Cells in the SHF are highly proliferative so that the pool of cardiac progenitor cells is maintained as the heart grows. FGF, canonical WNT and Hedgehog signalling promote proliferation [18]. Differentiation of SHF cells is triggered as they enter the poles of the heart. BMP signalling plays an important role in this process, antagonising FGF signalling [20, 21] and promoting the deployment and differentiation of cells that will contribute to the arterial pole. Notch and non-canonical WNT signalling are implicated also in cardiac differentiation [18]. At the venous pole of the heart, canonical WNT, Hedgehog and BMP signalling orchestrate the critical balance between proliferation and differentiation, where Hedgehog signalling promotes the migration of cells to the heart tube as well as their proliferation [22, 23]. These pathways are activated by signals from surrounding tissues as well as by cell autonomous signalling within the SHF [24]. Thus, the midline structures (neural tube and notochord) are sources of canonical WNT and Hedgehog signalling, promoting SHF proliferation. Hedgehog signalling also comes from the endoderm, while FGF signals come from surrounding endoderm and ectoderm as well as from the SHF itself. Neural crest cells, which migrate through the SHF into the arterial pole, mediate BMP inhibition of FGF signalling, such that in the absence of neural crest, excessive SHF cell proliferation results in reduced cardiomyocyte differentiation and defective growth of the heart. BMP signals are also produced by cells in the SHF, promoting neural crest survival, and by cardiomyocytes, promoting recruitment to the poles of the heart.

An extensive gene regulatory network operates in the SHF to control cardiac progenitor cell behaviour [18]. Key transcription factors modulate signalling pathways, illustrated by Tbx1 interference with the BMP signalling cascade to delay differentiation [24, 25] or Tbx1 activation of enhancers that control Fgf8 and Fgf10 expression in the SHF [24–26]. Islet1 also activates the Fgf10 enhancer [26] and has a SHF enhancer which is itself controlled by forkhead box (Fox) c2 [18].

Some genes expressed in the SHF are specific to the progenitor cell population in this cardiac context. Foxc2, for example, marks SHF progenitors, and Islet1 has long been considered to be the key marker of this field. However, more sensitive reporters now indicate that Islet1 is also expressed at a low level in FHF progenitors, although the Islet1 mutant phenotype suggests that it is mainly functional in the SHF [18].

Genes that are involved in myocardial differentiation also are expressed in the SHF. These include genes coding for the transcription factors, NK2 homeobox 5 (Nkx2-5), Gata4, Mef2c, serum response factor (SRF), Tbx5 and the chromatin remodelling protein SWI-/SNF-related, matrix-associated, actin-dependent regulator of chromatin, subfamily D, member 3 (Smarcd3; also called 60 kDa BRG-1/Brm associated factor subunit C1; Baf60c) which marks a subset of early Mesp1-positive cardiac progenitors [18]. These genes are expressed in differentiating cells of the cardiac crescent and in regions of the heart, where combinations of them activate sarcomeric genes. Ectopic expression of Gata4, Tbx5 and Smarcd3 in non-cardiogenic regions of the mouse embryo has been shown to activate cardiac muscle formation [27], indicating the role of these factors in cardiac progenitor cell determination as well as in cardiac differentiation. Their targets in the SHF are not well understood, although transcriptome analysis of Nkx2–5-expressing cells gives some insights [28]. Notably, Nkx2–5 functions in a negative feedback loop with Bmp2/Smad1 signalling that controls cardiac progenitor cell behaviour, which is probably of importance in the aetiology of Nkx2-5-related human congenital heart defects. Different combinations of factors are one reason why targets may differ between cardiac progenitors and myocardium. Another consideration is the level of expression which may affect function. The potential importance of modulating the level of these factors is indicated by Tbx1 downregulation of SRF and Tbx5 in the SHF [16, 25]. In the Tbx1 mutant, abnormal cardiac differentiation of SHF progenitors in the pericardial wall probably reflects this effect as well as deregulation of the balance between BMP and FGF signalling [24]. The SHF enhancer of the Fgf10 gene provides an example of how levels and combinations of transcription factors can be functionally important [26]. In this enhancer both Islet1 and Nkx2-5 bind to the same sites. In the SHF, Islet1 levels are high, and it competes successfully for binding, leading to activation of Fgf10, which also depends on Tbx1. In the heart, on the other hand, Nkx2-5 levels are very high and it displaces Islet1. This leads to downregulation of Fgf10 in keeping with the repressive function of Nkx2-5 observed for SHF genes, including Islet1, in the heart [28]. Since Nkx2-5 also functions as a transcriptional activator in the myocardium this dual role again depends on combinatorial effects of other factors.

Most transcription factors and components of signalling pathways that play a role in cardiogenesis are not cardiac specific. This is in contrast to skeletal muscle where the myogenic regulatory factors of the myogenic differentiation (MyoD) family are only expressed during myogenesis [29]. Analysis of mutant phenotypes is informative for the heart only when the gene does not play a vital role during early embryogenesis. Islet1 or Mesp1, for example, is mainly important for formation of the heart at early stages of development, whereas Fgf8 is required for gastrulation and conditional mutation of Fgf8 directed to cardiac progenitors is essential in order to understand its role in the SHF. Mesp1-Cre lines [13] have proved to be very valuable in targeting all cardiac progenitors, whereas an enhancer of the Mef2c gene targets cells in the SHF [30]. It follows that in the human population, many mutations in cardiac regulatory genes may be early embryonic lethal or result in multiple effects as seen in a number of syndromes where the heart is affected. This is the case, for example, of mutations in TBX1 which give rise to DiGeorge syndrome [31]. This includes major defects at the arterial pole of the heart, such as common arterial trunk or persistent truncus arteriosus as well as craniofacial abnormalities, due to the function of Tbx1 in the SHF as well as in endoderm and ectoderm in the pharyngeal region which is important for cardiac and cranial neural crest. Unless there is some cardiac-specific aspect of function, mutations in enhancers that direct gene transcription to the cardiac progenitors of the first or second heart fields are the best candidates for defects that are restricted to the formation of the heart. Many such enhancers, that specifically direct transcription in the SHF, have now been identified for genes such as Fgf8, Fgf10, Islet1, Nkx2–5 or Mef2c [18]. They are regulated by combinations of SHF transcription factors, as for Fgf10 [26].

The SHF is patterned on the anterior/posterior axis. Although gene expression has not yet been compared at a single-cell level, regional heterogeneity is apparent. The anterior part of the SHF is marked by genes that are not expressed throughout the field [18]. This is the case for Fgf8 or Fgf10 and also for the SHF Mef2c enhancer which is not active in the posterior SHF. This anterior/posterior distinction correlates with the contribution of cells expressing these genes to the arterial or venous pole, respectively. This is illustrated in the anterior SHF also, where Tbx1 is mainly expressed in cells that will contribute to myocardium at the base of the pulmonary trunk [24]; in contrast to Fgf8/10 or the Mef2c enhancer which mark progenitors of the right ventricle as well as outflow tract myocardium, both of the pulmonary trunk and the aorta. Differences in the origin of myocardium at the base of these arteries are reflected by differences in gene expression [16]. Pulmonary trunk myocardium, for example, is marked by semaphorin 3c (Sema3c) gene expression. In Tbx1mutants pulmonary trunk myocardium is reduced [24], which is probably the early abnormality that underlies the development of tetralogy of Fallot and other cardiac defects in DiGeorge syndrome patients.

The posterior SHF is marked also by distinct gene expression profiles. Homeobox (Hox) genes, differentially expressed along the axis of the embryo, are classically involved in anterior/ posterior patterning. Anterior members of the Hox gene clusters such as Hoxb1 are expressed in the posterior SHF, under the control of retinoic acid signalling. When this signalling pathway is perturbed, Hox gene expression is affected, and the posterior boundary of the SHF is modified [35]. Tbx5 is expressed in cardiac progenitor cells in the posterior SHF. In addition to atrial myocardium, the dorsal mesenchymal protrusion [33] also derives from the posterior SHF, from cells expressing Tbx5 and Islet1, and its development depends on Hedgehog and BMP signalling [22]. This structure is important in the context of congenital heart malformations since it provides the muscular base of the atrial septum, and anomalies in its formation from the posterior SHF lead to atrial and atrioventricular septal defects. Tbx18 is expressed in a particular lateral domain of the posterior SHF which is distinguished also by lack of Nkx2–5 and Islet1 expression [34]. This domain, which has been proposed to constitute a third heart field, contributes caval vein myocardium, which is marked by continuing Tbx18 expression. Retrospective clonal analysis distinguishes sub-lineages of the second myocardial cell lineage (equated with the SHF) which contribute to myocardium at the arterial or venous poles of the heart. However this clonal analysis does not distinguish a particular sub-lineage contributing to caval vein myocardium only, as might be expected if this arises from a third heart field [35].

The distinction between anterior and posterior domains of the SHF is made on the basis of gene expression; however, cells may move between these domains. This was first indicated by genetic tracing of cells that had expressed Hoxb1 in the posterior SHF [32], leading to labelling of myocardial derivatives at the venous pole. Unexpectedly Tbx1-dependent myocardium at the base of the pulmonary trunk was labelled also [16]. In keeping with this result, dye labelling of small groups of cells in the posterior domain of the SHF showed, after embryo culture, that some labelled cells were located in the outflow tract and had therefore moved anteriorly. In so doing these cells activated markers of the anterior SHF, as indicated by Fgf10 reporter gene expression [36]. Most cells do not appear to move extensively within the SHF and clonal analysis [2] suggests coherent, rather than dispersed, cell growth as the heart forms. However lineage studies for the venous pole of the heart showed unexpected clonality between myocardium at the left side of the venous pole (pulmonary vein, left caval vein and left atrium) and pulmonary trunk myocardium at the arterial pole [35]. This is consistent with anterior movement of a subset of progenitor cells located in the left side of the posterior SHF from which left venous pole myocardium derives. Unexpectedly, Cre genetic tracing experiments show that the classic anterior SHF markers, Tbx1 [37] and the Mef2c–AHF enhancer [16], are transiently expressed in progenitor cells giving rise to the atrial septum and dorsal mesenchymal protrusion. Indeed, recent experiments have shown that loss of function of Tbx1 impacts not only the anterior SHF but also the posterior SHF, resulting in venous as well as arterial pole defects [16]. The development of the dorsal mesenchymal protrusion is affected, leading to ostium primum and atrioventricular septal defects. Some anterior SHF cells may move posteriorly to contribute to parts of the venous pole; however, there is no evidence to date for this. Alternatively, diminished proliferation and premature differentiation of cells in the SHF in the absence of Tbx1 may perturb the demarcation between anterior and posterior regions of the SHF. In addition Tbx1 may play a role in the segregation of cells to anterior and posterior domains from a common early progenitor population.