SEARCH

SEARCH BY CITATION

Keywords:

  • apoptosis;
  • limb;
  • morphogenesis;
  • programmed cell death

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. ICD mechanism
  5. Role of ICD in digit separation
  6. Signaling during ICD activation
  7. Concluding remarks
  8. Acknowledgments
  9. References

Interdigital cell death (ICD) is the oldest and best-studied model of programmed cell death (PCD) in vertebrates. The classical view of ICD function is the separation of digits by promotion of tissue regression. However, in addition, ICD can contribute to digit individualization by restricting interdigital tissue growth. Depending on the species, the relative contribution of either regression or growth-restricting functions of ICD to limb morphogenesis may differ. Under normal conditions, most cells appear to die by apoptosis during ICD. Accordingly, components of the apoptotic machinery are found in the interdigits, though their role in the initiation and execution of cell death is yet to be defined. Fgf8 has been identified as a survival factor for the distal mesenchymal cells of the limb such that ICD can initiate following specific downregulation of Fgf8 expression in the ectoderm overlying the interdigital tissue. On the other hand, Bmps may promote cell death directly by acting on the interdigital tissue, or indirectly by downregulating Fgf8 expression in the ectoderm. In addition, retinoic acid can activate ICD directly or through a Bmp-mediated mechanism. Interactions at different levels between these factors establish the spatiotemporal patterning of ICD activation. Defining the regulatory network behind ICD activation will greatly advance our understanding of the mechanisms controlling PCD in general.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. ICD mechanism
  5. Role of ICD in digit separation
  6. Signaling during ICD activation
  7. Concluding remarks
  8. Acknowledgments
  9. References

Programmed cell death (PCD) is a genetically-controlled process by which certain cells destroy themselves at a particular developmental stage and within specific tissues or organs. PCD is required to remove transitory structures, to sculpt tissues and to eliminate damaged cells that can be harmful to the organism. Secondary palate shelf fusion (Cuervo et al. 2002), Müllerian duct regression (Roberts et al. 1999) and interdigital cell death (ICD; Ballard & Holt 1968; Jacobson et al. 1997) are some examples in vertebrate embryos where PCD is necessary for the correct formation of the organism.

Programmed cell death has been classified into three types according to the morphology that cells acquire while they are dying and the molecular machinery involved: PCD type I or apoptosis; PCD type II or autophagy, and PCD type III also described as lysosomal cell death (Clarke 1990). Of these, only PCD type I has been unambiguously detected in the mouse embryo in association with a developmental process (Clarke 1990; Penaloza et al. 2008). This does not exclude other types of PCD from contributing to cell elimination during mouse development, but more likely reflects that, under normal conditions, apoptosis is the predominant PCD mechanism. Autophagic cells have been particularly detected in tissues where abundant cell death is occurring. In contrast with the degeneration of certain tissues in invertebrates (Melendez & Neufeld 2008), in vertebrates all identified tissues that undergo PCD do so preferentially via apoptotic mechanisms in the developmental context. Cells undergoing apoptosis show a series of physical and biochemical changes such as plasmatic membrane blebbing, loss of mitochondrial membrane potential, caspase-activation, DNA fragmentation (e.g. terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling [TUNEL]-positive; (Gavrieli et al. 1992), and finally, cell disintegration into apoptotic bodies, which are engulfed by specialized cells.

Cell death in the limb mesenchyme

One of the best models to study PCD is the developing limb. This is due to the apparently evident association between tissue degeneration and cell death; however, a direct analysis of this potential association has only until recently emerged (Hernandez-Martinez et al. 2009) (see below). The first area of apoptosis detected in the developing limb is in the denominated anterior necrotic zone (ANZ). This area is detected in the anterior mesenchyme of the chick limb at Hamburger and Hamilton stage 21 HH; (Hamburger & Hamilton 1992); in the mouse limb it appears at E11 of embryonic development (where E0 is the day of coitus) having a more proximal position when compared with the chick ANZ. Around the same stage of development, at 22 HH, the opaque patch (OP) is detected in the chick limb mesenchyme, which represents apoptotic cells located between the two cartilage condensations that develop into the ulna and radius or the fibula and tibia. This area, which is an extensive region of cells reaching the proximal autopod mesenchyme, is detected at E10.5 in the mouse limb (Fernandez-Teran et al. 2006). The next apoptotic area arises at stage 24 HH in the posterior mesenchyme of the chick wing (Posterior Necrotic Zone – PNZ). Dying cells corresponding to the PNZ are scarcely or not detected in the mouse limb (Chen & Zhao 1998; Fernandez-Teran et al. 2006).

The most notable areas of cell death are evident at E13.5 of embryonic development in the mouse limb and at stage 30 HH, in which dying cells in these regions are located in the mesenchyme between the forming digits, hence referred to as interdigital cell death (ICD) (Fig. 1). Gradients of cell death have been reported to occur in the mesenchyme along the dorso-ventral limb axis; more cells die in the ventral side of interdigital regions of developing mouse limbs, though digit separation is more pronounced at the dorsal side (Salas-Vidal et al. 2001). In the chick limb mesenchyme an apparently equal number of cells die in the dorsal and ventral mesenchyme (Fernandez-Teran et al. 2006). Additionally, in the mouse, more dying cells are detected in the anterior than in the posterior interdigits (Salas-Vidal et al. 2001).

image

Figure 1.  Models of interdigital cell death (ICD). (A) Massive ICD model (sculpting function). The interdigital tissue is first preformed (left). Then, ICD initiates and continues massively (red dots) in order to regress the interdigital tissue (middle). Free digits (right limb) result from the elimination of the interdigital tissue only. (B) Progressive ICD model (differential growth function). Before ICD begins, digital and interdigital regions grow at about the same rate (same size arrows in left limb). ICD initiates (red dots) and limits the growth of interdigital regions such that digits grow faster than interdigits (different size arrows in middle limb). Digits individualize (right limb) as a result of differential growth.

Download figure to PowerPoint

ICD mechanism

  1. Top of page
  2. Abstract
  3. Introduction
  4. ICD mechanism
  5. Role of ICD in digit separation
  6. Signaling during ICD activation
  7. Concluding remarks
  8. Acknowledgments
  9. References

Under physiological conditions ICD occurs mostly by apoptosis. Many cells within the interdigital region show condensed chromatin, are TUNEL-positive, and are contained within macrophage-like cells (Garcia-Martinez et al. 1993; Rotello et al. 1994; Salas-Vidal et al. 1998). Caspase activity has been detected in the interdigital regions of mouse and chick limbs (Nakanishi et al. 2001; Zuzarte-Luis et al. 2006). Accordingly, a broad-spectrum caspase inhibitor markedly reduces apoptosis as determined by chromatin condensation, TUNEL staining and phagocytosis detection (Jacobsen et al. 1996; Chautan et al. 1999). Similarly, limbs of mice deficient in Apaf1, a key regulator of the intrinsic apoptotic pathway, also show reduction in apoptosis in the interdigital regions (Chautan et al. 1999). Interestingly, under conditions of reduced caspase activity provoked in the previous experiments, non-apoptotic cell morphologies have been observed in both chick and mouse developing limbs (Chautan et al. 1999; Zuzarte-Luis et al. 2007; Montero et al. 2010). Whether these non-apoptotic cell morphologies represent alternative forms of cell death that contribute to the natural ICD process remains to be determined, as their presence is related to experiments done in culture or when apoptosis is genetically affected. The recent increase in molecular tools to detect autophagy in vertebrates (Klionsky et al. 2008; Mizushima et al. 2010), without relying uniquely on lysosomal activity (that also detects phagocytosis) or intracellular morphology (not appropriate for quantitative analyses) may allow to get a better estimation of the contribution of this form of PCD to ICD.

The specific caspases required for ICD have not been identified. Caspase-3 and caspase-9 activity is detected during mouse ICD (Nakanishi et al. 2001). However, individualized digits are observed in limbs of mice lacking key enzymes controlling apoptosis such as caspase-3, caspase-7 or caspase-9 (Kuida et al. 1996, 1998; Zheng et al. 1999; Lakhani et al. 2006). A direct requirement for caspase-8 in ICD has not been evaluated because its absence causes early embryo lethality (Varfolomeev et al. 1998). It is expected that, due to caspase redundancy, mouse mutants carrying a combination of null and conditional loci will be required to identify the initiator and executioner caspases relevant in ICD.

Bcl2 family members are also expressed in the limb during ICD. The antiapoptotic Bcl2 is expressed in digits (Novack & Korsmeyer 1994), suggesting that apoptotic triggers are acting on all distal cells of the autopod, but protection against cell death allows proper digit formation. If this survival activity is relevant, other antiapoptotic Bcl2 family members would need to participate, since Bcl2 deficient mice do not show abnormalities in the limb (Veis et al. 1993). Bax and Bak are two proapoptotic Bcl2 family members that appear to have a role in ICD. Mice deficient in Bax and Bak genes survive to adulthood and show the presence of interdigital tissue in their limbs (Lindsten et al. 2000). Accordingly, Bax is expressed in the distal region of limbs at the time ICD is occurring (Hernandez-Martinez et al. 2009). The proapoptotic BH3-only proteins also play a role in ICD, since adult mice lacking Bim and Bmf possess limbs with remnant interdigital tissue (Hubner et al. 2010). BclxL, a gene that can produce antiapoptotic and proapoptotic proteins by differential splicing, is expressed in interdigits during ICD (Hernandez-Martinez et al. 2009), but because mice null for this gene die early in development (Motoyama et al. 1995), no role can be proposed at this time.

Reactive oxygen species (ROS) are considered relevant in the activation of apoptosis during development (Salas-Vidal et al. 1998). Interestingly, high peroxidase activity and high ROS levels are detected in digits and interdigits, respectively, when ICD is occurring. In cultured limbs, reducing peroxidase activity causes cell death in tips of digits, and reducing ROS promotes interdigital cell survival. Several antioxidant enzyme genes are expressed during ICD, but the extensive compensatory mechanisms controlling intracellular redox state complicate the study of this potentially relevant mechanism in ICD activation. Nonetheless, glutathione peroxidase 4 (Gpx4) appears to be the enzyme responsible for protecting digits from the proapoptotic activity of ROS: Gpx4 is the main glutathione peroxidase present in the limb during ICD and inhibiting this activity causes cell death in the tip of digits (Schnabel et al. 2006). It will be interesting to determine the relationship between this peroxidase activity and the function of antiapoptotic Bcl2 family members in digits.

It is apparent that almost every apoptotic cell detected by condensed chromatin or TUNEL in interdigits is contained within a phagocyte (Rotello et al. 1994; Salas-Vidal et al. 1998). Apoptotic cells produce two classes of signals: the “find-me” and the “eat-me” signals. The first recruits phagocytes to the specific site of apoptosis, whereas the latter promotes the engulfment of cells under apoptosis (Ravichandran & Lorenz 2007). Presently, the molecular mechanisms that govern phagocytosis during ICD have not been studied in detail.

Professional macrophages arriving via blood vessels and infiltrating the interdigital mesenchyme are detected when ICD begins (Hopkinson-Woolley et al. 1994). This phenomenon is easy to visualize in the whole-limb of transgenic mice carrying fluorescent reporters driven by macrophage-specific gene promoters (Rae et al. 2007). The fact that phagocytosis of apoptotic cells in interdigits occurs in the absence of professional macrophages, as in mice lacking the PU.1 gene, suggests that neighboring mesenchymal cells are able to respond to the “eat-me” signals by acquiring phagocytic activity (Wood et al. 2000). Unexpectedly, cells of the interdigital tissue are engulfed by macrophages in mutant limbs lacking Apaf1 and DNAse-II (Nagasaka et al. 2010). This observation contrasts with the marked reduction in the typical lysosomal activity associated with phagocytosis observed in mice lacking Apaf1 or caspase activity (Chautan et al. 1999). Therefore, it is apparent that non-apoptotic interdigital cells produce the “eat-me” signals necessary for their engulfment, but this event is not sufficient to activate the phagocyte’s degradative activity.

During ICD, other proteases besides caspases are active and participate in the remodeling of the interdigit. Matrix metalloproteinases (MMPs) are enzymes that mediate the degradation of the extracellular matrix (ECM). In particular, the gene encoding stromelysine-3 (metalloproteinase-11) is highly expressed in the interdigital regions, and is downregulated in Rarβ−/−/Rarγ−/− and Raldh2−/− mutant mice, which show reduced apoptosis in interdigits (Zhao et al. 2010; Dupe et al. 1999; see below). The regulation of metalloproteinase activity during ICD is poorly studied, nevertheless it is expected that extracellular matrix degradation should accompany any process involving abundant cell death. In particular, it will be interesting to determine whether ICD promotes ECM degradation (i.e. cataptosis) as occurs during the degradation of the secondary palate medial edge seam (Cuervo & Covarrubias 2004).

Role of ICD in digit separation

  1. Top of page
  2. Abstract
  3. Introduction
  4. ICD mechanism
  5. Role of ICD in digit separation
  6. Signaling during ICD activation
  7. Concluding remarks
  8. Acknowledgments
  9. References

Important differences have been found at both the initiation and progression of ICD between limbs of different species. In some cases these patterns of ICD are related to differential gene expression patterns (Weatherbee et al. 2006; Hernandez-Martinez et al. 2009). In the mouse limb, ICD begins in the distal mesenchyme underlying the apical ectodermal ridge (AER) and spreads proximally forming a delta shaped region (Salas-Vidal et al. 2001) (Fig. 1). A similar pattern is observed during the ICD of the snapping and painted turtles, which begins at the distal mesenchyme with very few cells dying in the proximal parts of interdigits (Fallon & Cameron 1977). In the chick limb, dying cells in the interdigits first appear as a cluster in the proximal mesenchyme, and subsequently, a second apoptotic cell cluster is detected in the distal mesenchyme. As development progresses both clusters connect together to form a common domain of cell death, the ICD area (Fernandez-Teran et al. 2006).

Interdigital cell death is frequently referred to as a classic example where PCD functions to sculpt an organ (Ballard & Holt 1968; Jacobson et al. 1997). The very restricted cell death pattern in the developing limb suggests that the degeneration of a preformed interdigital tissue is required to free digits (Fig. 1). In agreement with this model, species with webbed limbs, such as the duck (Ganan et al. 1998) and the bat (Weatherbee et al. 2006), show scarce ICD. Similarly, mutant chick and mouse limbs with reduced ICD are characterized by a phenotype in their limbs known as syndactyly, a generic name given to limbs with joined digits. Simple syndactyly can be considered where evident soft tissue remains between digits, and may not extend to the whole digit length. Tightly joined digits with very little tissue between them characterize a different type of syndactyly. More complex syndactyly involves bone element fusions and/or other digit abnormalities. Syndactyly is also found in limbs with supernumerary digits (i.e. polydactyly). In principle, absent or reduced ICD should result in a phenotype more closely resembling simple syndactyly. It is unlikely that other types of syndactyly result from simply a reduction in ICD. Commonly, reduced distal limb growth is associated with syndactyly. For example, elimination of the N-Myc gene in the limb mesenchyme leads to a syndactyl phenotype characterized by small skeletal elements and a severe reduction in cell proliferation that consequently result in fewer interdigital cells and the absence of ICD (Ota et al. 2007).

Recently, a different view of the process of digit separation was proposed (Salas-Vidal et al. 2001; Hernandez-Martinez et al. 2009). An analysis of digit and interdigit formation indicates that differential growth of digital and interdigital regions occurs during digit individualization (Salas-Vidal et al. 2001) (Fig. 1). This observation suggests that the role of ICD is the control of growth of interdigital regions and thus allows digits to protrude distally. Actually, mouse and chick interdigits never grow as far distal as digits (Hernandez-Martinez et al. 2009). In addition, interdigital distal proliferating cells are the ones that die during ICD, suggesting that the interdigital tissue containing dying cells is progressively formed as digits individualize (Hernandez-Martinez et al. 2009). Different contributions of ICD to interdigital tissue regression or growth may exist among limbs of different species (Figs 1,2). For instance, in mouse, contribution of ICD to tissue regression is minimal, whereas in chick, tissue regression appears to be essential (Hurle & Fernandez-Teran 1983), probably due to the large quantity of interdigital tissue generated in the developing hindlimbs. In contrast, in duck limbs and bat wings, not only should ICD be reduced but signals that promote interdigital growth must remain (Ganan et al. 1998; Weatherbee et al. 2006). Expression of Fgf8 in the interdigital mesenchyme of the developing bat wing (Weatherbee et al. 2006) and the gradual regression in the distal ectoderm of duck interdigital regions (Ganan et al. 1998) support this hypothesis. In mutant limbs, complete simple syndactyly (i.e. when interdigital tissue extends most digit length; Table 1) would be expected only when signals promoting cell death are missing and those required for growth remain. This appears to be the case in mice lacking the Bmp receptor Bmpr1a in the distal ectoderm in which Fgf8 expression is retained in the ectoderm overlying the interdigital regions (Pajni-Underwood et al. 2007). The partial syndactyly observed in mice lacking pro-apoptotic Bcl2 family members is in agreement with the idea that apoptosis is reduced without an increase in interdigital tissue growth (Lindsten et al. 2000; Hubner et al. 2010).

image

Figure 2.  Major regulators of distal and proximal interdigital cell death (ICD). Before ICD begins Fgf8 produced by the distal ectoderm prevents from the cell death-inducing activity of retinoic acid (RA) (A). Distal ICD begins after Bmp2/4 promotes Fgf8 downregulation in the ectoderm of the interdigital regions allowing RA to induce cell death (B). Proximal ICD is also controlled by RA but mediated by Bmp7 (also Bmp4 in the chick); however, the trigger has not been identified yet.

Download figure to PowerPoint

Table 1.   Types of syndactyly developed in mutant mice Thumbnail image of

Signaling during ICD activation

  1. Top of page
  2. Abstract
  3. Introduction
  4. ICD mechanism
  5. Role of ICD in digit separation
  6. Signaling during ICD activation
  7. Concluding remarks
  8. Acknowledgments
  9. References

Bone morphogenetic proteins family members

Most of what we know about signaling during ICD came from experiments done in chick or mouse limbs. Several members of the bone morphogenetic protein (Bmp) growth factor family are good candidates to play a role in the control of ICD. Bmp2, Bmp4, Bmp5 and Bmp7 genes are expressed in the interdigital mesenchyme of chick limbs around the stages at which cell death occurs (Zuzarte-Luis et al. 2004; Geetha-Loganathan et al. 2006). In the mouse, Bmp2 and Bmp7, but not Bmp4, are expressed in the interdigits before and during ICD, while Bmp4 is expressed in the mesenchyme underlying the distal ectoderm (Salas-Vidal et al. 2001) (Fig. 2). Despite the above observations, it is important to note that the expression pattern of most Bmp genes does not precisely correlate with the ICD pattern at different developmental stages; either the Bmp is expressed before any cell death is detected or it is not expressed in the same region where cells are dying. However, because Bmps are secreted factors, gene expression patterns are not sufficient to establish a functional correlation, Therefore, a rigorous analysis of gene expression patterns and protein distribution is needed to accurately identify potential gene functions. In some instances, it is possible that genes expressed in the interdigital regions, particularly at early stages, play no role in ICD but instead in the signaling functions of interdigits, as recently determined for the establishment of digit identity (Dahn & Fallon 2000; Suzuki et al. 2008).

Bmp genes are also expressed in the distal ectoderm during limb development where they regulate the expression of genes that control patterning and growth. Bmp2, Bmp4 and Bmp7 are expressed in the distal ectoderm at different stages of mouse and chick limb development (Francis et al. 1994; Lyons et al. 1995; Geetha-Loganathan et al. 2006). Among these, Bmp4 and possibly low levels of Bmp2 are the only genes expressed around the time ICD occurs.

In contrast with the restricted expression of Bmp genes and their association with the process of cell death, their receptors show a broader expression pattern. For instance, Bmpr1b is expressed in chondrogenic precursors of chick and mouse limb and proximal mesenchyme of the mouse limb bud (Zou et al. 1997; Yi et al. 2000; Barna et al. 2005), whereas Bmpr1a is expressed in the distal mesenchyme of chick limb and ubiquitously in the mouse limb (Zou et al. 1997; Ovchinnikov et al. 2006). Bmpr1a is also expressed in the distal ectoderm of the limb bud (Ahn et al. 2001). The spatiotemporal distribution of physiological secreted Bmp antagonists is also relevant for specific cell responses. Genes encoding Noggin, Gremlin, Follistatin and Chordin are expressed in the mouse and chick limb at several developmental stages (Feijen et al. 1994; Merino et al. 1999; Bardot et al. 2001; Zhang et al. 2002). At the time ICD occurs, Noggin expression is restricted to digits and Gremlin to the proximal interdigital areas (Brunet et al. 1998; Pajni-Underwood et al. 2007). The persistent Gremlin expression in the proximal interdigital mesenchyme in the mouse (Pajni-Underwood et al. 2007) and the duck (Merino et al. 1999) may explain the reduced proximal ICD in both species.

In support of a role of Bmps in ICD in the chick embryo, expression of a dominant negative form of the Bmpr1a in the mesenchyme of early limb buds inhibits cell death in the regions of PCD and results in a syndactyl phenotype (Yokouchi et al. 1996). On the contrary, expression of a constitutively active form of Bmpr1b leads to increased cell death (Zou et al. 1997). The cell death-inducing activity of several Bmp family members is evident when the limb mesenchyme is exposed to them before the onset of cell death (Rodriguez-Leon et al. 1999; Hernandez-Martinez et al. 2009). On the other hand, exposure of the chick interdigital mesenchyme to Bmp antagonists, such as Noggin and Gremlin, reduces ICD (Merino et al. 1999; Rodriguez-Leon et al. 1999). In the mouse, the lack of both Bmp2 and Bmp4 in the limb mesenchyme, which clearly affects ICD and prevents digit separation, is the only evidence of a possible direct role of Bmp in ICD (Bandyopadhyay et al. 2006). Nonetheless, their gene expression pattern in the mesenchyme does not overlap at this stage. Therefore, although a gene redundancy may exist at earlier limb developmental stages, if those Bmps are acting during ICD, a complex gene interaction needs to be considered. Because the selective elimination of Bmpr1a and/or Bmpr1b in the mesenchyme alters limb development, no role of Bmp in ICD can be inferred from the phenotypes obtained (Yi et al. 2000; Ovchinnikov et al. 2006).

Despite the data described above, there are some observations that are inconsistent with a direct role for Bmp in the initiation of ICD, at least in mice. Bmp7 induces cell death in interdigits of mouse limbs, but the administration of Noggin directly to the interdigital mesenchyme does not reduce ICD or alter its morphology (Hernandez-Martinez et al. 2009). Also, inhibition of Smad phosphorylation, a key step in Bmp signaling, has no effect on ICD (Hernandez-Martinez et al. 2009). Bmp genes expressed in the ectoderm, particularly Bmp4, could indirectly participate in the regulation of ICD. Strong evidence for this is the penetrant syndactyl phenotype in mice lacking Bmpr1a specifically in the distal ectoderm (Pajni-Underwood et al. 2007). Accordingly, the elimination of both Bmp2 and Bmp4 in the limb ectoderm reduces ICD and prevents digit separation (Maatouk et al. 2009). A similar effect on ICD is observed in transgenic mice that overexpress Noggin in the distal ectoderm (Guha et al. 2002; Wang et al. 2004). As mentioned below, this indirect Bmp function appears to be mediated by Fgf8. In the chick limb, it is apparent that direct and indirect Bmp functions participate in the control of ICD; the former being relevant for proximal ICD, whereas the latter is important for distal ICD (Hernandez-Martinez et al. 2009). The absence of proximal ICD in the mouse limb supports the role of ectodermal Bmp in the initiation of distal ICD.

Fibroblast growth factor family members

The fibroblast growth factors (Fgf) constitute a large family of growth factors, several of which are essential for limb patterning and morphogenesis. Fgf4, Fgf8, Fgf9 and Fgf17 genes are expressed in the AER of chick and mouse limb, contributing by different degrees to limb development. This was demonstrated by experiments in which Fgf8 loss of function alone cause abnormalities in the limb skeleton, but the severity of the malformations observed increase by removal of other Fgfs expressed in the AER (Lewandoski et al. 2000; Mariani et al. 2008). The Fgfs produced in the AER function as signals that specify distal structures, and promote proliferation and/or survival of the underlying mesenchyme (Sun et al. 2002; Mariani et al. 2008; Yu & Ornitz 2008).

Correlative evidence suggests that Fgf8 promotes cell survival in the interdigital mesenchyme. Genetic manipulations that reduce ICD and promote syndactyly correlate with the persistent Fgf8 expression in the ectoderm overlying the interdigital mesenchyme. This is the case in mice deficient in Bmpr1a in the ectoderm; deficient in Bmp2/Bmp4 either in the ectoderm or mesenchyme; transgenic mice with ectopic expression of Noggin in the ectoderm; and in mice deficient in both Msx1 and Msx2 (Guha et al. 2002; Wang et al. 2004; Lallemand et al. 2005; Pajni-Underwood et al. 2007; Maatouk et al. 2009). On the contrary, the induction of cell death by Bmp7 reduces Fgf8 expression (Hernandez-Martinez et al. 2009). Consistent with this cell survival role, Fgf8 expressed in the mesenchyme of the bat wing appears to counteract the cell death-inducing activity of Bmps in the interdigital mesenchyme, as it can only be revealed when FGF signaling is inhibited (Weatherbee et al. 2006). The fact that Fgf8 expression diminishes precisely at the time ICD begins suggests that loss of survival activity is the trigger for distal ICD (Salas-Vidal et al. 2001; Hernandez-Martinez et al. 2009) (Fig. 2).

Functionally, Fgfs promote cell survival in the interdigital mesenchyme. For instance, Fgfs antagonize the apoptosis-inducing effect of Bmp-soaked beads in the limb bud (Buckland et al. 1998). Furthermore, the topical administration of FGF2- or FGF4-soaked beads in the interdigital chick mesenchyme and Fgf8-soaked beads in the interdigital mouse mesenchyme reduce ICD (Macias et al. 1996; Hernandez-Martinez et al. 2009). More importantly, the overexpression of Fgf4 in the distal limb ectoderm in transgenic mice reduces ICD with the concomitant generation of syndactyly (Lu et al. 2006). Unfortunately, no definitive conclusion can be obtained from the analysis of mice deficient in Fgf8 alone or in combination with other Fgfs expressed in the distal ectoderm, since limbs do not show any evident increase in cell death in the distal mesenchyme, but rather the emergence of proximal ectopic cell death (Sun et al. 2002; Mariani et al. 2008). Nevertheless, as expected if a survival function is associated with an ectodermal Fgf, the pharmacological or genetic reduction of Fgf receptor (Fgfr) activity promotes death of cells of the distal mesenchyme (Yu & Ornitz 2008; Hernandez-Martinez et al. 2009). In addition, complex syndactyly has been observed in humans and mice carrying certain mutations in Fgf receptor genes (Wilkie et al. 2002; Yin et al. 2008); however, because Fgfs have many different functions during limb development, it is not possible to conclude that the phenotypes observed result from a direct effect on ICD.

The binding of Fgf8 to its receptor in the limb mesenchyme activates different intracellular pathways, with the Mapk pathway prominent among them (Corson et al. 2003). Activation of this pathway requires the serial phosphorylation/activation of kinases, and frequently ends in the regulation of a transcription factor by phosphorylation. Mapk1/2, also known as Erk1/2, are the last common kinases activated in this pathway. In concordance with the Mapk pathway being a major intracellular signaling pathway for Fgf during development, phosphorylated Erk1/2 is detected in association with Fgf activity in the developing limb and other embryo regions. Sef, Sprouty and Mkp3, molecules that negatively regulate Fgf signaling, generate a negative regulatory feedback loop, as the expression of their corresponding genes is induced by Fgf (Eblaghie et al. 2002; Li et al. 2007). In particular, Mkp3 expression is associated with Fgf8 activity in the developing limb. Therefore, the Mapk pathway probably mediates the survival activity of Fgf8 during ICD. This hypothesis is supported by experimental evidence showing a marked reduction in phosphorylated Erk1/2 and Mkp3 expression in the interdigital mesenchyme underlying the distal ectoderm at the time ICD initiates, and the requirement for active Erk1/2 for Fgf8 survival activity (Hernandez-Martinez et al. 2009). In contrast with the above observations, one study observed an increase in cell death after Pi3k inhibition and proposed that the Pi3k signaling pathway mediates the Fgf8 survival activity in the distal chick limb mesenchyme by activation of Mkp3 that, in turn, inhibits the apoptotic effects of the Mapk pathway (Kawakami et al. 2003). A possible explanation to this discrepancy derives from experiments confirming the detection of more death cells upon inhibition of the Pi3k pathway (Hernandez-Martinez et al. 2009) but, instead of resulting from increased apoptosis, they accumulate due to a reduction in phagocytic activity (Hernández-Martínez, R. and Covarrubias, L., manuscript in preparation).

Retinoic acid

Retinoic acid (RA) is a lipophilic metabolite of vitamin A. In the first step of synthesis, vitamin A (retinol) is oxidized to retinaldehyde, a reaction that is catalyzed by members of the alcohol deshidrogenase family (Aldhs). In the second synthesis step, retinaldehyde is converted to retinoic acid by the catalytic activity of retinaldehyde deshydrogenases (Raldhs). Inactivation of RA occurs by oxidation, mediated by members of the P450 family such as Cyp26a1, Cyp26b1 and Cyp26c1 enzymes (Fujii et al. 1997). In the embryo, the RA gradients are defined by the fine-tuning of expression of genes encoding synthetic and degradative enzymes as well as RA-binding proteins in the extracellular milieu and cytoplasm. RA binds to heterodimeric nuclear receptors composed of two types of subunit: Rar and Rxr. Each subunit type is represented by three members, which are encoded by different genes (Rarα, Rarβ, Rarγ, Rxrα, Rxrβ and Rxrγ). The expression pattern of these genes in the embryo reveals the possible heterodimers formed and the sites where RA is actively signaling.

Retinoic acid is synthesized in the interdigital regions around the time ICD occurs (Fig. 2). The expression pattern of Raldh2 indicates that the interdigital tissue is a source of RA (Niederreither et al. 1997). Cyp26a1 and Cyp26b1 control the levels of RA in the limb, whose genes are expressed in the distal ectoderm and the mesenchyme underlying the Fgf8-expressing ectoderm, respectively (Maclean et al. 2001). Several Rar and Rxr are expressed in the limb with different patterns (Dolle et al. 1989, 1994). In particular, Rarβ is expressed in the interdigital regions prior to the onset of ICD and, although the isoform Rarβ2/4 is more abundantly expressed in the proximal interdigital mesenchyme, it also extends to the distal portion (Mollard et al. 2000; Salas-Vidal et al. 2001). With a different pattern than Rarβ, the isoform Rarγ2, in addition to its marked expression in precartilagenous tissue around digits, is expressed in the distal interdigital mesenchyme (Mollard et al. 2000). The expression of all three Rxr genes is apparently ubiquitous during mouse embryogenesis (Dolle et al. 1994).

Retinoic acid is a potent inducer of cell death in several embryonic regions of the embryo. In fact, at least some of the limb abnormalities produced by the teratogenic activity of RA are due to cell death (Yashiro et al. 2004; Pennimpede et al. 2010). Indeed, this compound can induce apoptosis in the mesenchyme of both chick and mouse limbs (Kochhar et al. 1993; Rodriguez-Leon et al. 1999). The role of RA in PCD is supported by the marked reduction in ICD after blocking RA synthesis by the targeted inactivation of Raldh2 in mice (Zhao et al. 2010). Also synthetic antagonists of Rar type of receptors reduce ICD (Rodriguez-Leon et al. 1999; Hernandez-Martinez et al. 2009). However, the individual elimination of any Rar or Rxr gene does not cause any abnormality that can be associated with an effect on ICD (Kastner et al. 1994, 1997; Sucov et al. 1994; Krezel et al. 1996; Cash et al. 1997; Ghyselinck et al. 1997). The only combination of null Rar alleles that reduces ICD and cause syndactyly is Rarβ−/−/Rarγ−/− (Ghyselinck et al. 1997; Dupe et al. 1999; Zhao et al. 2010). Rarβ and Rarγ may be redundant genes in the interdigital region, but a complex gene interaction cannot be discarded. Probably the three Rxr are redundant in the limb, but Rxrα seems to be the most important as the RA effects on the limb are not observed in mice deficient in Rxrα (Kastner et al. 1994; Sucov et al. 1994; Krezel et al. 1996). In addition, the combination of null alelles Rxrα+/−;Rarγ−/− produces syndactyly (Kastner et al. 1997).

The mechanism by which RA induces apoptosis is unknown. In cell cultures, tissue-transglutaminase gene expression and levels of ROS increase upon RA treatment in association with increased apoptosis (Davies et al. 1985; Jiang & Kochhar 1992; Lee et al. 2003). Tissue-transglutaminase mRNA is detected in the interdigital regions and is absent in double mutant mice Rarβ−/−/Rarγ−/− (Dupe et al. 1999; Zhao et al. 2010). On the other hand, in addition to the high levels of ROS detected in the interdigital regions, RA induces an increase in ROS levels and dying cells in the developing limb, particularly evident in the tip of digits (Schnabel et al. 2006). Interestingly and in agreement with a direct role of RA in the activation of ICD, the promoter region of the proapoptotic genes Bax and Bak contain putative retinoic acid responsive elements (RARE) (Hernandez-Martinez et al. 2009). Accordingly, RA increases and an Rar antagonist decreases, Bax expression (Hernandez-Martinez et al. 2009).

Based on experiments performed in the chick limb, an alternative indirect mechanism has been proposed for the RA-mediated regulation of ICD (Rodriguez-Leon et al. 1999). Local application of RA in the interdigital mesenchyme of chick limbs upregulates Bmp4 and Bmp7 expression in association with an increase in cell death. Under this condition, Noggin prevents cell death suggesting that Bmp4 and/or Bmp7 mediate the RA-induced cell death. The role of endogenous RA in ICD through this mechanism is supported by the reduction in ICD and Bmp7 expression in the presence of a Rar antagonist (Rodriguez-Leon et al. 1999). Interdigital mouse Bmp7 expression is also positively regulated by RA as indicated by the downregulation observed in limbs treated with Rar antagonists (Hernández-Martínez, R. and Covarrubias, L., unpubl. data), and in Rarβ−/−/Rarγ−/− or Raldh2−/− mutant mice (Dupe et al. 1999; Zhao et al. 2010). However, as mentioned above, Bmps do not appear to directly activate ICD in the mouse. In concordance with this conclusion and in contrast with the observations in chick limbs, induction of cell death by exogenous RA in mouse limbs is not prevented by Noggin, despite Bmp7 expression being markedly increased (Hernandez-Martinez et al. 2009). An explanation for this discrepancy could be that only proximal ICD is directly controlled by Bmps, whereas distal ICD, which predominates in the mouse, is more dependent on ectodermal Fgf8 levels (Hernandez-Martinez et al. 2009) (Fig. 2). An additional indirect mechanism by which RA can regulate ICD is by downregulating Fgf8 expression and Fgf8 signaling (Hernandez-Martinez et al. 2009). Concomitantly, Fgf8 could also promote cell survival by reducing RA levels, as Fgf8 can downregulate Raldh2 expression and upregulate Cyp26b1 expression in the limb (Hernandez-Martinez et al. 2009).

Msx1/2 transcription factors

Msx1 and Msx2 (Msx1/2) are two closely related transcription factors expressed in the developing limb of mouse and chick (Yokouchi et al. 1991; Ganan et al. 1998; Lallemand et al. 2005). Msx1/2 expression in the interdigital mesenchyme suggests a possible participation in ICD (Houzelstein et al. 1997; Ganan et al. 1998; Rodriguez-Leon et al. 1999). Supporting this view, Msx1/2 expression increases in the chick limb mesenchyme in association with the cell death-inducing activity of Bmps (Rodriguez-Leon et al. 1999), and overexpression of Msx2 in the same tissue generates alterations in patterning, apparently due to an increase in cell death and levels of Bmp4 expression (Ferrari et al. 1998). In contrast, however, reduction in ICD in limbs with altered Bmp signaling does not always correlate with downregulation of endogenous Msx1/2 expression. For example, in the absence of Bmp2 and Bmp4, the mesenchyme shows no obvious reduction in interdigital Msx1/2 expression (Bandyopadhyay et al. 2006). Msx1/2 endogenous expression is also unchanged in the interdigits of transgenic mouse limbs overexpressing Noggin in the distal ectoderm, or in those of mice deficient in ectodermal Bmpr1a (Wang et al. 2004; Pajni-Underwood et al. 2007). Downregulation of Msx2 expression is observed in the interdigits of mice deficient in Bmp2 and Bmp4 in the distal ectoderm and in mice lacking Bmp2 and Bmp7 as a consequence of Rac1 inactivation in the mesenchyme (Maatouk et al. 2009; Suzuki et al. 2009).

If Msx1/2 plays a role in ICD, they unlikely do it alone, as expression of both Msx1 and Msx2 in the interdigits of mouse limbs precedes the onset of ICD (Salas-Vidal et al. 2001). In addition, during chick ICD, Fgf8 upregulates Msx2 expression in the interdigital mesenchyme in correlation with its survival activity (Hernandez-Martinez et al. 2009). How can we reconcile such observations with the fact that limbs of mice deficient in both Msx1 and Msx2 show syndactyly and reduced ICD (Lallemand et al. 2005)? Early in development Msx1/2 are expressed in the AER and later on, in addition to the expression in the interdigital regions, they are also expressed in the distal ectoderm where they remain in digits during ICD (Lallemand et al. 2005). Fgf8 expression in mice lacking Msx1 and Msx2 persists in the interdigital ectoderm (Lallemand et al. 2005). Therefore, Msx1/2 may participate in the control of ICD by negatively regulating Fgf8 expression, probably in response to Bmp signaling in the ectoderm, as described above.

Other genes

Wnt signaling could also participate in the control of ICD. Overexpression of β-catenin in the distal ectoderm of transgenic mouse limbs leads to a reduction in ICD (Villacorte et al. 2010). This effect is not caused by changes in Bmp2, Bmp4 or Bmp7 expression, but it appears to relate to an AER alteration resulting in persistent Fgf8 expression (Villacorte et al. 2010). Dkk1, a natural antagonist of Wnt signaling, is present in limb regions where PCD occurs (Grotewold & Ruther 2002). Interestingly, overexpressing Dkk1 in chick limb buds as with the topical administration of the gene product induces cell death in the mesenchyme (Grotewold & Ruther 2002). Mice deficient in Dkk1 develop syndactyly characterized by the fusion of digits (Mukhopadhyay et al. 2001). More studies will be needed to define the role of Wnt signaling in the control of ICD, both in the ectoderm and the mesenchyme.

Members of the ADAMTS family are a class of metalloproteinases that appear to participate in the regulation of ICD. Genes encoding several members of this family (e.g. ADAMTS5 and ADAMTS9) are expressed in the interdigital regions, though the expression appears absent in the distal regions where ICD begins (Mcculloch et al. 2009). A combination of null alleles for different ADAMTS reduces ICD, as determined by TUNEL and lysosomal activity, and causes other limb abnormalities (Mcculloch et al. 2009). The role of ADAMTS in the control of ICD is mediated by the processing of Versican, a protegycan present in digital regions (Mcculloch et al. 2009). The proapoptotic activity of processed Versican in interdigits and the ability of exogenously added Bmp4 to induce apoptosis in the absence of processed Versican (Mcculloch et al. 2009), suggest that a permissive or non-permissive environment for the action of proapoptotic inducers is generated by processed or unprocessed Versican, respectively. Other components of the ECM, such as Fibrillin 2 and Laminin α5, also play a role in autopod development, but their role in ICD remains to be determined directly (Miner et al. 1998; Arteaga-Solis et al. 2001).

Other genes participating in the regulation of ICD have been identified from the limb phenotype observed after loss-of-function studies performed by specific gene targeting in mice. For instance, genetic elimination of mesenchymal Rac1 during limb development decreases ICD with the concomitant downregulation of Bmp2, Bmp7 and Msx1/2 expression in the interdigital mesenchyme (Suzuki et al. 2009). Digits of Tgfβ2−/−/Tgfβ3−/− mice do not appear separated and ICD is not detected; however, syndactyly in this case is probably due to the marked reduction in autopod growth observed (Dunker et al. 2002). In another example, adult mice deficient in Jagged2 or deficient in both Notch1 and Notch2 have syndactyly in their limbs with the middle digits tightly fused by soft-tissue or bone (Pan et al. 2005). Although it is unlikely that a direct effect on ICD is the cause of these later phenotypes, a final conclusion will require an analysis of cell death in these mice. As several mouse mutagenesis projects move on, it is likely that more genes participating in ICD are identified by the syndactyl phenotype.

Concluding remarks

  1. Top of page
  2. Abstract
  3. Introduction
  4. ICD mechanism
  5. Role of ICD in digit separation
  6. Signaling during ICD activation
  7. Concluding remarks
  8. Acknowledgments
  9. References

Interdigital cell death can no longer be perceived as a simple sculpting process. Limb development is a dynamic process, where cell differentiation, migration, proliferation and death occur simultaneously in restricted domains generating distinctive morphogenetic patterns at every stage. The recasting of ICD as a dynamic process opens the path for new functions and regulatory mechanisms to be introduced into the model. Instead of being a single process, ICD appears to be a set of at least two PCD processes that occur about the same time but have different functions. As proposed here, the two ICD functions contribute to the individualization of digits, though by a different mechanism. The main purpose of distal ICD is to restrict interdigital tissue growth such that digits protrude distally, whereas the role of proximal ICD is to separate digits by degenerating the tissue between them, closer to the classical view of ICD (Figs 1,2). As expected, the prevalence of each process appears to depend on the species, since the requirements to shape the adult limb of each species are likely different.

The dissection of ICD in two processes also implies that each is subject to different regulatory mechanisms (Fig. 2). Common molecules regulate distal and proximal ICD; however, the mode of action and interaction appears to be different. We propose that RA induces distal ICD by a direct mechanism, whereas induction of proximal ICD by RA is a Bmp-mediated mechanism. On the other hand, Bmps appear to directly control proximal ICD, whereas Bmps regulate distal ICD by an Fgf8-mediated mechanism.

Gene manipulations in mice and chick have contributed to the identification of molecules involved in the control of ICD. However, due to the short time window in which ICD occurs and the multiple functions of the same molecules during limb development, novel strategies need to be developed in order to study the role of specific proteins in time windows as short as few hours. This will be the only way to determine the signaling network on which ICD initiation depends. ICD is a long-established model for PCD but, as we learn more about the process, it is evident that many questions still remain to be answered. It will not be a surprise if the answers to these questions are relevant to the function and regulation of PCD in other embryo regions.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. ICD mechanism
  5. Role of ICD in digit separation
  6. Signaling during ICD activation
  7. Concluding remarks
  8. Acknowledgments
  9. References

We apologize to those research groups whose work was not included in this review due to the limitation of space. We thank present and past members of the Laboratory on Tissue Degeneration and Regeneration for the fruitful discussions that contributed to developing some ideas presented in this review. This work was supported by DGAPA-UNAM (IN225910).

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. ICD mechanism
  5. Role of ICD in digit separation
  6. Signaling during ICD activation
  7. Concluding remarks
  8. Acknowledgments
  9. References
  • Ahn, K., Mishina, Y., Hanks, M. C., Behringer, R. R. & Crenshaw, E. B. 3rd 2001. BMPR-IA signaling is required for the formation of the apical ectodermal ridge and dorsal-ventral patterning of the limb. Development 128, 44494461.
  • Arteaga-Solis, E., Gayraud, B., Lee, S. Y., Shum, L., Sakai, L. & Ramirez, F. 2001. Regulation of limb patterning by extracellular microfibrils. J. Cell Biol. 154, 275281.
  • Ballard, K. J. & Holt, S. J. 1968. Cytological and cytochemical studies on cell death and digestion in the foetal rat foot: the role of macrophages and hydrolytic enzymes. J. Cell Sci. 3, 245262.
  • Bandyopadhyay, A., Tsuji, K., Cox, K., Harfe, B. D., Rosen, V. & Tabin, C. J. 2006. Genetic analysis of the roles of BMP2, BMP4, and BMP7 in limb patterning and skeletogenesis. PLoS Genet. 2, e216.
  • Bardot, B., Lecoin, L., Huillard, E., Calothy, G. & Marx, M. 2001. Expression pattern of the drm/gremlin gene during chicken embryonic development. Mech. Dev. 101, 263265.
  • Barna, M., Pandolfi, P. P. & Niswander, L. 2005. Gli3 and Plzf cooperate in proximal limb patterning at early stages of limb development. Nature 436, 277281.
  • Brunet, L. J., Mcmahon, J. A., Mcmahon, A. P. & Harland, R. M. 1998. Noggin, cartilage morphogenesis, and joint formation in the mammalian skeleton. Science 280, 14551457.
  • Buckland, R. A., Collinson, J. M., Graham, E., Davidson, D. R. & Hill, R. E. 1998. Antagonistic effects of FGF4 on BMP induction of apoptosis and chondrogenesis in the chick limb bud. Mech. Dev. 71, 143150.
  • Cash, D. E., Bock, C. B., Schughart, K., Linney, E. & Underhill, T. M. 1997. Retinoic acid receptor alpha function in vertebrate limb skeletogenesis: a modulator of chondrogenesis. J. Cell Biol. 136, 445457.
  • Chautan, M., Chazal, G., Cecconi, F., Gruss, P. & Golstein, P. 1999. Interdigital cell death can occur through a necrotic and caspase-independent pathway. Curr. Biol. 9, 967970.
  • Chen, Y. & Zhao, X. 1998. Shaping limbs by apoptosis. J. Exp. Zool. 282, 691702.
  • Clarke, P. G. 1990. Developmental cell death: morphological diversity and multiple mechanisms. Anat. Embryol. (Berl.) 181, 195213.
  • Corson, L. B., Yamanaka, Y., Lai, K. M. & Rossant, J. 2003. Spatial and temporal patterns of ERK signaling during mouse embryogenesis. Development 130, 45274537.
  • Cuervo, R. & Covarrubias, L. 2004. Death is the major fate of medial edge epithelial cells and the cause of basal lamina degradation during palatogenesis. Development 131, 1524.
  • Cuervo, R., Valencia, C., Chandraratna, R. A. & Covarrubias, L. 2002. Programmed cell death is required for palate shelf fusion and is regulated by retinoic acid. Dev. Biol. 245, 145156.
  • Dahn, R. D. & Fallon, J. F. 2000. Interdigital regulation of digit identity and homeotic transformation by modulated BMP signaling. Science 289, 438441.
  • Davies, P. J., Murtaugh, M. P., Moore, W. T. Jr, Johnson, G. S. & Lucas, D. 1985. Retinoic acid-induced expression of tissue transglutaminase in human promyelocytic leukemia (HL-60) cells. J. Biol. Chem. 260, 51665174.
  • Dolle, P., Fraulob, V., Kastner, P. & Chambon, P. 1994. Developmental expression of murine retinoid X receptor (RXR) genes. Mech. Dev. 45, 91104.
  • Dolle, P., Ruberte, E., Kastner, P., Petkovich, M., Stoner, C. M., Gudas, L. J. & Chambon, P. 1989. Differential expression of genes encoding alpha, beta and gamma retinoic acid receptors and CRABP in the developing limbs of the mouse. Nature 342, 702705.
  • Dunker, N., Schmitt, K. & Krieglstein, K. 2002. TGF-beta is required for programmed cell death in interdigital webs of the developing mouse limb. Mech. Dev. 113, 111120.
  • Dupe, V., Ghyselinck, N. B., Thomazy, V., Nagy, L., Davies, P. J., Chambon, P. & Mark, M. 1999. Essential roles of retinoic acid signaling in interdigital apoptosis and control of BMP-7 expression in mouse autopods. Dev. Biol. 208, 3043.
  • Eblaghie, M., Sanz-Ezquerro, J., Dickinson, R., Munsterberg, A., Keyse, S. & Tickle, C. 2002. 5 Pyst1/MKP3, a novel dual-specificity phosphatase, and FGF signalling in chick limb development. J. Anat. 201, 418.
  • Fallon, J. F. & Cameron, J. 1977. Interdigital cell death during limb development of the turtle and lizard with an interpretation of evolutionary significance. J. Embryol. Exp. Morphol. 40, 285289.
  • Feijen, A., Goumans, M. J. & Van Den Eijnden-Van Raaij, A. J. 1994. Expression of activin subunits, activin receptors and follistatin in postimplantation mouse embryos suggests specific developmental functions for different activins. Development 120, 36213637.
  • Fernandez-Teran, M. A., Hinchliffe, J. R. & Ros, M. A. 2006. Birth and death of cells in limb development: a mapping study. Dev. Dyn. 235, 25212537.
  • Ferrari, D., Lichtler, A. C., Pan, Z. Z., Dealy, C. N., Upholt, W. B. & Kosher, R. A. 1998. Ectopic expression of Msx-2 in posterior limb bud mesoderm impairs limb morphogenesis while inducing BMP-4 expression, inhibiting cell proliferation, and promoting apoptosis. Dev. Biol. 197, 1224.
  • Francis, P. H., Richardson, M. K., Brickell, P. M. & Tickle, C. 1994. Bone morphogenetic proteins and a signalling pathway that controls patterning in the developing chick limb. Development 120, 209218.
  • Fujii, H., Sato, T., Kaneko, S., Gotoh, O., Fujii-Kuriyama, Y., Osawa, K., Kato, S. & Hamada, H. 1997. Metabolic inactivation of retinoic acid by a novel P450 differentially expressed in developing mouse embryos. EMBO J. 16, 41634173.
  • Ganan, Y., Macias, D., Basco, R. D., Merino, R. & Hurle, J. M. 1998. Morphological diversity of the avian foot is related with the pattern of msx gene expression in the developing autopod. Dev. Biol. 196, 3341.
  • Garcia-Martinez, V., Macias, D., Ganan, Y., Garcia-Lobo, J. M., Francia, M. V., Fernandez-Teran, M. A. & Hurle, J. M. 1993. Internucleosomal DNA fragmentation and programmed cell death (apoptosis) in the interdigital tissue of the embryonic chick leg bud. J. Cell Sci. 106, (Pt) 201208.
  • Gavrieli, Y., Sherman, Y. & Ben-Sasson, S. A. 1992. Identification of programmed cell death in situ via specific labeling of nuclear DNA fragmentation. J. Cell Biol. 119, 493501.
  • Geetha-Loganathan, P., Nimmagadda, S., Huang, R., Scaal, M. & Christ, B. 2006. Expression pattern of BMPs during chick limb development. Anat. Embryol. (Berl.) 211(Suppl 1), 8793.
  • Ghyselinck, N. B., Dupe, V., Dierich, A., Messaddeq, N., Garnier, J. M., Rochette-Egly, C., Chambon, P. & Mark, M. 1997. Role of the retinoic acid receptor beta (RARbeta) during mouse development. Int. J. Dev. Biol. 41, 425447.
  • Grotewold, L. & Ruther, U. 2002. The Wnt antagonist Dickkopf-1 is regulated by Bmp signaling and c-Jun and modulates programmed cell death. EMBO J. 21, 966975.
  • Guha, U., Gomes, W. A., Kobayashi, T., Pestell, R. G. & Kessler, J. A. 2002. In vivo evidence that BMP signaling is necessary for apoptosis in the mouse limb. Dev. Biol. 249, 108120.
  • Hamburger, V. & Hamilton, H. L. 1992. A series of normal stages in the development of the chick embryo. 1951. Dev. Dyn. 195, 231272.
  • Hernandez-Martinez, R., Castro-Obregon, S. & Covarrubias, L. 2009. Progressive interdigital cell death: regulation by the antagonistic interaction between fibroblast growth factor 8 and retinoic acid. Development 136, 36693678.
  • Hopkinson-Woolley, J., Hughes, D., Gordon, S. & Martin, P. 1994. Macrophage recruitment during limb development and wound healing in the embryonic and foetal mouse. J. Cell Sci. 107, (Pt 5) 11591167.
  • Houzelstein, D., Cohen, A., Buckingham, M. E. & Robert, B. 1997. Insertional mutation of the mouse Msx1 homeobox gene by an nlacZ reporter gene. Mech. Dev. 65, 123133.
  • Hubner, A., Cavanagh-Kyros, J., Rincon, M., Flavell, R. A. & Davis, R. J. 2010. Functional cooperation of the proapoptotic Bcl2 family proteins Bmf and Bim in vivo. Mol. Cell. Biol. 30, 98105.
  • Hui, C. C. & Joyner, A. L. 1993. A mouse model of Greig cephalopolysyndactyly syndrome: the extra-toesJ mutation contains an intragenic deletion of the Gli3 gene. Nat Genet. 3, 241246.
  • Hurle, J. M. & Fernandez-Teran, M. A. 1983. Fine structure of the regressing interdigital membranes during the formation of the digits of the chick embryo leg bud. J. Embryol. Exp. Morphol. 78, 195209.
  • Jacobsen, M. D., Weil, M. & Raff, M. C. 1996. Role of Ced-3/ICE-family proteases in staurosporine-induced programmed cell death. J. Cell Biol. 133, 10411051.
  • Jacobson, M. D., Weil, M. & Raff, M. C. 1997. Programmed cell death in animal development. Cell 88, 347354.
  • Jiang, H. & Kochhar, D. M. 1992. Induction of tissue transglutaminase and apoptosis by retinoic acid in the limb bud. Teratology 46, 333340.
  • Kastner, P., Grondona, J. M., Mark, M., Gansmuller, A., Lemeur, M., Decimo, D., Vonesch, J. L., Dolle, P. & Chambon, P. 1994. Genetic analysis of RXR alpha developmental function: convergence of RXR and RAR signaling pathways in heart and eye morphogenesis. Cell 78, 9871003.
  • Kastner, P., Mark, M., Ghyselinck, N., Krezel, W., Dupe, V., Grondona, J. M. & Chambon, P. 1997. Genetic evidence that the retinoid signal is transduced by heterodimeric RXR/RAR functional units during mouse development. Development 124, 313326.
  • Kawakami, Y., Rodriguez-Leon, J., Koth, C. M., Buscher, D., Itoh, T., Raya, A., Ng, J. K., Esteban, C. R., Takahashi, S., Henrique, D., Schwarz, M. F., Asahara, H. & Izpisua Belmonte, J. C. 2003. MKP3 mediates the cellular response to FGF8 signalling in the vertebrate limb. Nat. Cell Biol. 5, 513519.
  • Klionsky, D. J., Abeliovich, H., Agostinis, P., Agrawal, D. K., Aliev, G., Askew, D. S., Baba, M., Baehrecke, E. H., Bahr, B. A., Ballabio, A., Bamber, B. A., Bassham, D. C., Bergamini, E., Bi, X., Biard-Piechaczyk, M., Blum, J. S., Bredesen, D. E., Brodsky, J. L.. Brumell, J. H., Brunk, U. T., Bursch, W., Camougrand, N., Cebollero, E., Cecconi, F., Chen, Y., Chin, L. S., Choi, A., Chu, C. T., Chung, J., Clarke, P. G., Clark, R. S., Clarke, S. G., Clave, C., Cleveland, J. L., Codogno, P., Colombo, M. I., Coto-Montes, A., Cregg, J. M., Cuervo, A. M., Debnath, J., Demarchi, F., Dennis, P. B., Dennis, P. A., Deretic, V., Devenish, R. J., Di Sano, F., Dice, J. F., Difiglia, M., Dinesh-Kumar, S., Distelhorst, C. W., Djavaheri-Mergny, M., Dorsey, F. C., Droge, W., Dron, M., Dunn, W. A. Jr, Duszenko, M., Eissa, N. T., Elazar, Z., Esclatine, A., Eskelinen, E. L., Fesus, L., Finley, K. D., Fuentes, J. M., Fueyo, J., Fujisaki, K., Galliot, B., Gao, F. B., Gewirtz, D. A., Gibson, S. B., Gohla, A., Goldberg, A. L., Gonzalez, R., Gonzalez-Estevez, C., Gorski, S., Gottlieb, R. A., Haussinger, D., He, Y. W., Heidenreich, K., Hill, J. A., Hoyer-Hansen, M., Hu, X., Huang, W. P., Iwasaki, A., Jaattela, M., Jackson, W. T., Jiang, X., Jin, S., Johansen, T.Jung, J. U., Kadowaki, M., Kang, C., Kelekar, A., Kessel, D. H., Kiel, J. A., Kim, H. P., Kimchi, A., Kinsella, T. J., Kiselyov, K., Kitamoto, K. & Knecht, E. 2008. Guidelines for the use and interpretation of assays for monitoring autophagy in higher eukaryotes. Autophagy 4, 151175.
  • Kochhar, D. M., Jiang, H., Harnish, D. C. & Soprano, D. R. 1993. Evidence that retinoic acid-induced apoptosis in the mouse limb bud core mesenchymal cells is gene-mediated. Prog. Clin. Biol. Res. 383B, 815825.
  • Krezel, W., Dupe, V., Mark, M., Dierich, A., Kastner, P. & Chambon, P. 1996. RXR gamma null mice are apparently normal and compound RXR alpha +/−/RXR beta−/−/RXR gamma−/− mutant mice are viable. Proc. Natl. Acad. Sci. U S A 93, 90109014.
  • Kuida, K., Haydar, T. F., Kuan, C. Y., Gu, Y., Taya, C., Karasuyama, H., Su, M. S., Rakic, P. & Flavell, R. A. 1998. Reduced apoptosis and cytochrome c-mediated caspase activation in mice lacking caspase 9. Cell 94, 325337.
  • Kuida, K., Zheng, T. S., Na, S., Kuan, C., Yang, D., Karasuyama, H., Rakic, P. & Flavell, R. A. 1996. Decreased apoptosis in the brain and premature lethality in CPP32-deficient mice. Nature 384, 368372.
  • Lakhani, S. A., Masud, A., Kuida, K., Porter, G. A. Jr, Booth, C. J., Mehal, W. Z., Inayat, I. & Flavell, R. A. 2006. Caspases 3 and 7: key mediators of mitochondrial events of apoptosis. Science 311, 847851.
  • Lallemand, Y., Nicola, M. A., Ramos, C., Bach, A., Cloment, C. S. & Robert, B. 2005. Analysis of Msx1; Msx2 double mutants reveals multiple roles for Msx genes in limb development. Development 132, 30033014.
  • Lee, Z. W., Kwon, S. M., Kim, S. W., Yi, S. J., Kim, Y. M. & Ha, K. S. 2003. Activation of in situ tissue transglutaminase by intracellular reactive oxygen species. Biochem. Biophys. Res. Commun. 305, 633640.
  • Lewandoski, M., Sun, X. & Martin, G. R. 2000. Fgf8 signalling from the AER is essential for normal limb development. Nat. Genet. 26, 460463.
  • Li, C., Scott, D. A., Hatch, E., Tian, X. & Mansour, S. L. 2007. Dusp6 (Mkp3) is a negative feedback regulator of FGF-stimulated ERK signaling during mouse development. Development 134, 167176.
  • Lindsten, T., Ross, A. J., King, A., Zong, W. X., Rathmell, J. C., Shiels, H. A., Ulrich, E., Waymire, K. G., Mahar, P., Frauwirth, K., Chen, Y., Wei, M., Eng, V. M., Adelman, D. M., Simon, M. C., Ma, A., Golden, J. A., Evan, G., Korsmeyer, S. J., Macgregor, G. R. & Thompson, C. B. 2000. The combined functions of proapoptotic Bcl-2 family members bak and bax are essential for normal development of multiple tissues. Mol. Cell 6, 13891399.
  • Lu, P., Minowada, G. & Martin, G. R. 2006. Increasing Fgf4 expression in the mouse limb bud causes polysyndactyly and rescues the skeletal defects that result from loss of Fgf8 function. Development 133, 3342.
  • Lyons, K. M., Hogan, B. L. & Robertson, E. J. 1995. Colocalization of BMP 7 and BMP 2 RNAs suggests that these factors cooperatively mediate tissue interactions during murine development. Mech. Dev. 50, 7183.
  • Maatouk, D. M., Choi, K. S., Bouldin, C. M. & Harfe, B. D. 2009. In the limb AER Bmp2 and Bmp4 are required for dorsal-ventral patterning and interdigital cell death but not limb outgrowth. Dev. Biol. 327, 516523.
  • Macias, D., Ganan, Y., Ros, M. A. & Hurle, J. M. 1996. In vivo inhibition of programmed cell death by local administration of FGF-2 and FGF-4 in the interdigital areas of the embryonic chick leg bud. Anat. Embryol. (Berl.) 193, 533541.
  • Maclean, G., Abu-Abed, S., Dolle, P., Tahayato, A., Chambon, P. & Petkovich, M. 2001. Cloning of a novel retinoic-acid metabolizing cytochrome P450, Cyp26B1, and comparative expression analysis with Cyp26A1 during early murine development. Mech. Dev. 107, 195201.
  • Mariani, F. V., Ahn, C. P. & Martin, G. R. 2008. Genetic evidence that FGFs have an instructive role in limb proximal-distal patterning. Nature 453, 401405.
  • Mcculloch, D. R., Nelson, C. M., Dixon, L. J., Silver, D. L., Wylie, J. D., Lindner, V., Sasaki, T., Cooley, M. A., Argraves, W. S. & Apte, S. S. 2009. ADAMTS metalloproteases generate active versican fragments that regulate interdigital web regression. Dev. Cell 17, 687698.
  • Melendez, A. & Neufeld, T. P. 2008. The cell biology of autophagy in metazoans: a developing story. Development 135, 23472360.
  • Merino, R., Rodriguez-Leon, J., Macias, D., Ganan, Y., Economides, A. N. & Hurle, J. M. 1999. The BMP antagonist Gremlin regulates outgrowth, chondrogenesis and programmed cell death in the developing limb. Development 126, 55155522.
  • Miner, J. H., Cunningham, J. & Sanes, J. R. 1998. Roles for laminin in embryogenesis: exencephaly, syndactyly, and placentopathy in mice lacking the laminin alpha5 chain. J. Cell Biol. 143, 17131723.
  • Mizushima, N., Yoshimori, T. & Levine, B. 2010. Methods in mammalian autophagy research. Cell 140, 313326.
  • Mollard, R., Viville, S., Ward, S. J., Decimo, D., Chambon, P. & Dolle, P. 2000. Tissue-specific expression of retinoic acid receptor isoform transcripts in the mouse embryo. Mech. Dev. 94, 223232.
  • Montero, J. A., Lorda-Diez, C. I., Certal, A. C., Moreno, N., Rodriguez-Leon, J., Torriglia, A. & Hurle, J. M. 2010. Coordinated and sequential activation of neutral and acidic DNases during interdigital cell death in the embryonic limb. Apoptosis 94, 223232.
  • Motoyama, N., Wang, F., Roth, K. A., Sawa, H., Nakayama, K., Negishi, I., Senju, S., Zhang, Q. & Fujii, S. 1995. Massive cell death of immature hematopoietic cells and neurons in Bcl-x-deficient mice. Science 267, 15061510.
  • Mukhopadhyay, M., Shtrom, S., Rodriguez-Esteban, C., Chen, L., Tsukui, T., Gomer, L., Dorward, D. W., Glinka, A., Grinberg, A., Huang, S. P., Niehrs, C., Izpisua Belmonte, J. C. & Westphal, H. 2001. Dickkopf1 is required for embryonic head induction and limb morphogenesis in the mouse. Dev. Cell 1, 423434.
  • Nagasaka, A., Kawane, K., Yoshida, H. & Nagata, S. 2010. Apaf-1-independent programmed cell death in mouse development. Cell Death Differ. 17, 931941.
  • Nakanishi, K., Maruyama, M., Shibata, T. & Morishima, N. 2001. Identification of a caspase-9 substrate and detection of its cleavage in programmed cell death during mouse development. J. Biol. Chem. 276, 4123741244.
  • Niederreither, K., Mccaffery, P., Drager, U. C., Chambon, P. & Dolle, P. 1997. Restricted expression and retinoic acid-induced downregulation of the retinaldehyde dehydrogenase type 2 (RALDH-2) gene during mouse development. Mech. Dev. 62, 6778.
  • Novack, D. V. & Korsmeyer, S. J. 1994. Bcl-2 protein expression during murine development. Am. J. Pathol. 145, 6173.
  • Ota, S., Zhou, Z. Q., Keene, D. R., Knoepfler, P. & Hurlin, P. J. 2007. Activities of N-Myc in the developing limb link control of skeletal size with digit separation. Development 134, 15831592.
  • Ovchinnikov, D. A., Selever, J., Wang, Y., Chen, Y. T., Mishina, Y., Martin, J. F. & Behringer, R. R. 2006. BMP receptor type IA in limb bud mesenchyme regulates distal outgrowth and patterning. Dev. Biol. 295, 103115.
  • Pajni-Underwood, S., Wilson, C. P., Elder, C., Mishina, Y. & Lewandoski, M. 2007. BMP signals control limb bud interdigital programmed cell death by regulating FGF signaling. Development 134, 23592368.
  • Pan, Y., Liu, Z., Shen, J. & Kopan, R. 2005. Notch1 and 2 cooperate in limb ectoderm to receive an early Jagged2 signal regulating interdigital apoptosis. Dev. Biol. 286, 472482.
  • Penaloza, C., Orlanski, S., Ye, Y., Entezari-Zaher, T., Javdan, M. & Zakeri, Z. 2008. Cell death in mammalian development. Curr. Pharm. Des. 14, 184196.
  • Pennimpede, T., Cameron, D. A., Maclean, G. A. & Petkovich, M. 2010. Analysis of Cyp26b1/Rarg compound-null mice reveals two genetically separable effects of retinoic acid on limb outgrowth. Dev. Biol. 339, 179186.
  • Rae, F., Woods, K., Sasmono, T., Campanale, N., Taylor, D., Ovchinnikov, D. A., Grimmond, S. M., Hume, D. A., Ricardo, S. D. & Little, M. H. 2007. Characterisation and trophic functions of murine embryonic macrophages based upon the use of a Csf1r-EGFP transgene reporter. Dev. Biol. 308, 232246.
  • Ravichandran, K. S. & Lorenz, U. 2007. Engulfment of apoptotic cells: signals for a good meal. Nat. Rev. Immunol. 7, 964974.
  • Roberts, L. M., Hirokawa, Y., Nachtigal, M. W. & Ingraham, H. A. 1999. Paracrine-mediated apoptosis in reproductive tract development. Dev. Biol. 208, 110122.
  • Rodriguez-Leon, J., Merino, R., Macias, D., Ganan, Y., Santesteban, E. & Hurle, J. M. 1999. Retinoic acid regulates programmed cell death through BMP signalling. Nat. Cell Biol. 1, 125126.
  • Rotello, R. J., Fernandez, P. A. & Yuan, J. 1994. Anti-apogens and anti-engulfens: monoclonal antibodies reveal specific antigens on apoptotic and engulfment cells during chicken embryonic development. Development 120, 14211431.
  • Salas-Vidal, E., Lomeli, H., Castro-Obregon, S., Cuervo, R., Escalante-Alcalde, D. & Covarrubias, L. 1998. Reactive oxygen species participate in the control of mouse embryonic cell death. Exp. Cell. Res. 238, 136147.
  • Salas-Vidal, E., Valencia, C. & Covarrubias, L. 2001. Differential tissue growth and patterns of cell death in mouse limb autopod morphogenesis. Dev. Dyn. 220, 295306.
  • Schnabel, D., Salas-Vidal, E., Narvaez, V., Sanchez-Carbente Mdel, R., Hernandez-Garcia, D., Cuervo, R. & Covarrubias, L. 2006. Expression and regulation of antioxidant enzymes in the developing limb support a function of ROS in interdigital cell death. Dev. Biol. 291, 291299.
  • Sheth, R., Bastida, M. F. & Ros, M. 2007. Hoxd and Gli3 interactions modulate digit number in the amniote limb. Dev Biol. 310, 430441.
  • Sucov, H. M., Dyson, E., Gumeringer, C. L., Price, J., Chien, K. R. & Evans, R. M. 1994. RXR alpha mutant mice establish a genetic basis for vitamin A signaling in heart morphogenesis. Genes Dev. 8, 10071018.
  • Sun, X., Mariani, F. V. & Martin, G. R. 2002. Functions of FGF signalling from the apical ectodermal ridge in limb development. Nature 418, 501508.
  • Suzuki, D., Yamada, A., Amano, T., Yasuhara, R., Kimura, A., Sakahara, M., Tsumaki, N., Takeda, S., Tamura, M., Nakamura, M., Wada, N., Nohno, T., Shiroishi, T., Aiba, A. & Kamijo, R. 2009. Essential mesenchymal role of small GTPase Rac1 in interdigital programmed cell death during limb development. Dev. Biol. 335, 396406.
  • Suzuki, T., Hasso, S. M. & Fallon, J. F. 2008. Unique SMAD1/5/8 activity at the phalanx-forming region determines digit identity. Proc. Natl. Acad. Sci. U S A 105, 41854190.
  • Varfolomeev, E. E., Schuchmann, M., Luria, V., Chiannilkulchai, N., Beckmann, J. S., Mett, I. L., Rebrikov, D., Brodianski, V. M., Kemper, O. C., Kollet, O., Lapidot, T., Soffer, D., Sobe, T., Avraham, K. B., Goncharov, T., Holtmann, H., Lonai, P. & Wallach, D. 1998. Targeted disruption of the mouse Caspase 8 gene ablates cell death induction by the TNF receptors, Fas/Apo1, and DR3 and is lethal prenatally. Immunity 9, 267276.
  • Veis, D. J., Sorenson, C. M., Shutter, J. R. & Korsmeyer, S. J. 1993. Bcl-2-deficient mice demonstrate fulminant lymphoid apoptosis, polycystic kidneys, and hypopigmented hair. Cell 75, 229240.
  • Villacorte, M., Suzuki, K., Hayashi, K., De Sousa Lopes, S. C., Haraguchi, R., Taketo, M. M., Nakagata, N. & Yamada, G. 2010. Antagonistic crosstalk of Wnt/beta-catenin/Bmp signaling within the Apical Ectodermal Ridge (AER) regulates interdigit formation. Biochem. Biophys. Res. Commun. 391, 16531657.
  • Wang, C. K., Omi, M., Ferrari, D., Cheng, H. C., Lizarraga, G., Chin, H. J., Upholt, W. B., Dealy, C. N. & Kosher, R. A. 2004. Function of BMPs in the apical ectoderm of the developing mouse limb. Dev. Biol. 269, 109122.
  • Weatherbee, S. D., Behringer, R. R., Rasweiler, J. J. T. & Niswander, L. A. 2006. Interdigital webbing retention in bat wings illustrates genetic changes underlying amniote limb diversification. Proc. Natl. Acad. Sci. U S A 103, 1510315107.
  • Wilkie, A. O., Patey, S. J., Kan, S. H., Van Den Ouweland, A. M. & Hamel, B. C. 2002. FGFs, their receptors, and human limb malformations: clinical and molecular correlations. Am. J. Med. Genet. 112, 266278.
  • Wood, W., Turmaine, M., Weber, R., Camp, V., Maki, R. A., Mckercher, S. R. & Martin, P. 2000. Mesenchymal cells engulf and clear apoptotic footplate cells in macrophageless PU.1 null mouse embryos. Development 127, 52455252.
  • Yashiro, K., Zhao, X., Uehara, M., Yamashita, K., Nishijima, M., Nishino, J., Saijoh, Y., Sakai, Y. & Hamada, H. 2004. Regulation of retinoic acid distribution is required for proximodistal patterning and outgrowth of the developing mouse limb. Dev. Cell 6, 411422.
  • Yi, S. E., Daluiski, A., Pederson, R., Rosen, V. & Lyons, K. M. 2000. The type I BMP receptor BMPRIB is required for chondrogenesis in the mouse limb. Development 127, 621630.
  • Yin, L., Du, X., Li, C., Xu, X., Chen, Z., Su, N., Zhao, L., Qi, H., Li, F., Xue, J., Yang, J., Jin, M., Deng, C. & Chen, L. 2008. A Pro253Arg mutation in fibroblast growth factor receptor 2 (Fgfr2) causes skeleton malformation mimicking human Apert syndrome by affecting both chondrogenesis and osteogenesis. Bone 42, 631643.
  • Yokouchi, Y., Ohsugi, K., Sasaki, H. & Kuroiwa, A. 1991. Chicken homeobox gene Msx-1: structure, expression in limb buds and effect of retinoic acid. Development 113, 431444.
  • Yokouchi, Y., Sakiyama, J., Kameda, T., Iba, H., Suzuki, A., Ueno, N. & Kuroiwa, A. 1996. BMP-2/-4 mediate programmed cell death in chicken limb buds. Development 122, 37253734.
  • Yu, K. & Ornitz, D. M. 2008. FGF signaling regulates mesenchymal differentiation and skeletal patterning along the limb bud proximodistal axis. Development 135, 483491.
  • Zakeri, Z., Quaglino, D. & Ahuja, H. S. 1994. Apoptotic cell death in the mouse limb and its suppression in the hammertoe mutant. Dev. Biol. 165, 294297.
  • Zhang, D., Ferguson, C. M., O’keefe, R. J., Puzas, J. E., Rosier, R. N. & Reynolds, P. R. 2002. A role for the BMP antagonist chordin in endochondral ossification. J. Bone Miner. Res. 17, 293300.
  • Zhao, X., Brade, T., Cunningham, T. J. & Duester, G. 2010. Retinoic acid controls expression of tissue remodeling genes Hmgn1 and Fgf18 at the digit-interdigit junction. Dev. Dyn. 239, 665671.
  • Zheng, T. S., Hunot, S., Kuida, K. & Flavell, R. A. 1999. Caspase knockouts: matters of life and death. Cell Death Differ. 6, 10431053.
  • Zou, H., Wieser, R., Massague, J. & Niswander, L. 1997. Distinct roles of type I bone morphogenetic protein receptors in the formation and differentiation of cartilage. Genes Dev. 11, 21912203.
  • Zuzarte-Luis, V., Berciano, M. T., Lafarga, M. & Hurle, J. M. 2006. Caspase redundancy and release of mitochondrial apoptotic factors characterize interdigital apoptosis. Apoptosis 11, 701715.
  • Zuzarte-Luis, V., Montero, J. A., Kawakami, Y., Izpisua-Belmonte, J. C. & Hurle, J. M. 2007. Lysosomal cathepsins in embryonic programmed cell death. Dev. Biol. 301, 205217.
  • Zuzarte-Luis, V., Montero, J. A., Rodriguez-Leon, J., Merino, R., Rodriguez-Rey, J. C. & Hurle, J. M. 2004. A new role for BMP5 during limb development acting through the synergic activation of Smad and MAPK pathways. Dev. Biol. 272, 3952.