In early murine embryonic development FGF-4 is the first member to be expressed, from the 4 cell stage onto the blastocyst, egg cylinder and primitive streak [11
]. Its deletion causes peri-implantation embryonic lethality (E4-5); early development appears normal up to the blastocyst stage but embryos die within hours after implantation owing to deficient inner cell mass formation and maintenance [12
]. FGF-4 signaling appears to be important as early as the fifth cell division to promote cell proliferation onto the blastocyst stage [13
]. FGF-4 probably signals through FGFR2 as this receptor is the first detected in development, although early expression of FGFR1, 3 and 4 have also been inconsistently reported (probably owing to the few reliable antibodies available)[14
]. Moreover, FGFR2 deletion recapitulates FGF-4 deletion, causing early embryonic lethality (E6-8) due to defects in inner cell mass. FGFR1 deletion is also lethal (E7.5-9.5) and appears to cause defects in gastrulation, mainly by affecting axial patterning and migration/proliferation of cells through the primitive streak, thus inhibiting mesoderm and endoderm specification [16
]. FGFR3 deletion on the other hand is not embryonic lethal but mice display skeletal malformations that may lead to premature death (see section on skeletal/mesenchymal stem cells below), whereas FGFR4 null mice show no obvious phenotype. The functions of this latter receptor in development and postnatal life remain unclear as well as that of FGFR5.
Structure and signaling downstream of FGFRs.
Murine embryonic stem cells (mESCs) have historically been derived from the inner cell mass of the blastocyst [18
] or the epiblast (although these ESCs are considered more primed for gastrulation and germ layer commitment). Since FGF-4−/−
embryos fail to develop because of defects in inner cell mass proliferation and germ layers specification, it has been assumed that FGF signaling in mESCs was required for their differentiation or lineage commitment. Of note, mESCs constitutively express FGF-4 which is thought to act in an autocrine manner. Undifferentiated mESCs were found to express high levels of FGFR1 and 4 which are maintained during differentiation [15
]. They also express FGFR2(IIIb) and FGFR3(IIIc) but upregulate FGFR2(IIIc) and FGFR3(IIIb) upon differentiation. FGF-4−/−
mESCs do not display defects in proliferation in vitro and are capable of multilineage differentiation, however the survival of those differentiated progeny is severely compromised, although the underlying mechanism for this phenotype is still unclear [20
]. Further studies using FGF-4−/−
mESCs or specific inhibitors of FGFR1 and 3 confirmed that inhibition of FGF signaling through these receptors could maintain mESCs in a self-renewing, pluripotent state [21
]. However, these studies also suggest that FGFR1/3 signaling, as well as the presence of specific HSPGs, may act more as a priming or permissive signal for differentiation rather than a differentiation cue itself. Taken together, these studies suggest that FGF signaling in murine embryogenesis and ESCs may have stage specific effects. FGF-4 signaling through FGFR2 stimulates ESCs proliferation from the fifth division to the establishment of the inner cell mass in the blastocyst, whereas signaling through FGFR1/3 in peri-implantation embryos and epiblast ESCs is important for germ layer specification. However, the exact timing of expression of the various receptor isoforms in early lineage specification and their role in self-renewal, priming and differentiation of mESCs remains unclear to date.
The study of molecular events in human post-implantation embryogenesis is complicated by ethical and technical limitations. It is however possible to study pre-implantation embryos from which are derived humans ESCs. It must be noted that hESCs might be more related to epiblast-derived (primed) mESCs in terms of properties and functionality. Nevertheless, hESCs have been found to express several molecular-mass isoforms of FGF-2, 11 and 13 (but not FGF-4) as well as the whole repertoire of FGFRs with the following relative abundance (mRNA levels): FGFR1 > FGFR3 > FGFR4 > FGFR2 [23
]. These levels are modulated during hESCs differentiation, showing an initial decrease followed by upregulation in more advanced differentiation. Early evidence suggested that FGF signaling might be important for proliferation and self-renewal of hESCs in vitro [24
]. These observations were confirmed by many groups and to date FGF-2 is a necessary supplement to hMSCs culture medium, independently of the presence or absence of a feeder layer. The maintenance of pluripotency (self-renewal) by FGF-2 on hESCs may be in part attributed to modulation of Wnt signaling through PI3K [28
], but a recent study again suggests that its effects may be stage specific or dependent on context [29
]. Indeed, FGF-2 may also be important to sustain Nanog expression during BMP-4 induced differentiation of hESCs and promote mesendoderm over trophoectoderm differentiation.
In summary, the roles of FGF signaling in murine and human embryogenesis are diverse and appear stage specific. They are probably modulated by the context (presence and type of HSPGs), the differentiation status of target cells, and the repertoire of FGFRs these cells express. However, FGF signaling is important for self-renewal and proliferation of primitive ESCs and for migration, proliferation and lineage commitment of more differentiated cells.