Skip to main content
NCBI home page
Cardiovascular Research logoLink to Cardiovascular Research
. 2015 Jan 5;105(3):271–278. doi: 10.1093/cvr/cvu269

Extracardiac control of embryonic cardiomyocyte proliferation and ventricular wall expansion

Hua Shen 1, Susana Cavallero 1, Kristine D Estrada 1, Ionel Sandovici 2,3, S Ram Kumar 4, Takako Makita 5, Ching-Ling Lien 5, Miguel Constancia 2,3, Henry M Sucov 1,*
PMCID: PMC4366578  PMID: 25560321

Abstract

Aims

The strategies that control formation of the ventricular wall during heart development are not well understood. In previous studies, we documented IGF2 as a major mitogenic signal that controls ventricular cardiomyocyte proliferation and chamber wall expansion. Our objective in this study was to define the tissue source of IGF2 in heart development and the upstream pathways that control its expression.

Methods and results

Using a number of mouse genetic tools, we confirm that the critical source of IGF2 is the epicardium. We find that epicardial Igf2 expression is controlled in a biphasic manner, first induced by erythropoietin and then regulated by oxygen and glucose with onset of placental function. Both processes are independently controlled by retinoic acid signalling.

Conclusions

Our results demonstrate that ventricular wall cardiomyocyte proliferation is subdivided into distinct regulatory phases. Each involves instructive cues that originate outside the heart and thereby act on the epicardium in an endocrine manner, a mode of regulation that is mostly unknown in embryogenesis.

Keywords: Heart development, Epicardium, Cardiomyocyte proliferation, Endocrine signalling

1. Introduction

The midgestation period of mouse development, approximately between embryonic days E10 and 14, is a time of dramatic morphogenic change in many organ systems, including the heart. The mouse heart at E9.5 has already completed the specification and positioning of all four chambers, inflow and outflow tracts, and valves, but all domains are still quite rudimentary. By E14, the heart is essentially morphologically mature, with thickened ventricular walls, fully formed atrioventricular valves, septated outflow tract, and a functional coronary circulation. Further growth of the heart occurs after E14, but this is primarily enlargement with little change in morphogenic organization.

The ventricle wall, which is a primary focus of this study, accounts for the force of cardiac ejection. The ventricular wall at mouse E9.5 consists of a thin layer (1–2 cell diameters) of myocardium. By E14, the ventricle wall has expanded to 10–15 cell diameters and is often called the compact zone because of its tight organization. The importance of proper formation of the ventricular wall for midgestation embryonic growth is evident in the number of mouse gene mutations which compromise this process and which result in midgestation embryo lethality.1

The epicardium, which is the outer mesothelial layer of the heart, migrates onto the surface of the myocardium during the E9.5–10.5 period.2,3 Thus, the formation of the epicardium is coincident with the onset of midgestation ventricular wall expansion. Many observations, including the underdeveloped compact zone that results from genetic2,4 or surgical5,6 elimination of the epicardium, have shown that the epicardium is in fact required for ventricular wall cardiomyocyte proliferation. We showed in co-culture assays that epicardial cells secrete a soluble activity that induces embryonic cardiomyocyte proliferation7 and recently identified the primary mitogenic factor made by mouse embryonic epicardial cells as IGF2.8 We further showed that mouse embryos globally lacking IGF2, and those conditionally lacking the IGF receptors IGF1R and IR in the heart, suffer from a midgestation deficiency in ventricular wall cardiomyocyte proliferation and a corresponding deficiency in compact zone morphogenesis. Using in vitro co-culture assays, we showed that blocking IGF signalling with a selective receptor antagonist prevented the induction of cardiomyocyte proliferation by epicardial cells.8 These results are consistent with IGF2 being an epicardial mitogen in vivo during midgestation heart development, although we did not have the genetic tools at that time with which to confirm this model.

Many signalling pathways potentially regulate or intersect with epicardial mitogenic activity. We have long been interested in retinoic acid (RA), which functions via a heterodimeric nuclear receptor transcription factor consisting of one RAR and one RXR.9 Three separate genes encode distinct RAR isoforms, and likewise, three genes encode RXR isoforms. Mouse mutants globally lacking RXRα show a dramatically underdeveloped compact zone and die around E14.5.10 An additional phenotype present in Rxra−/− embryos is a transiently underdeveloped liver. We showed11 that the liver phenotype of Rxra−/− embryos is the consequence of a delay in the expansion and differentiation of definitive erythroid progenitors that reside in the liver at this time. This process is controlled by the cytokine erythropoietin (EPO),12 and we showed that liver Epo expression was greatly reduced, albeit transiently, in Rxra−/− embryos.11 We also showed that the Epo gene is a direct transcriptional target of RA signalling. The molecular explanation for the transient nature of the liver phenotype in Rxra−/− embryos is a transition in Epo expression that is initiated by the onset of placental function.11,13

In addition to their erythropoietic phenotype, Epo and Epo receptor (EpoR) global mutants die around E14 with an underdeveloped heart ventricular wall.14 The similar heart phenotypes seen in mouse mutants lacking EPO, EPOR, or RXRα, and the observation that the Epo gene is a direct downstream transcriptional target of RA signalling, suggested the possibility that compromised EPO signalling might be an explanation for the heart phenotype of Rxra mutants.15 The EpoR gene is expressed in the epicardium and endocardium, but the Epo ligand gene is not expressed in the heart.14 We proposed that EPO from the midgestation liver might be an upstream regulator of epicardial Igf2 expression,15 although this model was based primarily on in vitro assays and was not evaluated by in vivo genetic manipulations.

Using a battery of genetic reagents, in this study, we demonstrate the epicardial function of Igf2 in supporting ventricular wall formation and directly demonstrate the importance of EPO signalling in the epicardium to support epicardial Igf2 expression. However, the role of EPO in supporting epicardial Igf2 expression is transient. We show that midgestation initiation of placental function supplants the requirement for EPO signalling in epicardial Igf2 expression, just as it also supplants control of liver Epo expression by RA signalling. Importantly, the effects of placental function on heart growth occur through a regulated programme of epicardial Igf2 gene expression and not simply by provision of growth substrates directly to the myocardium.

2. Methods

2.1. Mice

All mouse lines have previously been described: Nkx2.5Cre,16 Mesp1Cre,17 Tbx18Cre,18 TCre,19 Myh6Cre,20 Tie2Cre,21 Rara403,22 conditional Igf2,23 conditional EpoR,24 and conditional R26R.25 In all experiments, mutant embryos were compared with littermate wild-type controls. Noon of the day of observation of a mating plug was defined as E0.5. Pregnant female mice were anaesthetized by isoflurane inhalation and euthanized by cervical dislocation prior to isolation of embryos or placental tissue. Animal procedures were reviewed and approved by the USC IACUC board (protocol 10080) and were followed in accordance with institutional guidelines.

2.2. Morphometric heart analysis

E14.5 embryos were fixed overnight in 4% PFA then processed for paraffin embedding. Serial sections were taken and stained with haematoxylin and eosin. Transverse sections at the level of the aortic valve were used for measurement of ventricular wall thickness at a position 45° clockwise (right ventricle) or counterclockwise (left ventricle) to the ventricular septum–apex axis.

2.3. Whole-mount PECAM immunohistochemistry

E14.5 hearts were fixed overnight in 4% PFA, permeabilized in PBS with 0.5% NP-40, dehydrated in a methanol series, incubated in methanol/hydrogen peroxide, rehydrated, and blocked in PBSST (PBS containing 5% BSA and 0.1% Triton X-100). The primary antibody was rat anti-mouse PECAM (BD, 1:100). Biotinylated goat anti-rat IgG (Santa Cruz, 1:500) was used followed by streptavidin-peroxidase complex (Vectastain ABC kit, Vector). Antibody and ABC reagent incubations were carried out in PBSST at 4°C overnight. Following overnight incubations, hearts were washed five times (1 h each at 4°C) with PBSST. Immunoreactivity was visualized with DAB substrate (Invitrogen).

2.4. In situ hybridization

Digoxygenin (DIG)-labelled probes were made as described previously.8 Briefly, embryos or cultured thorax segments were cryopreserved in 30% sucrose, embedded in OCT, and then cryosectioned transversely at 10 μm. Control and mutant sections were placed on the same slides to ensure identical experimental conditions. Sections were fixed in 4% PFA in PBS, rehydrated, and treated with proteinase K and then triethanolamine in acetic anhydride. Hybridization was performed at 65°C for at least 16 h. Unhybridized probe was removed by RNaseA digestion. Signal was detected by POD-coupled anti-DIG primary antibody (Roche) and TSAplus Fluorescent Substrate Kit (PerkinElmer).

2.5. Thorax culture

Wild-type embryos from an ICR background, or littermate TCre/Rara403 and control embryos, were dissected. The thorax section of each embryo was isolated by removal of the head and abdominal region (including the diaphragm) using dissecting forceps and transferred into DMEM medium containing 1% BSA with varying amounts of glucose in 12-well culture dishes. Tissue was incubated with shaking at 37°C for 1.5 h, then rinsed quickly with PBS and fixed with 4% PFA, and embedded in OCT for cryosectioning.

2.6. Cell culture

Mouse MEC1 cells8 were grown in DMEM containing high glucose with 10% FBS. After reaching 80% confluence, cells were seeded into 6-well culture dishes and were cultured to near confluence. Cells were then cultured in high (25 mM glucose; Gibco 11 965) or no glucose (Gibco 11 966) media, or a mixture of the two to yield intermediate glucose concentrations, with 1% BSA for 24 h. For hypoxic cultures, media were first sparged with 94% N2;5% CO2;1% O2 gas, and plated cells were incubated in the same atmosphere in a sealed container in the incubator.

2.7. Polymerase chain reaction

RNA from MEC1 cells was isolated using RNA Mini Kit (Ambion). Equal amounts of RNA were used to synthesize cDNA using M-MLV reverse transcriptase (Invitrogen). The following primer sequences were used to amplify Igf2 cDNA fragments: 5′-GGCCTTCGCCTTGTGCTGCATC-3′ and 5′-GGATCCACGATCAGGGGACGATGACG-3′. 18S classic II (Ambion) and primers for beta-actin were used for the internal control. Q-PCR was performed in iQ SYBR Green Supermix (Bio-Rad) using LightCycler480 (Roche). The primers used for Igf2 Q-PCR were as follows: 5′-CCCTCAGCAAGTGCCTAAAG-3′ and 5′-TTAGGGTGCCTCGAGATGTT-3′. Quantification was normalized to Gapdh expression.

2.8. X-gal staining

Embryos or placentas were fixed with 2% PFA, cryopreserved, and embedded in OCT. Ten micrometre cryosections were stained with X-gal at 37°C overnight. For whole-mount staining, the samples were stained immediately after fixation.

2.9. Placenta IF

E13.5 placentas were fixed with 4% paraformaldehyde in PBS at 4°C overnight. After cryoprotection in 10% and 30% sucrose, the tissue was embedded and frozen in OCT. Eight micrometre sections were used for IF. Slides were briefly fixed with 4% PFA, washed with PBS, and permeabilized in 0.1% Triton X-100 at room temperature for 30 min. After blocking with 1× casein solution (Vector Labs SP-5020) at room temperature for 1 h, sections were incubated in rabbit anti-GFP (Invitrogen A6455; 1:500) and rat anti-mouse CD31 (BD Pharmingen 5 50 274; 1:100) in PBS containing 1% BSA and 10% donkey serum at 4°C overnight. Secondary antibodies donkey anti-rabbit (Invitrogen Alexa Fluor 488) and donkey anti-rat (Invitrogen Alexa Fluor 594) were used to detect GFP and CD31 with 45 min incubation at room temperature. Slides were mounted with mounting media containing DAPI.

2.10. Glucose measurement

Embryos were dissected in PBS on ice, then quickly frozen in liquid nitrogen, and kept at −80°C until use. Embryos of the appropriate genotypes were freeze-thawed three times to promote lysis, and then homogenized in 450 µl of glucose assay buffer (BioVision). Eight microlitres of lysate per sample was used for glucose measurement, using a glucose colorimetric assay kit (BioVision K606) and following the manufacturer's recommended protocol. A standard curve was made using a reference solution provided in the kit. Total embryo protein was determined with the Bio-Rad reagent. Calculation of whole embryo glucose concentrations used the glucose/embryo measurements and embryo volumes of 25 μl at E10.5, 110 μl at E12.5, and 265 μl at E14.5 as measured directly and consistent with previously reported measurements.26,27 These total embryo volumes do not take into account excluded volume, so free glucose concentrations are likely to be somewhat higher than calculated.

3. Results

3.1. Epicardial IGF2 controls ventricular wall expansion

We previously reported that global deficiency of Igf2 resulted in diminished ventricular cardiomyocyte proliferation and a poorly developed ventricular wall.8 Igf2 is expressed in epicardium and endocardium (Supplementary material online, Figure S1), may be expressed at a very low level in the myocardium,28 and is expressed in a number of other embryonic and extraembryonic tissues outside of the heart.29 To genetically define the tissue source of IGF2 in heart development, we combined a conditional Igf2 allele with a number of Cre drivers. Nkx2.5Cre drives highly efficient recombination in all three primary mesodermal cell types of the heart (epicardium, myocardium, and endocardium), but it is active in only a very small number of non-cardiac tissues.16 We replicated the global Igf2 null heart phenotype in Nkx2.5Cre/Igf2 conditional mutants (Figure 1), implying that Igf2 function in heart development is intrinsic to the heart. The same phenotype was observed using Tbx18Cre, which is active in the epicardium,18 but not when using Tie2Cre, which is active in endocardium and endothelium,21 nor with Myh6Cre, which is active only in the myocardium20 (Figure 1). Our results and measurements confirm the epicardium as the primary source of IGF2 that supports ventricular wall formation. If the endocardium or myocardium have any additional contributions to this process, these must be very small incremental roles.

Figure 1.

Figure 1

Open in a new tab

Epicardial IGF2 promotes ventricular wall morphogenesis. Shown are H&E stained sections of control and conditional Igf2 mutant E14.5 hearts, with the boxed regions shown at higher magnification. Scale bars = 100 μm. Brackets and cz, compact zone. Panels below show quantification of right and left ventricular wall thickness in E14.5 Igf2 conditional mutant embryos (mean ± SEM). *P < 0.0002; ns, not significant (P > 0.05) using an unpaired two-tailed t-test. The numbers of conditional mutant embryos examined were as follows: Nkx2.5Cre/Igf2 (10 from 5 litters); Tbx18Cre/Igf2 (8 from 4 litters); Tie2Cre/Igf2 (11 from 3 litters); Myh6Cre/Igf2 (6 from 2 litters).

The coronary vasculature initially forms in the subepicardial space,30 and it is theoretically possible that epicardial IGF2 might support heart development through promotion of coronary vasculogenesis. However, whole-mount PECAM1 staining at E14.5 demonstrated a normal distribution and degree of maturation of coronary vasculature in Nkx2.5Cre/Igf2 and Tbx18Cre/Igf2 mutants (Supplementary material online, Figure S2). That these vessels are functional is implied by the viability of IGF2- and IGF receptor-deficient mutant embryos in the late gestation period.8,31

3.2. Extracardiac RA signalling regulates epicardial Igf2 expression

Embryos compromised in IGF or RA signalling have a similar ventricular wall phenotype, suggesting a possible relation between these pathways. Epicardial Igf2 expression is substantially reduced in Rxra global mutant embryos (Supplementary material online, Figure S1A) and in embryos globally lacking expression of the RA synthetic enzyme RALDH2;15 the elimination of this domain of expression in Rxra and Raldh2 mutant embryos is consistent with epicardial IGF2 serving as a downstream effector of RA signalling in heart development. Igf2 is expressed normally in the endocardium of Rxra−/− mutants (Supplementary material online, Figure S1A), which implies a different mode of regulation of this gene in the endocardium vs. epicardium.

We used several Cre lines combined with a conditional dominant negative RA receptor allele (designated Rara403)22 to address the site(s) of RA signalling relevant to epicardial Igf2 expression and heart development. Mesp1Cre17 is highly active in all mesoderm of the heart and throughout mesoderm in the anterior region of the embryo, although it is inefficient in mesoderm of the liver and of the caudal half of the embryo; Brachyury (T)-Cre19 has mostly a reciprocal pattern of recombination efficiency (including efficient liver mesoderm activity) and is inefficient in the heart (Supplementary material online, Figure S3). We observed normal ventricular chamber morphology in Rara403 embryos combined with Mesp1Cre or with Nkx2.5Cre (Supplementary material online, Figure S4), but all TCre/Rara403 embryos at E14.5 displayed peripheral oedema (a sign of impaired heart function) and an underdeveloped ventricular wall (Figure 2A and B). Furthermore, epicardial but not endocardial Igf2 expression was abolished in TCre/Rara403 embryos at all examined midgestation time points (Figure 2C), although the epicardium was intact (Supplementary material online, Figure S5). The combination of Mesp1Cre with Rara403 results in highly penetrant outflow tract and arch artery phenotypes,32,33 both of which involve RA signalling processes earlier in development than the formation of the epicardium, implying that the absence of a ventricular phenotype cannot be trivially explained on the basis of timing or recombination efficiency. Thus, the sites of RA signalling relevant to ventricular wall formation are external to the heart and to the Mesp1Cre domain, and lie within the TCre domain.

Figure 2.

Figure 2

Open in a new tab

Phenotype following abolition of RA signalling in the TCre domain. (A) Whole mount views of control and TCre/Rara403 littermate E14.5 embryos; note oedema (arrows) in the conditional mutant. (B) Morphology of the heart at E14.5; note hypoplasia of the ventricular wall in the conditional mutant. (C) Igf2 expression by ISH; note absence of epicardial (but persistence of endocardial) Igf2 expression in the conditional mutant at all midgestation time points. (D) Restoration of epicardial Igf2 expression in E12.5 TCre/Rara403 mutants in thorax culture in the presence of 5 mM glucose. n ≥ 3 for both genotypes for histology and n ≥ 3 for all in situ hybridization analyses.

3.3. EPO signalling transiently regulates epicardial Igf2 expression

Epo and EpoR global nulls are compromised in ventricular wall development.14 To test whether this phenotype is related to the epicardium and to epicardial Igf2 expression, we combined a conditional EpoR allele24 with Nkx2.5Cre to eliminate EPOR function in all mesoderm of the heart and with Tbx18Cre to eliminate EPOR function in the epicardium. Neither manipulation had any impact on the presence or morphology of the epicardium, but in both cases, epicardial Igf2 expression at E10.5–11.5 was reduced to near undetectable levels (Figure 3; Supplementary material online, Figure S6). Endocardial Igf2 expression was not affected in either case. Subject to the assumption that EPOR function in the epicardium mediates EPO signals, these observations directly demonstrate the involvement of EPO signalling in promoting epicardial Igf2 expression at these time points. Epo is not expressed in the heart,14 and a prominent site of expression during the E10.0–11.5 period is the liver.11 In principle, EPO protein from the liver (or elsewhere in the abdominal cavity) could come directly into contact with the epicardium by transport through large openings in the incompletely formed diaphragm (the pericardioperitoneal canals) that are present through E11.5 and which close around E12.534 (see also Discussion).

Figure 3.

Figure 3

Open in a new tab

EPO signalling transiently regulates epicardial Igf2 expression. Igf2 expression was visualized by ISH. Note absence of epicardial expression in conditional EpoR mutants at E10.5 and E11.5, and restoration at E12.5 and E14.5. Endocardial expression was unchanged in the mutant. Scale bars = 50 μm. n ≥ 3 for all time points.

Although epicardial Igf2 expression was compromised in Nkx2.5Cre/EpoR and Tbx18Cre/EpoR mutants at E10.5–11.5, we noted restoration of expression in both mutant backgrounds at E12.5 and continuing through E14.5 (Figure 3; Supplementary material online, Figure S6). Consistent with this observation, both Nkx2.5Cre/EpoR and Tbx18Cre/EpoR mutants had normal heart morphology by E14.5 (Supplementary material online, Figure S7). Thus, while epicardial Igf2 expression from E10 to 14 is required for normal ventricular morphogenesis, EPO signalling only transiently controls epicardial Igf2 expression and is not required for the ultimate formation of the ventricular wall. The developmental time during which EPO regulates epicardial Igf2 expression (from the time of epicardium formation at E10 up to E11.5) is also the time during which Epo is expressed at high levels in the fetal liver in response to RA,11 and when the pericardioperitoneal canals are open.

3.4. A transition to placental control of epicardial Igf2 expression and heart morphogenesis

The delayed onset but ultimately normal level of epicardial Igf2 expression in Nkx2.5Cre/EpoR and Tbx18Cre/EpoR mutants implies that a different mode of regulation of epicardial Igf2 expression becomes established starting between E11.5 and 12.5. Importantly, though, this new mode of regulation must still be dependent on RA signalling, because epicardial Igf2 expression through E14.5 and ventricular morphology at E14.5 are both compromised in Rxra10,15 (see also Supplementary material online, Figure S1A) and TCre/Rara403 (Figure 2) mutants. The transition in control of epicardial Igf2 expression at E11.5–12.5 is coincident with the onset of placental function. Placental organization, specifically of the labyrinthine layer that is the major site of exchange between maternal and fetal blood, was previously noted to be compromised in Rxra−/− mutants.35 We observed a substantially diminished labyrinthine layer in TCre/Rara403 mutants (Figure 4A; Supplementary material online, Figure S8), similar to that previously reported for Rxra global mutants. TCre is active extensively in endothelial cells of the labyrinthine layer of the placenta, and in no other cell type present in appreciable quantity (Figure 4B), implying that the endothelial cell lineage is a likely target of RA signalling in the placenta.

Figure 4.

Figure 4

Open in a new tab

Placental phenotype in TCre/Rara403 mutants. (A) H&E sections of placentas reveal a smaller labyrinthine (L) layer in the TCre/Rara403 mutant. Scale bars = 500 μm. Higher magnification views of these sections are in Supplementary material online, Figure S8. n ≥ 3 for both genotypes at all time points. (B) Endothelial recombination in E13.5 placentas from embryos bearing paternally inherited TCre and the R26-YFP conditional reporter. Most cells are co-labelled with Pecam1; white arrowheads point to rare Pecam1+, GFP− cells. Recombination occurred also in rare non-endothelial placental cells (blue arrowheads), the identity of which is unknown. S, spongiotrophoblast layer. Scale bars = 500 μm (upper) and 100 μm (lower). n = 2. (C) Measurement of midgestation whole embryo glucose levels, on a per embryo basis or normalized to recovered protein. *P = 0.024; **P = 0.008; ns, not significant, using an unpaired two-tailed t-test. The number of embryos used for each determination was, for control embryos, 11 at E10.5, 13 at E12.5, and 10 at E14.5; for TCre/Rara403 embryos, 5 at E10.5, 9 at E12.5, and 6 at E14.5.

Because glucose transport is one of the primary functions of the placenta,36 we measured embryonic glucose levels to confirm the impairment of placental function in TCre/Rara403 mutants (Figure 4C). Approximate whole embryo glucose concentrations calculated from these numbers (see Methods) for wild-type embryos at E10.5, E12.5, and E14.5 are <50 µM, 0.5 mM, and 1 mM, respectively. In TCre/Rara403 mutants, we observed a 40% decrease in whole embryo glucose at E12.5, and more than two-fold decrease at E14.5, confirming deficient placental function.

We reasoned that glucose or other materials transported from the placenta to the midgestation embryo might regulate epicardial Igf2 expression and heart development after E11.5, and their deficiency might explain the heart phenotypes of TCre/Rara403 mutants. We used in vitro embryonic thorax cultures to model this response (Figure 5A). Epicardial Igf2 expression in cultured wild-type thorax segments was undetectable in no glucose conditions, induced in the presence of glucose in the physiological range of the midgestation embryo (see above), and further induced in high glucose conditions. In all cases, endocardial Igf2 expression was not impacted. Glucose restored epicardial Igf2 expression in TCre/Rara403 thorax cultures (Figure 2D), supporting the conclusion that abolished epicardial Igf2 expression and heart defects in this mutant in vivo (Figure 2B and C) are the indirect result of impaired placental function. In MEC1 mouse embryonic ventricular epicardial cells,8 Igf2 expression was minimal in the absence of glucose and increased with increasing glucose (Figure 5B and C), demonstrating that glucose directly induces epicardial Igf2 expression. Glucose did not induce Igf2 expression under hypoxic conditions (Figure 5D), implying that both placental substrates support epicardial Igf2 expression. Finally, in ventricular tissue isolated from thorax cultures, the proliferation intermediate phosphoErk was activated by increasing glucose levels (Figure 5E), consistent with placental glucose-induced activation of a proliferative response in the myocardium.

Figure 5.

Figure 5

Open in a new tab

Placental substrates control epicardial Igf2 expression. (A) Epicardial Igf2 expression in normoxic thorax culture is glucose responsive; detected by ISH. Scale bars = 50 μm. n ≥ 3 for all time points and conditions. (B and C) Igf2 expression in MEC1 epicardial cells is glucose responsive. Quantitative PCR analyses of glucose response of Igf2 expression in MEC1 cells are shown in B′ and C′; n = 3 for all samples. Asterisk in B′: P < 0.04 for all glucose-treated sample compared with no glucose; asterisk in C′: P < 0.02 for 5 mM glucose compared with no glucose and for 25 mM glucose compared with 5 mM glucose (unpaired two-tailed t-test). (D) Glucose-induced Igf2 expression in MEC1 epicardial cells requires normoxia. (E) Glucose induces pErk activation in ventricular tissue in thorax culture; detected by Western blotting. n = 2 for each condition.

Our studies support an overall model (Figure 6) in which epicardial Igf2 expression is controlled independently in two distinct phases, first by EPO from E10 to E11.5 and then by glucose and oxygen (and perhaps also additional materials) from the placenta. Both processes involve signals from a distant tissue acting on the epicardium, which then produces IGF2 that acts locally to induce cardiomyocyte proliferation and ventricular wall growth.

Figure 6.

Figure 6

Open in a new tab

Model of regulation of epicardial Igf2 expression and ventricular wall morphogenesis. Boxed regions magnifying the epicardium, subepicardial space, and myocardium are shown below. From E10.0 to 11.5, after formation of the epicardium and prior to establishment of placental circulation, embryonic tissue is hypoxic, RA in the liver is high, and the liver is a prominent source of EPO (green diamonds) that diffuses via coelomic fluid from the peritoneal cavity into the pericardial cavity through the pericardioperitoneal canals (open gaps in the diaphragm). EPO activates EPOR on the surface of the epicardium to induce expression of IGF2 (blue circles), which diffuses across the subepicardial space to activate myocardial IGF receptors and induce proliferation. After E11.5 and with the onset of placental function, embryonic tissue becomes normoxic and Epo expression in the liver declines, and at the same time the diaphragm closes. Placental glucose and oxygen from the heart lumen then maintain Igf2 expression in the epicardium.

4. Discussion

Our prior results indicated the importance of IGF2 signalling in ventricular wall morphogenesis.8 Through tissue-specific Igf2 gene disruption, we confirm that the epicardium is the source of this mitogen.

The unexpected realization, made possible because of our genetic approach, is that ventricular growth during midgestation mouse development occurs in two independent phases (Figure 6). The E10–11.5 preplacental phase occurs in the context of hypoxia and very low glucose, and employs EPO signalling via EPOR to regulate epicardial Igf2 expression. Epo is not expressed in the heart,14 and a prominent site of expression in the E10–11.5 period is the fetal liver, where it is expressed at high level under the transcriptional control of RA and hypoxia.11 EPO can reach the epicardium via coelomic fluid that moves easily through the pericardioperitoneal canals. With the onset of placental function around E11.5, Epo expression in the liver diminishes greatly,11,13 and glucose and oxygen (and perhaps additional factors) reach the heart through circulation and maintain epicardial Igf2 expression, presumably by diffusion from the heart lumen. Unlike large proteins that do not diffuse deeply into tissues and must be presented directly to responding cells, small molecules like glucose and oxygen diffuse over tens of cell diameters.37 The specific intracellular mechanisms by which glucose and oxygen support Igf2 expression in the epicardium are unknown but are under ongoing investigation. According to this understanding, the severe heart phenotype seen in Epo and EpoR mutants14 is explained by the initial loss of liver-epicardium EPO signalling followed by persistent hypoxia associated with the absence of definitive erythropoiesis. The similar heart phenotype in Rxra−/− and TCre/Rara403 mutants is explained by the initial loss of liver-epicardium EPO signalling followed by persistent hypoxia and reduced glucose associated with impaired placental function.

A large number of gene mutations in mice result in compromised placental development.38 In severe cases, these cause early or midgestation embryo lethality, whereas less severe placental impairment is associated with general embryo growth deficiency in late gestation. The placental phenotype of TCre/Rara403 mutants seems to fall somewhere in the middle of this spectrum, and as we show in this study, this is associated with impaired ventricular wall formation. A partially underdeveloped ventricular wall may have been present in some other mutant lines with intermediate placental severity and simply not recognized. Alternatively, epicardial Igf2 expression and heart morphogenesis might require the combined activity of glucose and oxygen, plus an additional placental product that is specifically absent in TCre/Rara403 mutants, such that placental defects in other mutants might not be associated with the same impairment of heart development.

While epicardial IGF2 signalling to the myocardium occurs at close proximity across the subepicardial space, the factors that control epicardial Igf2 expression reach the heart from a distance and thus qualify as endocrine signals. Endocrine interactions between organs at a distance are the norm in adult physiology but have not previously been recognized in heart development and are rarely employed in the embryo; for this reason, almost all developmental studies focus on local signalling events between adjacent and nearby tissues. Even placental function is generally thought to involve the passive provision of growth substrates rather than regulating a morphogenic programme of gene expression, such as occurs in the epicardium to influence heart growth. Our observations demonstrate that the embryo does indeed employ endocrine signalling mechanisms to facilitate development and specifically in heart morphogenesis.

The onset of placental function around E11.5 explains the transition in control of epicardial Igf2 expression and heart growth. Several other developmental transitions occur at the same time, including the dramatic reduction in liver Epo expression, maturation of hepatoblasts to hepatocytes, and closure of the pericardioperitoneal canals (completion of the diaphragm). Liver Epo expression is clearly also regulated by the onset of placental function,11,13 and we speculate that other developmental events at this same time may also be responsive to placental activity, perhaps also via glucose and oxygen acting as developmental endocrine signals. This might serve to co-ordinate the timing of distinct processes in multiple organs, including the heart, during midgestation mouse development.

Supplementary material

Supplementary material is available at Cardiovascular Research online.

Conflict of interest: none declared.

Funding

This work was supported by the National Institutes of Health grant HL070123 to H.M.S. H.S. was supported by a postdoctoral fellowship from the California Institute for Regenerative Medicine.

References

  • 1.Rossant J. Mouse mutants and cardiac development: new molecular insights into cardiogenesis. Circ Res. 1996;78:349–353. doi: 10.1161/01.res.78.3.349. [DOI] [PubMed] [Google Scholar]
  • 2.Sengbusch JK, He W, Pinco KA, Yang JT. Dual functions of [alpha]4[beta]1 integrin in epicardial development: initial migration and long-term attachment. J Cell Biol. 2002;157:873–882. doi: 10.1083/jcb.200203075. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Viragh S, Challice CE. The origin of the epicardium and the embryonic myocardial circulation in the mouse. Anat Rec. 1981;201:157–168. doi: 10.1002/ar.1092010117. [DOI] [PubMed] [Google Scholar]
  • 4.Moore AW, McInnes L, Kreidberg J, Hastie ND, Schedl A. YAC complementation shows a requirement for Wt1 in the development of epicardium, adrenal gland and throughout nephrogenesis. Development. 1999;126:1845–1857. doi: 10.1242/dev.126.9.1845. [DOI] [PubMed] [Google Scholar]
  • 5.Gittenberger-de Groot AC, Vrancken Peeters MP, Bergwerff M, Mentink MM, Poelmann RE. Epicardial outgrowth inhibition leads to compensatory mesothelial outflow tract collar and abnormal cardiac septation and coronary formation. Circ Res. 2000;87:969–971. doi: 10.1161/01.res.87.11.969. [DOI] [PubMed] [Google Scholar]
  • 6.Manner J, Perez-Pomares JM, Macias D, Munoz-Chapuli R. The origin, formation and developmental significance of the epicardium: a review. Cells Tissues Organs. 2001;169:89–103. doi: 10.1159/000047867. [DOI] [PubMed] [Google Scholar]
  • 7.Chen TH, Chang TC, Kang JO, Choudhary B, Makita T, Tran CM, Burch JB, Eid H, Sucov HM. Epicardial induction of fetal cardiomyocyte proliferation via a retinoic acid-inducible trophic factor. Dev Biol. 2002;250:198–207. doi: 10.1006/dbio.2002.0796. [DOI] [PubMed] [Google Scholar]
  • 8.Li P, Cavallero S, Gu Y, Chen THP, Hughes J, Hassan AB, Bruning J, Pashmforoush M, Sucov HM. IGF signaling directs ventricular cardiomyocyte proliferation during embryonic heart development. Development. 2011;138:1795–1805. doi: 10.1242/dev.054338. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Mangelsdorf DJ, Evans RM. The RXR heterodimers and orphan receptors. Cell. 1995;83:841–850. doi: 10.1016/0092-8674(95)90200-7. [DOI] [PubMed] [Google Scholar]
  • 10.Sucov HM, Dyson E, Gumeringer CL, Price J, Chien KR, Evans RM. RXR alpha mutant mice establish a genetic basis for vitamin A signaling in heart morphogenesis. Genes Dev. 1994;8:1007–1018. doi: 10.1101/gad.8.9.1007. [DOI] [PubMed] [Google Scholar]
  • 11.Makita T, Hernandez-Hoyos G, Chen TH, Wu H, Rothenberg EV, Sucov HM. A developmental transition in definitive erythropoiesis: erythropoietin expression is sequentially regulated by retinoic acid receptors and HNF4. Genes Dev. 2001;15:889–901. doi: 10.1101/gad.871601. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Wu H, Liu X, Jaenisch R, Lodish HF. Generation of committed erythroid BFU-E and CFU-E progenitors does not require erythropoietin or the erythropoietin receptor. Cell. 1995;83:59–67. doi: 10.1016/0092-8674(95)90234-1. [DOI] [PubMed] [Google Scholar]
  • 13.Makita T, Duncan SA, Sucov HM. Retinoic acid, hypoxia, and GATA factors cooperatively control the onset of fetal liver erythropoietin expression and erythropoietic differentiation. Dev Biol. 2005;280:59–72. doi: 10.1016/j.ydbio.2005.01.001. [DOI] [PubMed] [Google Scholar]
  • 14.Wu H, Lee SH, Gao J, Liu X, Iruela-Arispe ML. Inactivation of erythropoietin leads to defects in cardiac morphogenesis. Development. 1999;126:3597–3605. doi: 10.1242/dev.126.16.3597. [DOI] [PubMed] [Google Scholar]
  • 15.Brade T, Kumar S, Cunningham TJ, Chatzi C, Zhao X, Cavallero S, Li P, Sucov HM, Ruiz-Lozano P, Duester G. Retinoic acid stimulates myocardial expansion by induction of hepatic erythropoietin which activates epicardal Igf2. Development. 2011;138:139–148. doi: 10.1242/dev.054239. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Moses KA, DeMayo F, Braun RM, Reecy JL, Schwartz RJ. Embryonic expression of an Nkx2-5/Cre gene using ROSA26 reporter mice. Genesis. 2001;31:176–180. doi: 10.1002/gene.10022. [DOI] [PubMed] [Google Scholar]
  • 17.Saga Y, Miyagawa-Tomita S, Takagi A, Kitajima S, Miyazaki J, Inoue T. MesP1 is expressed in the heart precursor cells and required for the formation of a single heart tube. Development. 1999;126:3437–3447. doi: 10.1242/dev.126.15.3437. [DOI] [PubMed] [Google Scholar]
  • 18.Cai CL, Martin JC, Sun Y, Cui L, Wang L, Ouyang K, Yang L, Bu L, Liang X, Zhang X, Stallcup WB, Denton CP, McCulloch A, Chen J, Evans SM. A myocardial lineage derives from Tbx18 epicardial cells. Nature. 2008;454:104–108. doi: 10.1038/nature06969. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Perantoni AO, Timofeeva O, Naillat F, Richman C, Pajni-Underwood S, Wilson C, Vainio S, Dove LF, Lewandoski M. Inactivation of FGF8 in early mesoderm reveals an essential role in kidney development. Development. 2005;132:3859–3871. doi: 10.1242/dev.01945. [DOI] [PubMed] [Google Scholar]
  • 20.Agah R, Frenkel PA, French BA, Michael LH, Overbeek PA, Schneider MD. Gene recombination in postmitotic cells. Targeted expression of Cre recombinase provokes cardiac-restricted, site-specific rearrangement in adult ventricular muscle in vivo. J Clin Invest. 1997;100:169–179. doi: 10.1172/JCI119509. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Kisanuki YY, Hammer RE, Miyazaki J, Williams SC, Richardson JA, Yanagisawa M. Tie2-Cre transgenic mice: a new model for endothelial cell-lineage analysis in vivo. Dev Biol. 2001;230:230–242. doi: 10.1006/dbio.2000.0106. [DOI] [PubMed] [Google Scholar]
  • 22.Rajaii F, Bitzer ZT, Xu Q, Sockanathan S. Expression of the dominant negative retinoid receptor, RAR403, alters telencephalic progenitor proliferation, survival, and cell fate specification. Dev Biol. 2008;316:371–382. doi: 10.1016/j.ydbio.2008.01.041. [DOI] [PubMed] [Google Scholar]
  • 23.Haley VL, Barnes DJ, Sandovici I, Constancia M, Graham CF, Pezzella F, Buhnemann C, Carter EJ, Hassan AB. Igf2 pathway dependency of the Trp53 developmental and tumour phenotypes. EMBO Mol Med. 2012;4:705–718. doi: 10.1002/emmm.201101105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Tsai PT, Ohab JJ, Kertesz N, Groszer M, Matter C, Gao J, Liu X, Wu H, Carmichael ST. A critical role of erythropoietin receptor in neurogenesis and post-stroke recovery. J Neurosci. 2006;26:1269–1274. doi: 10.1523/JNEUROSCI.4480-05.2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Soriano P. Generalized lacZ expression with the ROSA26 Cre reporter strain. Nat Genet. 1999;21:70–71. doi: 10.1038/5007. [DOI] [PubMed] [Google Scholar]
  • 26.Kulandavelu S, Qu D, Sunn N, Mu J, Rennie MY, Whiteley KJ, Walls JR, Bock NA, Sun JC, Covelli A, Sled JG, Adamson SL. Embryonic and neonatal phenotyping of genetically engineered mice. ILAR J. 2006;47:103–117. doi: 10.1093/ilar.47.2.103. [DOI] [PubMed] [Google Scholar]
  • 27.Wong MD, Dorr AE, Walls JR, Lerch JP, Henkelman RM. A novel 3D mouse embryo atlas based on micro-CT. Development. 2012;139:3248–3256. doi: 10.1242/dev.082016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Murrell A, Heeson S, Bowden L, Constancia M, Dean W, Kelsey G, Reik W. An intragenic methylated region in the imprinted Igf2 gene augments transcription. EMBO Rep. 2001;2:1101–1106. doi: 10.1093/embo-reports/kve248. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Bondy CA, Werner H, Roberts CT, LeRoith D. Cellular pattern of insulin-like growth factor-I (IGF-I) and type I IGF receptor gene expression in early organogenesis: comparison with IGF-II gene expression. Mol Endocrinol. 1990;4:1386–1398. doi: 10.1210/mend-4-9-1386. [DOI] [PubMed] [Google Scholar]
  • 30.Red-Horse K, Ueno H, Weissman IL, Krasnow MA. Coronary arteries form by developmental reprogramming of venous cells. Nature. 2010;464:549–553. doi: 10.1038/nature08873. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Louvi A, Accili D, Efstratiadis A. Growth-promoting interaction of IGF-II with the insulin receptor during mouse embryonic development. Dev Biol. 1997;189:33–48. doi: 10.1006/dbio.1997.8666. [DOI] [PubMed] [Google Scholar]
  • 32.Li P, Pashmforoush M, Sucov HM. Retinoic acid regulates differentiation of the secondary heart field and TGFbeta-mediated outflow tract septation. Dev Cell. 2010;18:480–485. doi: 10.1016/j.devcel.2009.12.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Li P, Pashmforoush M, Sucov HM. Mesodermal retinoic acid signaling regulates endothelial cell coalescence in caudal pharyngeal arch artery vasculogenesis. Dev Biol. 2012;361:116–124. doi: 10.1016/j.ydbio.2011.10.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Norden J, Grieskamp T, Christoffels VM, Moorman AF, Kispert A. Partial absence of pleuropericardial membranes in Tbx18- and Wt1-deficient mice. PLoS ONE. 2012;7:e45100. doi: 10.1371/journal.pone.0045100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Sapin V, Dolle P, Hindelang C, Kastner P, Chambon P. Defects of the chorioallantoic placenta in mouse RXRalpha null fetuses. Dev Biol. 1997;191:29–41. doi: 10.1006/dbio.1997.8687. [DOI] [PubMed] [Google Scholar]
  • 36.Lager S, Powell TL. Regulation of nutrient transport across the placenta. J Pregnancy. 2012;2012:179827. doi: 10.1155/2012/179827. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Franko AJ, Sutherland RM. Oxygen diffusion distance and development of necrosis in multicell spheroids. Radiat Res. 1979;79:439–453. [PubMed] [Google Scholar]
  • 38.Watson ED, Cross JC. Development of structures and transport functions in the mouse placenta. Physiology (Bethesda) 2005;20:180–193. doi: 10.1152/physiol.00001.2005. [DOI] [PubMed] [Google Scholar]

Articles from Cardiovascular Research are provided here courtesy of Oxford University Press

RESOURCES