Similarities between the tail and the extremities
The idea that limb and tail buds present similar development was first mentioned by Hans Grüneberg (1956) and later suggested again by other authors (Schubert et al., 2000). At the same time it also matches the idea of Axis paramorphism as long as the tail is considered as an appendage itself.
Histologically the VER and the AER correspond to an ectodermal epithelial tissue that covers proliferative mesenchyme. In both cases the epithelium/mesenchyme interaction is important for the proliferation of mesodermal cells (Ohta et al., 2007).
In zebrafish there is a ‘tail organizer’, in the sense of been the source of signaling pathway components such as Wnt, Bmp and Nodal (Liu et al., 2004), in a similar way that the ZPA is a source of Shh in the tetrapod limb bud (Bouldin et al., 2010). Additionally, in mouse, a mutation in the gene stratifin produces the phenotype called repeated epilation in which the VER and the AER are very thin and there is an abnormal development of both limbs and tail (Salzgeber and Guénet, 1984; Herron et al., 2005).
Concerning the development of limbs, the Shh pathway has been studied extensively and it is associated with the antero-posterior polarization processes (Bouldin et al., 2010). The presence of Shh was detected in the posterior regions of tetrapod limbs (Riddle et al., 1994), teleost fish (Reifers et al., 1998), the little skate (L. erinacea) and sharks (Chiloscyllum punctatum (Müller and Henle, 1838) and S. canicula) (Hadzhiev et al., 2007; Sakamoto et al., 2009) (Table 1). The caudal fin of zebrafish expresses transcripts of several genes (e.g. ptc and shh) present in the Shh signaling pathway (Krauss et al., 1993; Hadzhiev et al., 2007). In mouse there is expression in the caudal region, however it is in the future spinal cord area (Solloway and Robertson, 1999). As this expression occurs later in development (day 9.5) it is probably not related with the formation of the tail itself.
Table 1. Summary of genes expressed on fin/extremities and caudal fin/tail. *Indirect evidence (for details see on section: Similarities between the tail and the extremities). In bold, genes expressed consistently in all the discussed structures.
Elements from the Bmp pathway are expressed recurrently in both structures (Table 1). For example, Bmp2 is expressed in the chicken AER (Akita et al., 1996), the mouse limb bud (Moon et al., 2000) and the zebrafish pectoral fin (Neumann et al., 1999). It is also present in the mouse VER from the earliest stages until the growth of the tail finishes. Another gene from this pathway is bmp4, in mouse it is present in the AER (Akita et al., 1996), but not in the VER (Catala et al., 1996). In addition, many BMPs have been detected in the caudal fin primordium of zebrafish (Hadzhiev et al., 2007). A final example is Bmp11, which is present in the tail bud and also in the limb bud of Xenopus (Gamer et al., 1999).
Several proteins of the Wnt pathway are found in vertebrate limbs and tails (Table 1). Wnt3a is expressed in mice limbs (Visel et al., 2004), as well as the most caudal portion of the tail bud (Takada et al., 1994). Mice carrying null alleles for wnt3 have truncated tail bud development, but there was no major effect on the extremities (Greco et al., 1996). It could suggest the expression of other genes with redundant functions or the fact that wnt3 is actually involved in other developmental processes on the limb. For chicken wnt3a has been reported in the AER (Kengaku et al., 1998). In zebrafish, Wnt3a is expressed in the AER (Ng et al., 2002) and morpholino knockdowns of wnt8 and wnt3a completely blocked the formation of the tail (Thorpe et al., 2005). Consistent with this phenotype, wnt8 expression is detected at the tip of the tail in zebrafish (Kelly et al., 1995). Moreover, the exogenous application of Wnt8c on the flank of chicken embryos induces the formation of an ectopic limb (Kawakami et al., 2001).
Wnt5a and wnt5b are also expressed in the chicken AER (Loganathan et al., 2005). The first one has a role related with the growth of the underlying mesenchyme (Dealy et al., 1993). The same gene is expressed in the pectoral fins of medaka, Oryzias latipes (Temminck and Schlegel, 1846) (Yokoi et al., 2003). In mouse, wnt5a is involved in the proliferation of branchial arches, facial protrusions, limb bud, VER (Goldman et al., 2000), fingers and genitals. In the mutant wnt5a-/- many of these tissues, including the tail and limbs, present a truncation in their growth (Yamaguchi et al., 1999). The expression of this gene in the branchial arches could also be considered as an argument in favor of the Archipterygium Hypothesis. In addition, wnt5b has a pattern of expression in the tail that is very similar to the one observed for wnt3a (Takada et al., 1994). On the other hand, during the regeneration process of the Xenopus tadpole tail, it is possible to detect the expression of wnt3a and wnt5a (Lin and Slack, 2008).
Another example is wnt11, which is expressed in the tail bud of zebrafish (Makita et al., 1998), chicken (Tanda et al., 1995) and Xenopus (Ku and Melton, 1993), as well as in the limbs of chicken (Tanda et al., 1995) and mouse (Christiansen et al., 1995). Finally, the effector of the Wnt pathway, Lef1, is expressed in the mouse limbs and tail (Oosterwegel et al., 1993).
A very important gene family for limb development corresponds to the Fgf genes; interestingly very few of these genes are expressed in the tail (Table 1). On mouse AER the genes fgf4, fgf8, fgf9 and fgf17 are expressed, but only the latter is present in the VER (Goldman et al., 2000). In zebrafish, fgf10 is expressed in the pectoral fin, tail and gill arches (Thisse and Thisse, 2004). Another gene in this family, fgf24, is expressed in the mesenchyme of the pectoral fin (Draper et al., 2003) and in the tail bud of zebrafish (Abe et al., 2007). No orthologues were found for this gene in tetrapods, but it is present in Chondrichthyes (Draper et al., 2003).
Functionally in mice, the maintenance of the AER depends only on Fgf10 (Norton et al., 2005) and there is no presence of this transcript in the VER (Goldman et al., 2000). The mutant mouse for this gene lacks lungs and anterior and posterior limbs (Sekine et al., 1999). Along the same line, during the regeneration process of the Xenopus tadpole tail there is expression of fgf8, fgf9, fgf10 and fgf20 (Lin and Slack, 2008).
The Sprouty family of proteins is antagonist of receptor tyrosine kinases, including FGF receptors. Spry1, spry2 and spry4 are expressed in the mouse extremities and in the VER (de Maximy et al., 1999; Goldman et al., 2000) (Table 1). In addition, spry2 is expressed in the zebrafish pectoral fin (Fürthauer et al., 2004).
Also, there are a number of common transcription factors between the two structures (Table 1). Several genes of the Msx family (msxA, msxB, msxC and msxD) are expressed in pectoral fins and the fin fold, including the caudal fin of zebrafish (Akimenko et al., 1995). In mice, msx1, functionally related to msxb in zebrafish (Akimenko et al., 1995), is expressed in the VER (Lyons et al., 1992) and AER (Tribioli et al., 2002). Other example is evx1, which is expressed in the mouse limb and the zebrafish fin (Brulfert et al., 1998), as well as in the tails of both organisms (Beck et al., 2001). Another transcription factor that is found in a wide variety of appendages is dll (Panganiban et al., 1997).
The Tbx transcription factors are also important in limb and tail development (Table 1). In chicken, tbx3 is expressed in the AER and in the tail bud, among other structures (Gibson-Brown et al., 1998). In the Japanese newt, Cynops pyrrhogaster (Boie, 1826), cptbx2 is expressed in the tail and the limb (Sone et al., 1999).
The Hox genes are usually related with segmental differentiation, but they also present shared expression between tail and limbs (Table 1). In Mexican axolotls, Ambystoma mexicanum (Shaw and Nodder, 1798), there is expression of hoxb13 and hoxc10 (short transcript) in the tip of the tail as well as in the hindlimb and in lower levels of the forelimbs (Carlson et al., 2001). In mice, hoxd11 (Gérard et al., 1997) and hoxd13 (Dollé et al., 1991) are expressed in the limb and tail bud. In zebrafish, the genes hoxc6a, hoxd12a (Thisse and Thisse, 2004), hoxc8a (Thisse et al., 2001), hoxd13 and hoxa13b (Thisse and Thisse, 2005) are expressed in the pectoral fin and the tail bud. Finally, in S. canicula there is also expression of hoxd in the tail fin (Freitas et al., 2006).
Ledent (2002) proposed a possible relationship between the adult caudal fin of fishes and the autopod of tetrapods. The author suggests that the Hox genes could be responsible for the axis bending which causes the heterocercal condition in fishes in the same way as they are responsible for the proximodistal finger specification of the tetrapod limb. In this scenario, Hox genes would have been recruited secondarily for limb development.
All these similarities between the genetic mechanism involved in limb and tail formation are also congruent with the Axis paramorphism idea (Minelli, 2000, 2003). On this conceptual framework, both structures could be considered as repetitions of the main body axis. Note that the tail is also a structure that presents sexual dimorphism. It has been documented on the tail length of birds (Winquist and Lemon, 1994) and snakes (King, 1989); number of vertebrae in salamanders (Ficetola et al., 2013); and colours on birds (Dakin and Montgomerie, 2013) and fish (Godin and McDonough, 2003).
While it is often possible to identify mutations with a limb phenotype having no consequence in the tail or vice versa, this could be explained by the existence of functional redundancies in one of the tissues.