epicardiumET2724IntronENSDARG00000057801 (pard3a)
PEO, EpicardiumGut, neuromasts, sclerotome, skinEpicardium, atrial muscle, valvesET30A20Downstream (93.8 kb)kcnh5
B. Region-specific transgenics
AtriumKidney, ear, neuromasts,N.A.ET33-mi75C16Downstream (114.8 kb)ENSDARG00000046087 (similar to Bromodomain containing-gene)
AtriumEyes, brain, notochord, neural crestN.A.GW38A2Upstream (21.5 kb)ENSDARG00000057284 (similar to serine/threonine kinase Pim)
VentricleNeuromasts, eyes, ear, vessels (trunk), horizontal myoseptumAV and VB valvesGW64B8Downstream (4.0 kb)zgc:92785 (similar to Proteasome inhibitor PI31 subunit)
BA, myocardium at AV junction and subsets of myocardial cellsNilBA subendothelium, AV and VB valvesET31N.A.N.A.Repetitive element
Epicardium, BA, ventral aortaBrain, branchial arch, enteric neuronsBA subendothelium, epicardiumET2724IntronENSDARG00000057801 (pard3a)
BATail tipVB valvesGW2CN.A.N.A.Repetitive element

One subset of CET defines cell layers of the heart, including the endocardium, myocardium, and epicardium (Table 1A). Another CET subset defines various structural elements of the heart, including its chambers, the atrium and ventricle as well as the valve at the atrio-ventricular (A-V) boundary and bulbus arteriosus (BA), which is connected to the ventral aorta (Table 1B). Taken together, our collection of CETs defines many cell types and regions of the heart by expression of EGFP in specific domains. It represents a toolbox for in vivo studies of various aspects of heart development and function.

Transgenics With Cell Layer-Specific Expression

The endocardium is the innermost cell layer of the heart. It is exposed to the blood flow and interacts with the myocardium in the formation of cardiac valves. It may play an important role in the regulation of heart regeneration in adult zebrafish (Lepilina et al., 2006). Therefore, it is of great interest to study this cell layer to understand the differences in the regenerative capacity of the teleost and mammalian hearts. Until now, endocardial development has been studied using pan-endothelial transgenics, such as Tg(fli1a:EGFP) (Fig. 1A), which means that in these studies the endocardial-specific functions may remain masked. We identified four CETs, ET33-1A, ET33-mi60A, ET33-mi74A, and ET33-mi84A, with EGFP expression in the endocardium. In one of these transgenics (ET33-1A), EGFP is expressed only in the endocardium and is absent from the rest of the vasculature (Fig. 1C–N; Movie 1, which is available online, shows a time-lapse sequence of a beating ET33-1A heart at a single plane while Movie 2 represents the 4D reconstruction). The endocardium-specific EGFP expression aligns MF20-positive myocardium in the atrium and ventricle (Fig. 1D, K), and its presence in the BA increases from 3–7 dpf (Fig. 1F, H) as the endocardial cushion in the BA thickens. By 3 dpf, the endocardial cells invaginate into the inner space at the A-V junction (Fig. 1E, E′), acquiring a folded leaflet morphology as described by Scherz et al. (2008). Whole mount in situ hybridization (WISH) using anti-gfp probe on 3-dpf ET33-1A larvae confirmed endocardial localization of transcripts (Fig. 1I, J), which is also maintained in the adult heart (Fig. 1L–N).

thumbnail image

Figure 1. CETs that define development of the endocardium in vivo. A: Anterior vasculature in Tg(fli-GFP). B: Bright-field view of anterior body. C: Endocardium-specific EGFP expression in ET33-1A at low magnification. Arrow indicates fluorescence only in the heart. D, K: Confirmation of EGFP-positive cells in the endocardial layer as evidenced by double immunohistochemistry with MF20 (red), which marks the myocardium. EGFP-positive cells are in the inner layer: (D) 3 dpf, (K) 7 dpf. E–H: In vivo imaging of the endocardium at high magnification. G: High-resolution 3D reconstruction of the endocardium in vivo revealing the intricate network of endocardial cells. Arrowhead (E′) and arrows (F–H) point to endocardial leaflets at A-V junction and BA, which in the latter thickens from 3 dpf (F) to 7 dpf (H), respectively. I, J: Lateral (I) and ventral (J) views of the heart after WISH using anti-gfp probe. Arrows indicate the heart. L, M: Low-resolution fluorescent (L) and bright-field images (M) of adult heart. N: Adult heart sections immunostained with anti-EGFP and F-actin. All images (except A) are of ET33-1A. C, G: Fluorescent images. E, E′, F, H: Composite fluorescent/DIC images. A, atrium; BA, bulbus arteriosus; V, ventricle.

Download figure to PowerPoint

The insertion sites of the transposon identified genomic regions containing genes that may have a specific function in the developing endocardium (Table 1A). In some of these transgenics, endocardial EGFP expression starts in the heart field from 19 hr (ET-mi84A), whereas in others it appears at the onset of heartbeat (e.g., ET33-mi74A, Movie 3) or at 48 hpf in the endocardium-specific ET33-1A transgenic line. This sequence of events probably reflects the consecutive steps in the regulation of the molecular machinery involved in endocardial development. Thus, in combination these lines provide an opportunity to analyze various stages of endocardial development in vivo.

The myocardium is an intermediate layer of the heart. We identified six CETs with EGFP expression in the myocardium, ET33-mi3A, ET33-mi32B, ET33-mi36B, ET33-mi89B, ET33-mi93A, and GW60A. Similar to Tg(cmlc2:dsRed, the EGFP intensity in the ventricular myocardium of the 3-dpf ET33-mi3A is stronger than that of the atrium (Fig. 2A), as trabeculation takes place. The EGFP expression of ET33-mi3A co-localizes with MF20 antibody (Fig. 2G, H). Figure 2C–H as well as Movies 4 and 5 (rendered) clearly show the difference in cell morphology between atrial and ventricular myocardium. In the adult heart, EGFP expression in the ventricle remains stronger compared to other chambers (Fig. 2B).

thumbnail image

Figure 2. CETs that define development of the myocardium in vivo. A: EGFP-positive embryonic myocardium in the ventricle is thicker compared to that in the atrium. B: EGFP expression in adult heart sections of myocardium-specific ET33-mi3A. C, D: Single optical sections of ventricular myocardium in the live embryo at high magnification (lateral view). E: High-resolution 3D-reconstruction of myocardium live. F–H: Myocardium immunolabelled with EGFP/MF20(red)/DAPI(blue). A, C, D, E: In vivo imaging of live embryos at 3 dpf. I,J: Lateral and ventral view of larva after WISH with anti-gfp probe. K, L: Lateral and frontal view of larva after WISH with anti-fhla probe. All images are ET33-mi3A. A, atrium; BA, bulbus arteriosus; V, ventricle.

Download figure to PowerPoint

The mapping of insertion sites in the myocardial CETs reveal a number of genomic sites containing genes that may play a role during development of the myocardium (Table 1A). The gfp (Fig. 2I–J) expression pattern in the myocardium and adjacent tissues is recapitulated by the transcript of the gene fhla (Fig. 2K, L; Table 1A), which is adjacent to the insertion site in ET33-mi3A on Chr. 14. The gfp expression domains outside of the myocardium represent muscles associated with eyes and branchial arches (Fig. 2I, J).

As the external-most layer of the heart, the epicardium forms as the last of the heart layers, when its cells surround the myocardium. It has been shown that morphogenetic signaling from the epicardium is necessary for proper heart morphogenesis and development of cardiac vasculature (Männer et al., 2001; Mikawa and Gourdie, 2006). This outer heart layer develops from a transient structure, the proepicardial organ (PEO), which morphologically is reminiscent of a cluster of grapes (Serluca, 2008). Despite its essential roles, very few genes have been linked to the formation of the PEO and epicardium. Therefore, it remains the least understood cell layer of the heart. There is so far no report of an epicardial transgenic fish line.

In the embryonic ET27 transgenics, EGFP is expressed in a thin layer of spindle-shaped cells, which by 80 hpf sparsely overlays the cmlc2–dsRed-expressing myocardium (Fig. 3A–D). In the adult ET27, this domain of EGFP expression is present on the heart surface immediately above the layer of F-actin-positive cardiomyocytes (Fig. 3E) in the atrium, ventricle, and the BA containing smooth muscle. F-actin staining is evenly detected in all cardiomyocytes. This is in contrast to the staining for the marker of sarcomeric myosin heavy chain (MyHC) MF20, which stains brightly only the superficial layer of cardiomyocytes immediately below the epicardium and fades away in deep cardiomyocytes in the adult heart (see Fig. 6E′). This is the first report of an epicardium transgenic fish line.

thumbnail image

Figure 3. CETs that define the epicardium in vivo. A, B: ET27 with EGFP-positive cells covering the ventricle; note that EGFP expression in the BA precedes that in the epicardium; C, D: The EGFP-positive epicardial cells (arrow) enveloping the internal cmlc2 highlighted myocardium at high magnification. E: Cross-section of the adult ET27 heart at low magnification, with GFP-labelled epicardium on the external surface of the BA, atrium and ventricle (arrow). B–D: Double transgenic embryos of ET27 and Tg(cmlc2:dsRed). All except D are fluorescent images. D: Composite fluorescent/DIC images. A, atrium; BA, bulbus arteriosus; V, ventricle.

Download figure to PowerPoint

Taken together, the set of transgenics that we identify here can be used as research tools to study development of all three cell layers that form the heart.

Transgenics With Region-Specific Expression in Larvae

The heart of zebrafish consists of two chambers, an atrium and a ventricle. In addition, there are several transition regions that include, from the venous to arterial pole of the heart: the sino-atrial (SA) node, which acts as the pacemaker of the heart; the AV valve; the ventriculo-bulbar (V-B) valve; and the BA connecting the heart to ventral aorta. Several of these areas may represent more conserved elements of the ancestral cardiac tube, including the SA node, A-V node, and BA. It was noted that in the absence of well-defined morphological features, the BA, SA valve, and A-V valve were difficult to identify unambiguously in fish and amphibians (reviewed in Moorman and Christoffels, 2003). Therefore, the availability of transgenics expressing fluorescent proteins in these regions provides the possibility to initiate in vivo studies of these regions.

In ET31, EGFP expression observed prior to heart looping along the myocardium (data not shown) becomes restricted to the A-V junction by 40 hpf, in addition to a subset of weakly expressing myocardial cells (Movie 6; this time-lapse movie has been rendered for clarity resulting in Movie 7) and the BA. To confirm the localization of EGFP at the A-V junction, ET31 was crossed with Tg(cmlc2:dsRed) and expression in this double transgenic line was compared to that in another double transgenic line ET33-1A:Tg(cmlc2:dsRed). In the latter, there is an unlabeled group of cells flanked by the endocardial cells and the cmlc2 marked working myocardium (Fig. 4D, E). These cells were labeled in ET31 (Fig. 4B, C; white arrow). This group of myocardial cells may represent the specialized myocardium demonstrated by Milan et al. (2006) to be the slow conduction tissue. The specific expression of gfp was also detected at the A-V junction of ET31 (Fig. 4F–H). In the adult ET31 heart, GFP expression was maintained at the A-V valve, BA (Fig. 6E, E″), and V-B valves (Fig. 6E′).

thumbnail image

Figure 4. The CET line, ET31, marks a unique subset of myocardium at the early A-V valve. A: ET31 crossed with Tg(cmlc2:dsRed) showing a subset of myocardium positive for EGFP at the A-V junction and chambers (arrowhead). Note the EGFP expression at the BA. B, C: High magnification of A-V junction reveals differential GFP intensity between a different subset of myocardial cells. Arrows point to non-overlapped ET31 EGFP expression. D, E: High magnification of A-V junction lined with endocardial cells adjacent to cmlc2 marked myocardium (shown for comparison). Arrows point to A-V junction that ET31 marks. F–H: anti-gfp WISH demonstrates that the expression domain at the A-V node represents a ring. A–C: Double transgenic embryos of ET31 and Tg(cmlc2:dsRed). D, E, Double transgenic embryos of ET33-1A and Tg(cmlc2:dsRed). B, D: Fluorescent images. C, E: Composite fluorescent/DIC images to reveal cellular morphology. A, atrium; BA, bulbus arteriosus; A-V, atrio-ventricular valve; V, ventricle.

Download figure to PowerPoint

Three chamber-specific CETs, GW38A (Movie 8) and ET33-mi75C with atrial EGFP expression and GW64B with larval ventricle-specific EGFP expression (Fig. 5A–I; Movie 9), were identified. In the latter, the expression of gfp was also detected at the transcript level (Fig. 5B). The ventricle-specific expression was maintained at least until 7dpf (Fig. 5E,F), but by 1-month post-fertilization EGFP was detected at the A-V and V-B valves (Fig. 5G–I).

thumbnail image

Figure 5. Developmental changes in a CET line with early ventricle-specific expression (GW64B). A: EGFP expression at 3 dpf was found in the ventricle (ventral view). B: Same stage larvae stained by WISH using anti-gfp probe (lateral view). C, D, F, I: Double immunolabelling with MF20(red) performed on 3-dpf (C, D), 7-dpf (F), and adult heart section (I). C–F: The ventricular expression revealing trabeculated myocardium was maintained until 7 dpf. In the adult heart, EGFP expression was found at the A-V junction (G), and at the VB junction, where it defines valves (H, I). All images are GW64B. A, atrium; A-V, atrio-ventricular valve; BA, bulbus arteriosus; V, ventricle; VB, ventriculo-bulbar valve.

Download figure to PowerPoint

The BA-specific transgenics include the aforementioned ET27 (Fig. 6A–C′), ET31 (Fig. 6D–E″). and GW2C (Fig. 6F–G′) with an expression domain restricted to the arterial pole of the heart. In ET27, EGFP expression also encompasses the ventral aorta (Fig. 6A) and epicardium. The Hedgehog signaling pathway is required for development of the ventral aorta and mutants affected in this signaling pathway (smo and syu) lack ventral aorta. Indeed, analysis of EGFP expression pattern in the smo/ET27 composite embryo (Fig. 6B) revealed a reduction of EGFP expression with disappearance of the part of the expression domain corresponding to the ventral aorta. The remaining domain of EGFP expression adjacent to the ventricle probably represents the underdeveloped BA. This proves the utility of CET transgenics for in vivo analysis of developmental events in mutants.

thumbnail image

Figure 6. Developmental changes in CET lines with early BA-specific expression. A: EGFP expression in the BA and ventral aorta (VA) in wildtype 3-dpf ET27. B: smo mutation affects the ventral aorta, but not the domain of EGFP expression corresponding to BA. C, C′: ET27 adult heart sections shown at low (C) and high (C′) magnification. D: ET31 embryonic expression in BA. E, E′, E″: Cardiac valves; VB (E′) and AV (E″) marked in adult heart of ET31 shown on F-actin-stained sections. F: BA restricted embryonic expression of GW2C. G: GW2C adult heart sections at low magnification. G′, H: Serial sections of VB valve showing its different morphology at high magnification. I, I′: ET33-1A adult heart sections shown at low (I) and high (I″) magnification, for comparison of endocardal cell morphology in BA with ET27 (C′). A, B, D, F: Ventral view at 3dpf; C, C′: Sections of the adult heart immunostained for EGFP and MF20. E′, E″, H, I: Sections counterstained with F-actin. A-V, atrio-ventricular valve; BA, bulbus arteriosus; VB, ventriculo-bulbar valve; V, ventricle.

Download figure to PowerPoint

The expression of EGFP is maintained in the BA of adult zebrafish in all three CETs in a pattern similar to that in the endocardium-specific ET33-1A (Fig. 6C, C′, E, G, I, I′), though GW2C and ET31 exhibit weaker expression. Despite its similarity to endocardium, but unlike the endocardium in ET33-1A, this domain is limited to the BA only and, therefore, probably represents the BA-specific subendothelium (Hu et al., 2001). In addition, it is interesting to note that in the adult hearts of ET31 and GW2C, which display embryonic BA expression, EGFP is specifically localized to cells characteristic of valvular tissue (Fig. 6E–E″, G–H).

We have identified a set of CET transgenics that could be used to study the in vivo development of many elements of the embryonic, larval (Fig. 7) and adult heart of zebrafish (Table 1). Such vital markers should be useful for studies of the mechanisms of heart formation, development, and aging as well as hereditary diseases affecting the heart.

thumbnail image

Figure 7. Diagram of a 3-dpf zebrafish heart illustrating distribution of EGFP expression domains in the CET lines. In bold are the ET lines used to make the figures that appear in this report.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

The cardiac enhancer trap (CET) transgenics represent versatile research tools necessary to study most if not all aspects of cardiac development (septation is a notable exception) and function. Given the fact that the zebrafish heart could be studied in vivo with high resolution (Liebling et al., 2006, Scherz et al., 2008), studies with these tools may shed light on those problems of cardiac development where progress has been hampered due to an absence of tissue- or region-specific transgenics expressing fluorescent proteins or non-specific changes in expression of transgenes derived from direct transgenesis. These transgenics faithfully recapitulate transcriptional activity at a particular genomic site and/or specific cell lineage throughout the lifetime of the animal. Thus, they are reliable as sources of various cell types, which allow monitoring changes taking place in these cells or lineages or genomic sites during aging of animals. Therefore, when compared to direct transgenics these transposon-mediated ET lines probably represent the gold standard in terms of reliability of transgene expression throughout the lifespan of vertebrate animals.

The newly developed CET lines could find applications in addressing a number of problems of heart development and physiology. For example, until now analysis of the development of the cardiac endothelium has been accomplished using the pan-endothelial Tg(fli:GFP) or Tg(tie2:GFP) transgenic lines (Lawson and Weinstein, 2002; Motoike et al., 2000). The broad range of marker expression in these lines may not reflect elements of the developmental program specific for the cardiac endothelium. In contrast, the subset of endocardium-specific transgenics could be used to define distinct phases in the early development of the endocardium. In addition, the endocardium-specific ET33-1A could be used to uncover genetic differences between the endocardium and the rest of the vasculature. This adds to the research capability in the study of endocardium.

Although transgenics that mark cardiomyocytes are available, the new insertion trap lines may help to extend the analysis of developmental genes active during cardiomyocyte differentiation to several new candidate genes as well as genomic regions that harbor them. In particular, four out of six insertions (ET33-mi32B, ET33-mi36B, ET33-mi89B; ET33-mi93A; Table 1A) are in the vicinity of genes that are poorly characterized, whereas the other two could be used to study the roles of Wnt signaling (GW60A) or the LIM-domain-only gene fhla (ET33-mi3A) in the myocardium. The human homologue FHL1 expressed in skeletal and heart muscles is implicated in a number of diseases affecting muscles (Quinzii et al., 2008; Schessl et al., 2008; Shalaby et al., 2009; Windpassinger et al., 2008). Given the practical importance of such myocardium-specific genes as potential markers in studies of human cardiomyopathy and development of therapeutic approaches, their identification could be useful for further studies of myocardium development and disease.

Our current knowledge about the development of the epicardium is rather limited (Männer et al., 2001; Serluca, 2008). In the developing rodent heart, epicardial progenitors contribute to the cardiomyocyte lineage (Wessels and Pérez-Pomarez, 2004; Zhou et al., 2008), which suggests the presence of new cardiomyocytes in the superficial layer of the myocardium. Thus, it is noteworthy that the MF20 Mab staining of the adult heart revealed the highest level of MyHC expression in the superficial cardiomyocytes, where one may expect to see a population of recently born epicardium-derived cardiomyocytes (Fig. 6G, J; Wessels and Pérez-Pomarez, 2004; Zhou et al., 2008). This is in line with the idea that the epicardium acts as a source of new cardiomyocytes (Lepilina et al., 2006). Here we report the first transgenic line (ET27) showing EGFP-positive epicardium, which could be used to trace the development of these cells and expand the knowledge about in vivo development of this relatively poorly defined cell layer. Our analysis confirmed that the most external epicardial layer in the zebrafish, similar to higher vertebrates, forms by migrating over the myocardium surface after the formation of the two inner layers. This is reminiscent of what has been described during development of the pancreas, where the internal endocrine cells develop before the external exocrine cells (Field et al., 2003; Kardong, 2002; Wan et al., 2006), or where the external mesothelium develops after the inner layers of the epithelium and mesoderm in the swim bladder (Winata et al., 2009). Therefore, this developmental schedule of formation from the inside to the outside may represent a common element of a general mechanism during the development of all visceral organs.

It was noted earlier that in the absence of morphological features that clearly distinguish the BA, the S-A valve, and the A-V valve, these regions were difficult to identify unambiguously in fish and amphibians. They are of particular interest since they represent the most conserved elements of the ancestral cardiac tube and it would be useful to label them precisely, in particular for comparative evolutionary studies.

Cardiomyocytes differentiate locally into distinct chambers at a specific region along the linear heart tube, but remain nascent at the junction between the chambers (Moorman and Christoffels, 2003). Despite the importance of chamber maturation for cardiac function, only a few genes have been linked to the development of the two heart chambers very distinct in their morphology and function. The genomic information offered by the chamber-specific CETs (ET33-mi75C, GW38A, GW64B) may allow identification of such genes.

In higher vertebrates, the non-chamber myocardium is known to acquire characteristics of conduction tissue that are responsible for the transmission of electrical impulse, in contrast to the working myocardium that functions in cardiac contraction (Moorman et al., 1998). In the embryonic zebrafish, Milan et al. (2006) functionally defined the presence of a ring of specialized conduction tissue at the A-V junction. In ET31, there is a similar ring of EGFP-positive cells at the A-V junction that may represent these specialized myocardial cells. In addition, EGFP in the adult heart of ET31 is localized to non-chamber valve tissue. A conduction system transgenic line has not been generated so far. More detailed analysis, involving electrophysiological characterization, will be required to establish ET31 as a cardiac conduction transgenic line.

Our study represents the first attempt to present a collection of CET transgenics. It is noteworthy that all these “cardiac” transgenics were identified after an “off-shelf” screen of a much larger collection of ET transgenics for EGFP expression in the CNS (Kondrychyn et al., 2009). Unavoidably, presentation of this collection of CET lines raises many questions. In this preliminary report, we are able to answer only some of these. Importantly, given the fact that transposon insertions in most of these transgenics were mapped to specific genomic regions, information about the genomic environment in which the reporter genes are actively transcribed is provided. And even when these lines are not mapped due to insertions into repetitive DNA or technical limitations, they could be used as a source of specific cell types that could be analyzed further by methods of genomics and proteomics (Poon et al., unpublished data). If necessary, the transposons in these lines could be remobilized and, due to the phenomenon of local hopping novel ET transgenic lines with closely linked insertions and similar expression patterns, could be generated and mapped (Kondrychyn et al., 2009; Parinov et al., 2004; reviewed in Korzh, 2007).

Without doubt, many more answers will be found when this CET collection becomes a standard toolbox for developmental cardiologists similar to the first set of ET transgenics generated in this laboratory (Parinov et al., 2004; Choo et al., 2006). These early ET lines became, for instance, useful tools for researchers of cell differentiation in the lateral line of zebrafish (Hern´ndez et al., 2007; Nagiel et al., 2008; Nechiporuk and Raible, 2008; López-Schier and Hudspeth, 2006; Sarrazin et al., 2007), kidney (Vasyliev et al., 2009), and CNS (Bianco et al., 2008; Ke et al., 2007; Regan et al., 2008). Future studies of CET transgenics may be as fruitful.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Zebrafish

Zebrafish were maintained according to established protocols (Westerfield, 2003) in agreement with the IACUC regulations (Biological Resource Center of Biopolis, IACUC license no. 050096) and rules of the IMCB zebrafish facility. Embryos were staged in hours (hpf) or days (dpf) or months (mpf) post-fertilization. Pigmentation was inhibited with 0.2 mM 1-phenyl-2-thiourea (PTU) in egg water. The “off-shelf” screen for cardiac ET lines was performed using a collection of the ET lines generated and maintained in this laboratory (Parinov et al., 2004; Choo et al., 2006; Kondrychyn et al., 2009).

The ET(krt4:EGFP)sqet27, ET(krt4:EGFP)sqet30A, ET(krt4:EGFP)sqet31, ET(krt4:EGFP)sqet33-1A, ET(krt4:EGFP)sqet33-mi3A, ET(krt4:EGFP)sqet33-mi32B, ET(krt4:EGFP)sqet33-mi36B, ET(krt4:EGFP)sqet33-mi60A, ET(krt4:EGFP)sqet33-mi74A, ET(krt4:EGFP)sqet33-mi75C, ET(krt4:EGFP)sqet33-mi84, ET(krt4:EGFP)sqet33-mi89B, ET(krt4:EGFP)sqet33-mi93A, ET(krt4:EGFP)sqet33-mi103, ET(krt4:EGFP)sqgw-2C, ET(krt4:EGFP)sqgw-38A, ET(krt4:EGFP)sqgw-60A, and ET(krt4:EGFP)sqgw-64B lines are referred to as ET27, ET30A, ET31, ET33-1A, ET33-mi3A, ET33-mi32B, ET33-mi36B, ET33-mi60A, ET33-mi74A, ET33-mi75C, ET33-mi84, ET33-mi89B, ET33-mi93A, ET33-mi103, GW2C, GW38A, GW60A, and GW64B, respectively.

These CET lines are available from the authors upon request.

TAIL PCR

TAIL PCR was performed as described in Parinov et al. (2004).

RNA In Situ Hybridization and Immunohistochemistry

Whole mount in situ hybridization (WISH) was carried out as previously described (Korzh et al., 1998) using the gfp and fhla anti-sense RNA probes. The gfp probe was synthesized from the pBKCMV Tol2-EGFP enhancer trap construct using T3 RNA polymerase. To generate the fhla probe, 1,189 bp of the gene was cloned from 2-dpf zebrafish embryos into pGEMTeasy vector (Promega, Madison, WI). Photographs of WISH embryos were taken using a Leica dissecting photomicroscope (Leica, Wetzlar, Germany).

Fluorescence immunohistochemistry was performed on wholemount zebrafish embryos and on sections of adult zebrafish heart following protocols previously described (Korzh et al., 1998). The antibodies were used at the following dilutions: MF20 (Developmental Studies Hybridoma Bank, Iowa City, IA) at 1:20, GFP (Abcam, Cambridge, MA) at 1:200 and secondary antibodies Alexa 594 or Alexa 488 coupled goat-anti-mouse (Molecular Probes, Eugene, OR) at 1:500. Nuclei were counterstained with DAPI.

Adult hearts were fixed in 4% PFA, mounted in 1.5% agar for cryosectioning and 15-μm sections were collected for immunohistochemistry. For F-actin staining, sections were incubated with rhodamine phalloidin (Molecular Probes).

Confocal Imaging

Stained embryos and sections were imaged either with confocal microscopy Zeiss LSM5 (Carl Zeiss, Jena GmbH, Germany) or Olympus Fluoview (Olympus, Japan). For in vivo imaging, embryos were dechorionated at the selected stages, anaesthetized with 0.2% tricaine, and oriented ventral to the bottom by embedding them in 1% low-melting agarose (LMA) in embryo water on a confocal dish. Microscopic observations were done using a dissecting fluorescent microscope SZX12 (Olympus, Tokyo, Japan). To capture still images, embryos were treated with 2,3-butanedione monoxime (BDM; Sigma, St. Louis, MO) to arrest heart contraction before mounting in agarose for confocal scanning using an inverted Zeiss LSM 510 or Olympus Fluoview. For in vivo imaging of heart contraction in real time, movies of live zebrafish embryos were made using the slit scanning confocal microscope Carl Zeiss LSM 5 LIVE. Briefly, a Z-series of time-lapse movies at an interval of 2.5 μm were taken and processed for 4-D reconstruction with imaging software Imaris (Bitplane AG, Zurich, Switzerland).

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

We are grateful to personnel of the zebrafish facility of the IMCB for fish maintenance and Sudipto Roy for the use of a Leica fluorescent dissecting photomicroscope. We are also thankful to Debbie Yelon, Ian Scott, and Nadia Rosenthal for suggestions, David Milan and Geoffrey Burn for the Tg(cmlc2:dsRed) line, and Uwe Straehle for extensive discussions. V.K.'s lab in the IMCB is supported by a research grant from the Agency for Science, Technology and Research (A-STAR) of Singapore.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
DVDY_22203_sm_SuppMov1.mov2125KMovie 1. Heart of the endocardium-specific ET33-1A transgenic line in vivo at 3 dpf in ventral view.
DVDY_22203_sm_SuppMov2.mov4107KMovie 2. Heart of the endocardium-specific ET33-1A transgenic line in vivo at 3 dpf in ventral view after rendering.
DVDY_22203_sm_SuppMov3.mov2956KMovie 3. Heart of the endocardium-specific ET33-mi74A transgenic line in vivo at 1 dpf in lateral view.
DVDY_22203_sm_SuppMov4.mov1177KMovie 4. Heart of the myocardium-specific ET33-mi3A transgenic line in vivo at 3 dpf in ventral view.
DVDY_22203_sm_SuppMov5.mov1073KMovie 5. Heart of the myocardium-specific ET33-mi3A transgenic line in vivo at 3 dpf in ventral view after rendering.
DVDY_22203_sm_SuppMov6.mov884KMovie 6. Heart of the atrio-ventricular junction-specific ET31A transgenic line in vivo at 3 dpf in ventral view.
DVDY_22203_sm_SuppMov7.mov967KMovie 7. Heart of the atrio-ventricular junction-specific ET31A transgenic line in vivo at 3 dpf in ventral view after rendering.
DVDY_22203_sm_SuppMov8.mov1264KMovie 8. Heart of the atrium-specific GW38A transgenic line in vivo at 3 dpf in frontal view.
DVDY_22203_sm_SuppMov9.mov174KMovie 9. Heart of the ventricle-specific GW64B transgenic line in vivo at 3 dpf in lateral view.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.