Mechanistic insight into how Shh patterns the vertebrate limb Edwina McGlinn and Clifford J Tabin The hands and feet of a newborn baby are a beautiful reminder of the complexity of embryonic patterning. Classical studies on how these structures form have led to a theoretical framework for understanding, in general, how discrete groups of cells can instruct differential fates across a wider field through the action of long-range signals. The discovery just more than a decade ago that localized expression of Sonic hedgehog (Shh) differentially patterns structures across the limb field, resulting in digits with unique characteristics, provided a starting point for readdressing these models at a molecular level. Current research has revealed unexpected complexity in how a gradient of Shh activity is both established and received, prompting re-evaluation of the nature of patterning mechanisms within the limb. Addresses Department of Genetics, Harvard Medical School, 77 Avenue Louis Pasteur, Boston, MA 02115, USA Corresponding author: Tabin, Clifford J (tabin@genetics.med.harvard.edu) Current Opinion in Genetics & Development 2006, 16:426���432 This review comes from a themed issue on Pattern formation and developmental mechanisms Edited by Rick Firtel and Magdalena Zernicka-Goetz Available online 23rd June 2006 0959-437X/$ ��� see front matter # 2006 Elsevier Ltd. All rights reserved. DOI 10.1016/j.gde.2006.06.013 Introduction Understanding how the fate of a cell is determined during embryogenesis is an extraordinarily challenging task. An important way of conceptualizing this problem, first pro- posed by Lewis Wolpert [1,2], is the idea of ���positional information���, whereby cells are first instructed of their location within a developing field and then subsequently respond appropriately in order to attain the correct fate. Discrete signaling centers within the early embryo have been proposed to orchestrate positional information by the release of diffusible signals termed morphogens. Within a field of cells, these morphogens elicit differential inductive responses depending on the distance of the cell from the signaling center and hence the precise morpho- gen concentration. Patterning along the anterior���poster- ior (AP) axis of the vertebrate limb ��� from the thumb to the little finger ��� has proved to be an extremely favorable paradigm for studying these basic principles of embryonic patterning. Outgrowth and patterning of the vertebrate limb is coordinated by specialized signaling centers identified through classical embryological manipulations (Figure 1a). A discrete region of the posterior mesoderm was capable of dramatically reorganizing the AP axis when grafted into the anterior���distal margin of a second limb bud, inducing mirror-image duplications in digit number and identity [3]. This region, termed the zone of polarizing activity (ZPA), is remarkably conserved at early stages of development among vertebrate species, despite obvious differences among phyla insofar as the shape and size of resulting skeleton (Figure 1b). The demonstration that the secreted molecule Shh is expressed exclusively within the ZPA and is capable of recapitulating the polarizing potential of the ZPA [4] suggested that Shh was indeed the posteriorizing factor and led to intense research on how this signal orchestrates morphological output. Far from Shh activity reflecting a linear diffusion gradient of protein as initially thought, recent work has shown that the activity of Shh is exqui- sitely sculpted along the AP axis by both the regulation of Shh protein distribution (spatial gradient) and the time of exposure (temporal gradient). This review focuses on the relative contribution of these mechanisms to limb pat- terning, along with recent insights into how the Shh signal is transduced and how size control in the limb is achieved. Establishment of an Shh activity gradient in the limb Tight posterior restriction of Shh activity is paramount for correct AP patterning of the limb. Shh mRNA is localized to the posterior [4] and, importantly, Shh has the ability to sense and regulate its own mRNA levels in order to compensate for fluctuations that would perturb patterning [5]. During normal development, a gradient of Shh pro- tein can be observed extending from the posterior [6]. Experimentally, when the concentration of exogenous Shh in the anterior chick mesoderm was incrementally increased by manipulating ZPA cell number [7] or abso- lute Shh concentration [8], ectopic digits with increasing posterior identity were sequentially specified, supporting a concentration-dependent model of Shh action. These data are consistent with how Shh is known to pattern the neural tube, for which a spatial gradient alone specifies differential neuronal identity [9]. A key breakthrough in understanding how Shh patterns the limb came following the discovery that Shh-expressing cells proliferate dra- matically, and themselves contribute to a significant fraction the final limb skeleton [10 ], challenging the idea that an Shh activity gradient is established purely by diffusion or transport. Current Opinion in Genetics & Development 2006, 16:426���432 www.sciencedirect.com

Temporal gradient When expression of Cre-recombinase was driven under theendogenous Shh-promoter,allcellswitha historyof Shh expression were permanently marked in a reporter mouse- line [10 ]. In this experiment, it became evident that Shh- expressing cells of the ZPA undergo extensive prolifera- tion over time to contribute to all of digits 5 and 4, along with a subset of cells within digit 3 and the ulna. These data correlate well with the skeletal structures that were missing following genetic removal of Shh in mice, with the exception of digit 2, which does not form in Shh / mutant limbs yet is composed of cells that never express Shh [11]. Giventhatdigit1formsindependentlyofShhactivity[11], this means that digit 2 is the only digit entirely reliant on a non-cell-autonomous effect of Shh. Fate mapping of Gli1-expressing mouse cells, which provides a robust read-out of Shh pathway activation, confirmed that digits 2 to 5 are responsive to Shh [12 ]. Importantly, release of Shh from expressing cells was required for formation of digit2,confirmingthataspatialgradientofShhcontributes to digit patterning [10 ]. The observation that Shh-des- cendant cells do not comprise the entire digit 3 suggests that both cell-autonomous and non-cell-autonomous Shh function patterns this digit however, digit 3 is still able to form in the absence of long-range Shh signaling [10 ]. To explore the temporal contribution of Shh-expressing cells to the descendant pool, expression of an inducible Cre driven under the endogenous Shh-promoter was used to mark cells at varying stages of development [10 ]. It was shown that Shh-expressing cells contributing to digit 3 move away from the ZPA early in development, with digit 4 and then digit 5 being sequentially specified with increasing developmental time (temporal gradient). These data are consistent with earlier observations in chick where the same anterior cell population was able to contribute to either an extra digit 2 or an extra digit 3 dependent on the time of exposure to a constant source of exogenous Shh [8] this suggests a ���ratchet��� effect [13] whereby cells undergo a sequential and irreversible spe- cification of fate from anterior to posterior identity, a model first proposed by Cheryl Tickle [14]. A summary of how Shh patterns each individual digit is presented in Figure 2. It should be noted that early fate-mapping experiments in chick are consistent with significant expansion of cells from the ZPA [15], suggesting phylo- genetic conservation of this patterning mechanism. It remains unclear how extended exposure to Shh mod- ulates digit identity. In particular, it is unknown whether cells fated to become a digit 3, a digit 4 or a digit 5 receive a uniform concentration of Shh, albeit for different lengths of time. The observation that Gli1 expression in the very posterior limb is reduced as more posterior digits become specified [12 ] suggests that, over time, this particular locus loses its competency to respond to Shh, but whether this extends to other Shh target genes such as Ptc1, Ptc2 or Bmp2 has not been tested. Mechanistic insight into how Shh patterns the vertebrate limb McGlinn and Tabin 427 Figure 1 Signaling centers that orchestrate final morphology of the vertebrate limb. (a) Key signaling centers in the early vertebrate limb include the zone of polarizing activity (ZPA) and the apical ectodermal ridge (AER) [57]. (b) Final skeletal structure varies between species. In particular, digit number exhibits the greatest degree of variation, exemplified here in mouse and chick. www.sciencedirect.com Current Opinion in Genetics & Development 2006, 16:426���432