The alteration of a number of transcription factors have been found to impact limb length in mammals through their roles in patterning and differentiation of the limb bud. For example, Paired Related Homeobox 1 (Prx1 or Prrx1) is a homeobox transcription factor known for its role in mesodermal cell proliferation and fate in the developing limb. In an elegant experiment, the limb specific transcriptional enhancer of mouse Prx1 was replaced by the orthologous enhancer from bat, Carollia perspicillata (Cretekos et al., 2008). This manipulation resulted in increased expression and an expansion of the expression domains of mouse Prx1, and an increase in the overall length of the mutant mouse limbs (Cretekos et al., 2008).

HOX genes, a group of highly conserved transcription factors that are essential for limb patterning also impact the length of the limb structures (Zakany and Duboule, 2007). Mouse knockouts of HoxD13, HoxA13, or HoxD12 result in both the truncation of the limb pattern and reduction of the overall limb size (Fromental-Ramain et al., 1996; Hérault et al., 1996; Cho et al., 2008). Increased and sustained expression of the HoxD locus occurs in the developing forelimb buds in bats. While these differences in expression do not result in noticeable differences in the growth and size of the fore and hind limb buds at the early stages, once differentiated, the skeletal elements in the autopod segment of bat forelimbs undergo a dramatic elongation, resulting in their proportionally larger size. Thus, loss of limb specific Hox genes appear to result in shortened limbs by negatively impacting pattern formation, while increased Hox expression positively correlates with limb size by increasing growth during the elongation stage of limb development.

Sox9 and paralogs Sox5 and Sox6 are members of the SRY-related HMG-box family of transcription factors and effect limb size through their regulation on chondrogenesis (Liu and Lefebvre, 2015). During embryonic limb development, Sox9 is considered the master chondrogenic factor, required for differentiation of mesenchymal precursor cells into chondrocytes (Lefebvre et al., 2001; Liu and Lefebvre, 2015). Sox9 then works in concert with Sox5 and Sox6 to drive differentiation and proliferation of chondrocytes (Lefebvre et al., 2001; Liu and Lefebvre, 2015). Activating mutations in Sox9 in mice results in a long limb phenotype (Long et al., 2020), while inhibiting mutations in the same gene results in short limb phenotypes (Akiyama et al., 2002, 2007). Furthermore, mouse knockouts of Sox5 and Sox6 in the limb bud mesenchyme results in chondrodysplasia with shortened limbs (Smits et al., 2001; Dy et al., 2008). These observations highlight the importance of this family of transcription factors on the regulation of growth during limb development.

Once the limb bud is patterned and the skeletal tissues have differentiated, the regulation of the long bone growth plates greatly contributes to the overall size of the adult limb. The genes that regulate limb growth at this stage are involved with paracrine factor signaling. One example of this is Indian hedgehog (Ihh) signaling, which positively regulates cell proliferation within the growth plates of the long bones. When Ihh signaling is inactivated, through null Ihh or mutations in Ihh transducers or effectors, the resulting mammalian limbs are severely shortened (Mo et al., 1997; St-Jacques et al., 1999; Long et al., 2001; Razzaque et al., 2005; Ruiz-Perez et al., 2007; Sohaskey et al., 2008; Joeng and Long, 2009; Caparrós-Martín et al., 2013; Yoshida et al., 2015; Zhang et al., 2015; Kim et al., 2018). In contrast, overexpressing Ihh in the developing chick limb through viral transfection resulted in increased limb length (Bren-Mattison et al., 2011). These effects on limb size are generally tied to altered Ihh signaling during the processes of chondrocyte proliferation and differentiation and osteoblast differentiation in the growth plates in the long bones (Minina et al., 2002).

Interestingly, FGF activity has a differential impact on cell division depending on the stage of limb development. Studies in mammals and amphibians have shown that FGF signaling is essential for proliferation in the limb bud mesenchyme, while during post-embryonic limb maturation, FGFs participate in a negative feedback loop with Ihh in the growth plates (Coffin et al., 1995; Mancilla et al., 1998; Minina et al., 2002; Purushothaman et al., 2019). Gain-of-function mutations in both human and mice FGFR3 result in achondroplasia characterized by a short limb phenotype (Iwata et al., 2000, 2001; Lee et al., 2017; Segev et al., 2000), while knockout of FGFR3 in mice produces a long limb phenotype (Eswarakumar and Schlessinger, 2007; Toydemir et al., 2006; Tseng et al., 2010; Wen et al., 2016) (Table 1). Furthermore, knockout of FGFR3 has been directly tied to increased Ihh and BMP signaling within the elongating skeletal tissue in mice (Wen et al., 2016).

BMPs also participate in a negative feedback loop with FGFs in the developing limb, and the inhibition of FGFs by BMPs is particularly important for the activation of Sox9 expression, which is essential for chondrogenesis of the developing skeletal tissue in avian and mammalian limb buds (Chimal-Monroy et al., 2003; Yoon et al., 2006; Norrie et al., 2014). The negative feedback between BMP and FGF signaling is also present in the growth pates of mammalian long bones (Olsen et al., 2000; Chen et al., 2012; Studer et al., 2012; Wei et al., 2016). Overexpression of BMP2 and BMP4 ligands increases skeletal element size during chick limb development (Duprez et al., 1996). Moreover, inhibiting BMP signaling, via mutations in the receptors or downstream genes, leads to a shortened limb phenotype in mouse models (Evers et al., 1996; Barna et al., 2000; Settle et al., 2003; Klüppel et al., 2005; Lallemand et al., 2005; Aizawa et al., 2012; Bhattacharyya et al., 2015; Zhang et al., 2020). BMP signaling is essential for chondrocyte proliferation and differentiation in mouse growth plates (Yoon et al., 2006). Additionally, when BMP signaling is not present, FGFR1 expression is elevated, which further represses the elongation of the long bones in both chicken and mouse models (Chimal-Monroy et al., 2003; Yoon et al., 2006; Norrie et al., 2014).

Both long and short limb phenotypes are additionally observed in mutations that affect TGFβ signaling. TGFβ’s regulate the construction and destruction of skeletal tissue by modulating the activity of osteoblasts and osteoclasts, respectively (Tang et al., 2009). Mutations in the human TGFβ1 gene causes Camurati-Engelmann disease, one characteristic of which is elongated limbs (Kinoshita et al., 2000; Janssens et al., 2003). Additionally, mutations in Fibrillin1 (FBN1), a TGFβ-binding partner, can lead to congenital syndromes (Marfan syndrome and Weill-Marchesani) in humans that result in either elongated or shortened limbs (Goff et al., 2011; Quarto et al., 2012). Fibrillin1 is an extracellular matrix glycoprotein necessary for microfibril associated signal transduction (Goff et al., 2011; Quarto et al., 2012). The human mutations largely reside in the TGFβ-binding domain, decreasing FBN1’s ability to sequester TGFβ ligands in the extracellular matrix, and increasing the bioavailability of TGFβ ligands (Goff et al., 2011; Quarto et al., 2012). It is unknown how the increased TGFβ activity observed in both Marfan syndrome and Weill-Marchesani syndromes lead to long and short limbs respectively, but the key difference might rely on the cell types that TGFβ signaling is hyperactivated in.

C-type natriuretic peptides, mostly known for their role in kidney function, also play a crucial role in limb sizing through chondrocyte regulation (Potter et al., 2009). Natriuretic peptide type C (NPC) activates the receptor (NPR-B or NPR2) to drive the synthesis of the second messenger, cGMP (Potter et al., 2009). In human patients, loss-of-function mutations in NPR2 result in shortened limbs, while gain-of-function mutations cause Acromesomelic Dysplasia, Maroteaux Type, characterized by elongated limbs (Bartels et al., 2004; Faivre et al., 2000; Hannema et al., 2013; Ianakiev et al., 2000; Jiao et al., 2007; Kant et al., 1998; Lane and Dickie, 1968; Miura et al., 2012, 2014). The limb length phenotypes due to these mutations appear to be caused by effects on chondrocyte proliferation and differentiation, supporting the hypothesis that the regulation of the long bone growth pates is critical in determining overall scaling of the limb (Lane and Dickie, 1968; Kant et al., 1998; Faivre et al., 2000; Ianakiev et al., 2000; Bartels et al., 2004; Jiao et al., 2007; Potter et al., 2009; Miura et al., 2012, 2014; Hannema et al., 2013).

Limb development and elongation requires that the cells are healthy enough to respond to the factors that regulate allometric growth. Thus, it is not surprising that gene mutations that negatively impact various aspects of cell physiology in the limb bud and immature limb will ultimately impact limb size. All the genes that fall under this category, including those that regulate lipid biosynthesis (Wanders et al., 1992; Clayton et al., 1994; Ofman et al., 1998; Thai et al., 2001; Rodemer et al., 2003; Nimmo et al., 2010; Itzkovitz et al., 2012), ion transport (Li et al., 1999; Kornak et al., 2001; Neutzsky-Wulff et al., 2008; Camacho et al., 2010; Weinstein et al., 2014), cell proliferation, and DNA damage repair (Vernersson Lindahl et al., 2013) have only been found to negatively impact limb size in mice and humans, suggesting that these factors may play permissive rather than instructive roles.

The maintenance of the appropriate limb size during tissue homeostasis depends on both the developmental stage of the animal, and whether it is a determinant or indeterminately growing species (Figure 1). In animals that have determinant growth, the lenth of limbs can be impacted up until the initiation of adulthood. In humans, limb elongation ends in late puberty, when the growth plates fuse and are no longer susceptible to the signals that promote their growth (reviewed in Shim, 2015). For example, altered nerve signaling in the limbs of pre-adult humans can result in a phenomenon known as macrodactyly, where one or more digits grows disproportionally larger than the other (Tsuge and Ikuta, 1973; Frykman and Wood, 1978; Razzaghi and Anastakis, 2005). In contrast, indeterminately growing species grow throughout their entire lives, and thus maintain active growth plates as adults (Riquelme-Guzmán et al., 2021). This indicates that growth plate activity must be continuously regulated in these limbs to maintain a size that is proportionally appropriate.

Regeneration of adult limbs presents additional challenges that are nonexistent during embryonic/larval development. The injured limb is much larger than it was during embryonic development, and this larger size must be re-established to regain full function. While humans cannot regenerate their limbs, researchers are actively working to understand the mechanisms by which other species, such as the mouse and the Mexican axolotl (Ambystoma mexicanum), are capable of regenerating with the hopes that the knowledge is transferable to humans. While mice regenerate digit tips, the axolotl are able to regenerate complete limb structures (McCusker et al., 2015; Dolan et al., 2018). Thus, the factors that regulate the growth of the regenerate can have a large impact on the overall size of the limb in the axolotl model. Axolotl are also an indeterminately growing species. This creates an interesting paradigm since the regenerating limb must grow to a size larger than it was at the time of amputation to accommodate the animal’s growing body length. How this growth is regulated is unknown, and studies on this aspect of regeneration in the axolotl are challenging because of the extended period it takes for a regenerated limb to reach its “completed” size.