Wt1 and Epicardial Fate MappingNovelty and Significance
Rationale: The embryonic epicardium is a crucial cell source of the cardiac fibrous skeleton as well as of the coronary system. Genetic lineage tracing systems based on Wt1 regulatory sequences provided evidence that epicardium-derived cells also adopt a myocardial fate in the mouse.
Objective: To define the adequacy of Wt1-based lineage tracing systems for epicardial fate mapping.
Methods and Results: Using in situ hybridization analysis and immunofluorescence on tissue sections, we detected endogenous expression of Wt1 mRNA and Wt1 protein in the proepicardium and epicardium and also in endothelial cells throughout cardiogenesis. Expression analysis of a sensitive GFP reporter showed that recombination mediated by cre recombinase in the Wt1creEGFP line occurs randomly and sporadically in all cells of the embryo. Recombination in cardiomyocytes was found in the linear heart tube before establishment of a (pro)epicardium. In contrast, the tamoxifen-inducible Wt1creERT2 mouse line mediated poor and variable recombination in the epicardium. Recombination in cardiomyocytes was not detected in this case.
Conclusions: Frequently used Wt1 based cre-mediated lineage tracing systems are not suitable for epicardial fate mapping because of endogenous endothelial expression of Wt1, ectopic recombination (Wt1creEGFP), and poor recombination efficiency (Wt1creERT2) in the developing heart. We conclude that claims of a cardiomyocyte fate of epicardial cells in the mouse are not substantiated.
The outer lining of the mature vertebrate heart, the epicardium, develops after cardiac looping from a cluster of mesothelial cells of the septum transversum. Groups of cells detach from this proepicardium, adhere to the myocardium, and spread out to form a continuous epithelial layer.1 Cell lineage tracings performed by retroviral reporter constructs and chimera analysis in the chick have shown that a subset of epicardial cells undergoes an epithelial–mesenchymal transition differentiates within the subepicardial space and myocardium into interstitial and perivascular fibroblasts and smooth muscle cells (SMCs) of the coronary arteries.2–5 Some reports also suggested an additional coronary endothelial fate of epicardium-derived cells in birds.6,7
Two independent lineage tracings by cre/loxP technology confirmed fibroblasts and SMCs as epicardial derivatives but suggested an additional myocardial fate of epicardial cells in the mouse. Using a cre knock-in allele of the T-box gene 18 (Tbx18), cardiomyocytes of the left ventricle and the interventricular septum were recognized as epicardial descendants,8 and lineage tracing with two cre knock-in alleles of the Wilms tumor 1 gene (Wt1; Wt1creEGFP, Wt1creERT2) detected a substantial contribution of epicardial cells to the myocardium of the chambers and the interventricular septum starting from embryonic day (E) 10.5.9 Furthermore, using the same two Wt1cre lines, a myocardial differentiation of epicardial cells after myocardial infarction subsequent to pretreatment with thymosin β4 was demonstrated.10 However, other cre/loxP-based epicardial lineage tracing efforts did not find evidence for a myocardial fate of epicardial cells in the mouse.11–13 In addition, it was recently reported that Tbx18 exhibits endogenous expression not only in the proepicardium and epicardium but also in cardiomyocytes of the left ventricle and the interventricular septum starting at E10.5, ie, exactly in the regions claimed to be home of epicardium-derived cardiomyocytes in the study by Cai et al.14
Given the enormous medical interest in a potential myocardial fate of epicardial cells in mammals, we wished to reevaluate the adequacy of previously used Wt1-based cre/loxP lineage tracing systems. We provide evidence by expression and reporter analyses that an endothelial and myocardial fate of epicardial cells cannot be demonstrated in the mouse using this genetic tool.
Mice and Genotyping
Mice with a knock-in of the cre recombinase gene in the Wt1 locus (Wt1tm1(EGFP/cre)Wtp or Wt1creEGFP),9 mice with a knock-in of the tamoxifen inducible cre-recombinase in the Wt1 locus (Wt1tm2(cre/ERT2)Wtp or Wt1creERT2),9 and the double fluorescent cre reporter line (Gt(ROSA)26Sortm4(ACTB-tdTomato,-EGFP)Luo/J or Rosa26mTmG)15 were obtained from the Jackson Laboratory (Bar Harbor, ME). All mice were maintained on an outbred (NMRI) background. Animal care was in accordance with national and institutional guidelines. Detailed Methods are provided in the Online Supplement.
Wt1 Expression in the Heart Is Not Restricted to Epicardial Cells
Genetic lineage tracing by the cre/loxP technology relies on the fact that cre is expressed only in the precursor tissue but not in the differentiated cell type for which contribution should be tested. For Wt1-based epicardial fate mapping systems, it requires that Wt1 is expressed in the (pro)epicardium but not in any other cell type of the developing and mature heart. Cardiac expression of Wt1 has been reported for the (pro)epicardium, but also for cells of unknown nature in the subepicardium and the interventricular septum,16–18 prompting us to reinvestigate Wt1 expression during cardiac development (Figure 1).
In situ hybridization analysis on embryo sections did not detect Wt1 mRNA in E8.5 hearts (Online Figure I). At E9.5, strong Wt1 expression was found in the proepicardium and in the few epicardial cells that have settled onto the myocardium in the atrioventricular sulcus. At E14.5, the entire epicardial layer was positive for Wt1, as were individual cells in the interventricular septum and in the subepicardium and myocardium of the ventricles (Figure 1A, B).
Cardiac expression of Wt1 protein was not found in E8.5 embryos by immunofluorescence analysis (Online Figure I). At E9.5, high levels of Wt1 protein were confined to nuclei of the proepicardium and the first epicardial cells in the sulcus region. Weak Wt1 expression was occasionally observed in nuclei of the myocardium and endocardium (Figure 1C). At E14.5, all epicardial cells were positive for nuclear Wt1 protein, whereas in the interventricular septum the subepicardium and myocardium of the ventricles individual Wt1-positive cells were detected. Expression in the myocardium occurred mostly in tight association with luminal spaces (Figure 1D). Coexpression of Wt1 and the endothelial marker isolectin B419 at E9.5 and E14.5 confirmed endogenous expression of Wt1 in endothelial cells of the endocardium and/or of the coronary vasculature (Figure 1E–G). Because of technical constraints (antibodies against Wt1 and the two well-established nuclear myocardial marker proteins Nkx2.5/Prox1 were all raised in rabbits), it was not possible to unambiguously demonstrate that Wt1+ isolectin B4− cells in the E9.5 myocardium and the E14.5 subepicardium (open triangle in Figure 1E, F) are myocardial in nature. Thus, Wt1 mRNA and Wt1 protein expression are not restricted to the (pro)epicardium in the developing heart; they are also found in differentiated endothelial cells. A weak myocardial expression at E9.5 cannot be excluded at this point.
Wt1creEGFP Mediates Sporadic and Ectopic Recombination in the Heart and the Rest of the Embryo
Although endogenous expression of Wt1 in endothelial cells of the developing heart does not allow a Wt1-based lineage tracing system to claim an epicardial origin of this cell type, fibroblasts, SMCs, and (possibly) cardiomyocytes still may be analyzed for such a descent using this genetic tool. However, this requires that the cre activity pattern recapitulates endogenous expression of Wt1 in a faithful manner. To stringently test if the Wt1creEGFP line that was previously used for this approach fulfills this criterion,9 we performed a careful analysis of the recombination activity of this line during cardiogenesis using the Rosa26mTmG mouse line as a sensitive reporter. In this reporter line, recombination is easily visualized by bright membrane-bound green fluorescent protein (GFP) expression replacing a membrane-bound red fluorescent protein, and anti-GFP immunofluorescence analysis on sections additionally allows reliable cellular resolution of cre recombination events.15
To our surprise, Wt1creEGFP/+Rosa26mTmG/+ embryos exhibited highly variable patterns of green fluorescence in the entire embryo at E8.5 when neither Wt1 mRNA nor Wt1 protein was detected (Figure 2A–C, Online Figure I). GFP expression analysis on embryo sections confirmed the stochastic presence and random distribution of GFP-positive cells in the heart at this stage. Recombination occurred in cardiomyocytes and endocardial cells, as shown by coexpression with the myocardial marker Nkx2.5 and the endothelial marker isolectin B4, respectively (Figure 2D–F, Online Figure II). At E9.5, GFP expression did not match endogenous Wt1 expression but was again highly variable in all embryonic tissues (Online Figures I and III). GFP-positive cardiomyocytes and endocardial cells were randomly distributed throughout the heart and also were found in regions where the epicardium had not yet settled (Online Figure III). Analysis of E12.5 and E18.5 embryos revealed a similar variability of GFP expression in all embryonic tissues; variable numbers of cardiomyocytes and endothelial cells were positive for GFP in the heart. At E18.5, we detected additional recombination in the heart in a fraction of periostin-expressing fibroblasts and Acta2-expressing SMCs (Online Figures IV and V). We conclude that Wt1creEGFP-mediated reporter activity does not reflect endogenous expression of Wt1, but occurs randomly and sporadically in all embryonic tissues including the myocardium, endocardium, and vessels of the heart.
Wt1creERT2 Recombines Inefficiently in the Epicardium and Its Cellular Derivatives
Wt1creERT2 represents a second Wt1 allele that was used for epicardial fate mapping as well as for “epicardium”-specific gene deletion experiments.10,20,21 This allele mediates expression of a fusion protein of cre with a variant of the estrogen receptor that allows activation of cre activity by administration of tamoxifen in a temporally controlled fashion.9 Our initial tests revealed that injection of high doses of tamoxifen before E11.5 results in embryonic lethality. Injection of 4 mg of tamoxifen intraperitoneally to pregnant dams at E11.5 led to earliest and highest recombination in our experience, although this procedure resulted in premature delivery at E16.5.
Analysis of GFP expression in Wt1creERT2/+Rosa26mTmG/+ hearts at E16.5 revealed low and highly variable levels of recombination in the heart. Epicardial recombination was incomplete, and limited GFP expression was detectable throughout the ventricular compact myocardium and the interventricular septum (Figure 3A–C). Double immunofluorescence analysis for GFP and MF20 failed to detect recombination within cardiomyocytes, whereas cells double-positive for GFP and isolectin B4 (endothelial cells) and GFP and Acta2 (SMCs) revealed rare recombination events in both cell types (Figure 3D–F). Thus, the Wt1creERT2 line is inadequate to follow the cellular descendents of the (pro)epicardial entity. The lack of recombination in cardiomyocytes indicates that these cells are not epicardium-derived or, alternatively, that Wt1-expressing cardiomyocytes are not present after E10.5, even though the low recombination efficiency may conceal some rare events.
Our results show that Wt1 expression in the heart is not restricted to the epicardium and that available Wt1cre knock-in lines do not faithfully recapitulate endogenous expression of Wt1 (Wt1creEGFP) and recombine poorly (Wt1creERT2), respectively.
Expression analyses of Wt1 on the level of mRNA and protein reported on subepicardial expression of the gene/protein in addition to strong (pro)epicardial expression during cardiac development. It was suggested that expression occurs in epicardium-derived cells and, hence, that epicardial cells maintain Wt1 expression after they have left the epicardial continuity and become mesenchymal.16–18 Our expression studies do not support this notion but strongly argue that subepicardial expression of Wt1/Wt1 is mainly found in endothelial cells of the endocardium and the coronary vasculature. Expression levels in endothelial cells are lower than in the epicardium but are likely to comprise the majority of coronary vessels that form as a plexus that grows from the sinus venosus toward the apex of the heart from E11.5 to E14.5. Notably, Wt1 expression is reactivated in endothelial cells of the coronary vasculature after myocardial infarction,22 indicating a conserved regulatory module for endothelial Wt1 expression in development and regeneration. We detected weak expression of Wt1 protein but not Wt1 mRNA in the E9.5 myocardium. Alternative assays to in situ hybridization may address in the future whether low levels of Wt1 transcription actually occur in cardiomyocytes at this stage. Endogenous expression of Wt1 in the endothelium/endocardium, and possibly the myocardium in the developing heart clearly excludes Wt1-based cre lines to trace an epicardial contribution to myocardial, endothelial, and endothelium derived cells in the mouse heart. Nonepicardial expression of Wt1 also was reported in the zebrafish heart, which prompted those investigators to exclude this gene from epicardial fate mapping efforts in this species.23
Based on the strong epicardial expression of Wt1, two cre lines have been developed to trace and manipulate epicardial cells and their descendants. Our present study clearly has shown that the Wt1creEGFP line mediates sporadic and highly variable recombination in the heart as well as in the rest of the embryo at stages when a (pro)epicardium is not present in the embryo, and in a pattern that is incompatible with endogenous Wt1 expression. The underlying reason for this random activation of cre activity in this line remains unclear. We have excluded the presence of a neo cassette, which might interfere with proper transcriptional activation of cre. At present, we assume that integration of the targeting vector was imprecise or incorrect placing cre under different regulatory elements. In contrast, recombination efficiency obtained with the tamoxifen-inducible Wt1creERT2 line in epicardial cells was extremely low, allowing for following the fate of a small set of epicardium-derived cells only. Furthermore, epicardial fate analysis was restricted to stages E11.5 onward because injection of tamoxifen at earlier stages resulted in embryonic death, as reported in other studies.24,25
We conclude that an epicardial origin of myocardial and endothelial cells in the heart cannot be deduced using Wt1-based cre/loxP technology. Our data additionally call for cautious interpretation and reinvestigation of data obtained by conditional gene deletion experiments using Wt1-based cre lines.
Sources of Funding
This work was supported by grants from the German Research Foundation (D.F.G.) for the Cluster of Excellence REBIRTH (From Regenerative Biology to Reconstructive Therapy) and for the Clinical Research Group KFO136 at Hannover Medical School (A.K.).
Brief UltraRapid Communications are designed to be a format for manuscripts that are of outstanding interest to the readership, report definitive observations, but have a relatively narrow scope. Less comprehensive than Regular Articles but still scientifically rigorous, BURCs present seminal findings that have the potential to open up new avenues of research. A decision on BURCs is rendered within 7 days of submission.
The online-only Data Supplement is available with this article at http://circres.ahajournals.org/lookup/suppl/doi:10.1161/CIRCRESAHA.112.273946/-/DC1.
Non-standard Abbreviations and Acronyms
- embryonic day
- green fluorescent protein
- smooth muscle cell
- T-box gene 18
- Wilms tumor 1
- Received May 16, 2012.
- Revision received May 25, 2012.
- Accepted June 1, 2012.
- © 2012 American Heart Association, Inc.
- Gittenberger-de Groot AC,
- Vrancken Peeters MP,
- Mentink MM,
- Gourdie RG,
- Poelmann RE
- Wilm B,
- Ipenberg A,
- Hastie ND,
- Burch JB,
- Bader DM
- Merki E,
- Zamora M,
- Raya A,
- Kawakami Y,
- Wang J,
- Zhang X,
- Burch J,
- Kubalak SW,
- Kaliman P,
- Izpisua Belmonte JC,
- Chien KR,
- Ruiz-Lozano P
- Grieskamp T,
- Rudat C,
- Ludtke TH,
- Norden J,
- Kispert A
- Moore AW,
- McInnes L,
- Kreidberg J,
- Hastie ND,
- Schedl A
- Smith CL,
- Baek ST,
- Sung CY,
- Tallquist MD
- Wagner KD,
- Wagner N,
- Bondke A,
- Nafz B,
- Flemming B,
- Theres H,
- Scholz H
- Kikuchi K,
- Gupta V,
- Wang J,
- Holdway JE,
- Wills AA,
- Fang Y,
- Poss KD
Novelty and Significance
What Is Known?
The embryonic epicardium is a source of trophic signals that affects the myocardium. It also is a source of cells that form the coronary vasculature and the fibrous skeleton of the heart.
During cardiac development, Wt1 is expressed in the proepicardium and epicardium.
Genetic lineage tracings based on Wt1 regulatory sequences have provided suggestive evidence that epicardium-derived cells also adopt a myocardial fate in the mouse.
What New Information Does This Article Contribute?
Wt1 expression is not restricted to epicardial cells. During cardiac development, the protein also is expressed in endothelial and endocardial cells and, weakly, in cardiomyocytes.
The Wt1creEGFP line mediates sporadic and highly variable recombination in all cardiac cell types, including cardiomyocytes, as well as ectopic recombination at extracardiac sites.
Wt1creERT2-mediated recombination in the epicardium is incomplete and highly variable but does not occur in late cardiomyocytes.
The epicardium is an epithelial monolayer that covers and mechanically protects cardiac muscle. During embryogenesis, the epicardium also acts as a signaling center that promotes myocardial growth and coronary plexus formation, and is a source of cells for the fibrous skeleton of the heart, smooth muscle cells, and fibroblasts of coronary vessels. Although these functions have been confirmed for all vertebrates, a genetic (cre/loxP-mediated) lineage tracing study based on regulatory elements of the epicardially expressed Wt1 gene have suggested that in the mouse, epicardial cells also substantially contribute to the myocardium. This finding is of high clinical relevance because it may open new avenues for deriving cardiomyocytes for myocardial cell therapy. Our study shows that Wt1 expression not only is restricted to the epicardium but also is expressed in differentiated cells types of the heart, namely endothelial cells and cardiomyocytes. Also, we provide evidence that the Wt1creEGFP line that was used for epicardial fate mapping before mediates ectopic and sporadic recombination in cardiomyocytes as well as in other cell types in which Wt1 is not expressed. In addition, the inducible Wt1creERT2 line recombines poorly in the epicardium, making it difficult to follow all the descendants of this tissue. Thus, cardiomyocyte fate of epicardial cells could not be substantiated using this approach.