- Email: [email protected]
Limb development: Farewell to arms
R368
Dispatch
Limb development: Farewell to arms Anthony Graham and Imelda McGonnell
Forelimbs and hindlimbs are, clearly, quite different, and it has long been appreciated that their differences are assigned early in development; the genetic basis of these differences has been more mysterious, however. Recent work has now shown that the homeobox gene Pitx1 imparts identity to the developing hindlimb bud. Address: Molecular Neurobiology Group, 4th floor Hodgkin Building, Kings College London, Guys Campus, London SE1 9RT, UK. Current Biology 1999, 9:R368–R370 http://biomednet.com/elecref/09609822009R0368 © Elsevier Science Ltd ISSN 0960-9822
Limbs have long been used as model systems for unravelling the fundamental concerns of developmental biology and, relatively speaking, we know a lot about how these structures develop. Yet, while we know much about how limbs per se develop, we know very little about how limb identity is assigned — what makes a forelimb distinct from a hindlimb? This is clearly an important question, and one that is now being answered through recent studies on the homeobox gene Pitx1 which demonstrate that this gene plays a critical role in specifying hindlimb identity. Both the forelimb and the hindlimb develop from buds that emerge in the lateral region of the trunk. The buds subsequently grow out from the body and, as they do so, the various components of the limbs are laid down in succession. The proximal long bones of the upper limbs are laid down first, while the phalanges of the digits appear last. Furthermore, we now also have a fairly detailed picture of the molecular basis of limb development. For example, it is apparent that the initiation of limb bud outgrowth involves members of the fibroblast growth factor (FGF) family of signalling molecules, while anterior–posterior patterning within the limb involves the signalling molecule Sonic hedgehog. Morphologically, the differences between the forelimb and hindlimb are very clear (see Figure 1). Many of the bones of the hindlimb are longer and heavier than their forelimb counterparts. The articulations and types of joint also differ, and one of the most striking differences between the forelimb and the hindlimb lies in the fact that, while the hand flexes posteriorly at the wrist, the foot extends directly from the ankle. There are also elements that are unique to each limb, such as the patella and calcaneous (heel bone) of the hindlimb, and the pisiform element in the forelimb. Besides these skeletal differences, the muscle and tendon patterns in each limb are also distinct, and in avians there are ectodermal differences, in that the
forelimb forms feathers whereas the distal hindlimb is covered with scales. The genetic basis of the differences between the forelimb and hindlimb has been less clear, however. For a gene to control the hindlimb-specific or forelimb-specific pattern, it must not only be expressed in one limb type, but it must also be expressed from early stages throughout the limb field. Some genes, such as HoxC10 and HoxC11, are expressed differentially in limbs, but at late stages and in restricted patterns [1] (Figure 1). More recently, genes that may be involved in expressing limb identity have been described. The first such genes to be identified, Tbx4 and Tbx5, encode members of the T-box family of transcription factors; later work identified Pitx1 (or Ptx1), which encodes a bicoid-like transcription factor [2–6] (Figure 1). Tbx5 is expressed in the forelimb bud, whereas Tbx4 and Pitx1 are expressed in the early hindlimb; importantly, embryonic manipulations in avian embryos have also demonstrated a tight correlation between the expression of these genes and limb identity. Classical experiments have shown that limb identity is specified early in development, and that this is independent of the limb ectoderm. Thus, when hindlimb mesenchyme is grafted into the forelimb it still develops as hindlimb. Consistent with the view that Tbx4, Tbx5 and Pitx1 play a role in the elucidation of limb identity, it has been shown that when forelimb mesenchyme is grafted into the hindlimb it maintains Tbx5 expression, and that when hindlimb mesenchyme is grafted into the forelimb the expression of both Tbx4 and Pitx1 persists [2–5]. Of these genes, Pitx1 seems to be particularly important, as it is expressed in the hindlimb primordium before Tbx4 is expressed [5]. Moreover, three recent papers have provided strong genetic evidence that Pitx1 plays a role in the elaboration of hindlimb identity [7–9]. Two groups have analysed the function of Pitx1 by using retroviral vectors to express the gene in the developing forelimb bud [8,9], and in both cases the consequence of this misexpression was startling. The infected forelimbs no longer displayed the expected posterior flexure of the ‘autopod’ (hand and digits) at the wrist with respect to the ‘zeugopod’ (lower arm). Rather, in these limbs the autopod remained in the same anteroposterior plane with respect to the zeugopod, a situation that one normally associates with the hindlimb. Expression of Pitx1 in the developing forelimb also had consequences for the skeletal elements. In the normal
Dispatch
R369
Figure 1 Schematic representation of the developing limb buds, indicating expression of limbspecific genes (left) and the skeletal elements of the forelimb and hindlimb (right).
Humerus Radius Tbx5
II
Forelimb Ulna
IV III Phalanges
Tibia Hindlimb
Femur Fibula
Pitx1 Tbx4 HoxC10/11
I
II III IV
Phalanges Current Biology
avian wing there are three digits, II, III and IV, along the anteroposterior axis. Of these, digit III is the longest, whereas digit II is characteristically short. In the leg, by contrast, there are four digits, but leg digits II, III and IV are of similar size, whereas the anterior-most digit I is much smaller. Interestingly, in Pitx1-expressing forelimbs, digits II, III and IV are of similar length, and in some cases infected wings also produced an ectopic small anterior digit, similar to digit I of the leg [8,9].
shorter. The femur and the tibia, in particular, also become narrower, and in these mutants the tibia and fibula become similarly sized, resembling the relationship between the radius and ulna of the forelimb. The reduced diameter of these bones is due to impaired ossification and calcification, and there is no articular cartilage proximally. These animals also display alterations in the joints of the hindlimb. The femur, which normally articulates only with the tibia, now also articulates with the fibula, a situation similar to that seen in the articulation between humerus, radius and ulna at the elbow.
Another indication that Pitx1 can promote a hindlimb developmental programme has come from the analysis of the muscles and tendon patterns in the Pitx1-expressing limbs. The leg has a different number of muscles from the wing, both dorsally and ventrally, and each has characteristic shape and points of origin and insertion. Pitx1-expressing wings were found to have muscle and tendon patterns characteristic of a leg, particularly in dorsal regions [8]. Indeed, this may in part explain the changed flexure and orientation of the wing autopod mentioned above. It should, however, also be noted that Pitx1 overexpression in the forelimb did not cause a complete transformation to a hindlimb phenotype. Although the shape of the phalanges of the infected forelimb resembled those of a hindlimb, the number of phalanges was what one would expect of forelimb.
The Pitx1 mutant mice also display alterations in limbspecific elements. The calcaneous — the hindlimb-specific heelbone — is much reduced and altered in shape. The patella — another hindlimb-specific element — is missing, as are the zucker nodes. Szeto et al. [9] also found an ectopic element on the proximal tarsus resembling a pisiform-like element, which is normally a forelimbspecific condensation. Another point of note was that most of the structures affected were dorsally derived, suggesting that Pitx1 is more concerned with patterning this region of the limb bud. Furthermore, as with the avian studies, analysis of the digits in the Pitx1 mutant mice also revealed that there was not a complete transformation of hindlimb to forelimb.
Complementing these avian studies, two groups have also looked at the effects of inactivating the Pitx1 gene in knockout mice [7,9]. The result of these studies also indicate that Pitx1 has an important role in imparting hindlimb identity. In the Pitx1 mutant mice, the pelvic girdle is smaller and the long bones — femur, tibia and fibula — significantly
Importantly, these studies on the function of Pitx1 have given us some insight into the genetic pathway controlling hindlimb identity. In the chick experiments, it was found that overexpression of Pitx1 in the forelimb resulted in the stimulation of Tbx4 expression [8,9], which is normally only expressed in the hindlimb at
R370
Current Biology, Vol 9 No 10
these stages. Pitx1 expression in the forelimb also resulted in the expression of other hindlimb-specific genes, such as HoxC10 and HoxC11 [8]. Driving expression of Pitx1 in the forelimb did not, however, suppress the forelimb-specific gene Tbx5 [8,9]. This is an important result suggesting, firstly that Tbx5 is under independent control, and secondly that the incomplete forelimb-to-hindlimb transformation may be due to persistent Tbx5 expression in the infected limb. Supporting these results, the Pitx1 mutant mice were found to show a severe reduction, but not complete absence, of Tbx4 expression in the hindlimb [7,8]. Again, this demonstrates a link between Pitx1 and Tbx4 expression, and also provides an explanation for the partial transformation observed in these animals. The new results discussed here are important because, for the first time, they have uncovered the genetic basis of how identity is imparted to a developing limb. Pitx1 acts to promote a hindlimb programme of development and, in part, does so by promoting Tbx4 expression. Clearly, the expression of Pitx1 itself must be controlled by earlier, more globally acting genes, and the prime candidates for fulfilling such a role are the Hox genes. Previous work has shown that genes of the Hox9 group are expressed in the lateral plate mesoderm before and during limb outgrowth, and importantly that specific complements of these genes pick out forelimb, flank and hindlimb territories [10]. Consequently, it will be of great interest to determine how the link between the axial patterning role of the Hox genes feeds in to Pitx1 expression, and thus to the specification of limb identity. Acknowledgements We would like to thank Cathie McMahon for critical input.
References 1. Nelson CE, Morgan BA, Burke AC, Laufer E, DiMambro E, Murtaugh LC, Gonzales E, Tessarollo L, Parada LF, Tabin C: Analysis of Hox gene expression in the chick limb bud. Development 1996, 122:1449-1466. 2. Ohuchi H, Takeuchi J, Yoshoika H, Ishimaru Y, Ogura K, Takahasi N, Ogura T, Noji S: Correlation of wing-leg identity in ectopic FGF-induced chimeric limbs with the differential expression of chick Tbx5 and Tbx4. Development 1998, 125:51-60. 3. Issac A, Rodriguez-Esteban C, Ryan A, Altabef M, Tsukui T, Patel K, Tickle C, Izpisua-Belmonte JC: Tbx genes and limb identity in chick embryo development. Development 1998, 125:1867-1875. 4. Gibson-Brown JJ, Agulnik SI, Silver LM, Niswander L, Papaioannou VE: Involvement in T-box genes Tbx2–Tbx5 in vertebrate limb specification and development. Development 1998, 125:2499-2509. 5. Logan M, Simon H, Tabin C: Differential regulation of T-box and homeobox transcription factors suggest roles in controlling chick limb-type identity. Development 1988, 125:2825-2835. 6. Lanctot C, Lamelot B, Drouin J: The bicoid-related homeoprotein Ptx-1 defines the most anterior domain of the embryo and differentiates posterior from anterior lateral mesoderm. Development 1997, 124:2807-2817. 7. Lanctot C, Moreau A, Chamberland M, Tremblay ML, Drouin J: Hindlimb patterning and mandible development require the Ptx1 gene. Development 1999, 126:1805-1810. 8. Logan M, Tabin CJ: Role of Pitx1 upstream of Tbx4 in specification of hindlimb Identity. Science 1999, 283:1736-1739.
9. Szeto DP, Rodriguez-Esteban C, Ryan AK, O'Connell SM, Liu F, Kioussi C, Gleiberman AS, Izpisua-Belmonte JC, Rosenfeld MG: Role of the bicoid-related homeodomain factor Pitx1 in specifying hindlimb morphogenesis and pituitary development. Genes Dev 1999, 13:484-494. 10. Cohn MJ, Patel K, Krumlauf R, Wilkinson DG, Clark JDW, Tickle C: Hox9 genes and vertebrate limb specification. Nature 1997, 387:97-101.
Dispatch
Limb development: Farewell to arms Anthony Graham and Imelda McGonnell
Forelimbs and hindlimbs are, clearly, quite different, and it has long been appreciated that their differences are assigned early in development; the genetic basis of these differences has been more mysterious, however. Recent work has now shown that the homeobox gene Pitx1 imparts identity to the developing hindlimb bud. Address: Molecular Neurobiology Group, 4th floor Hodgkin Building, Kings College London, Guys Campus, London SE1 9RT, UK. Current Biology 1999, 9:R368–R370 http://biomednet.com/elecref/09609822009R0368 © Elsevier Science Ltd ISSN 0960-9822
Limbs have long been used as model systems for unravelling the fundamental concerns of developmental biology and, relatively speaking, we know a lot about how these structures develop. Yet, while we know much about how limbs per se develop, we know very little about how limb identity is assigned — what makes a forelimb distinct from a hindlimb? This is clearly an important question, and one that is now being answered through recent studies on the homeobox gene Pitx1 which demonstrate that this gene plays a critical role in specifying hindlimb identity. Both the forelimb and the hindlimb develop from buds that emerge in the lateral region of the trunk. The buds subsequently grow out from the body and, as they do so, the various components of the limbs are laid down in succession. The proximal long bones of the upper limbs are laid down first, while the phalanges of the digits appear last. Furthermore, we now also have a fairly detailed picture of the molecular basis of limb development. For example, it is apparent that the initiation of limb bud outgrowth involves members of the fibroblast growth factor (FGF) family of signalling molecules, while anterior–posterior patterning within the limb involves the signalling molecule Sonic hedgehog. Morphologically, the differences between the forelimb and hindlimb are very clear (see Figure 1). Many of the bones of the hindlimb are longer and heavier than their forelimb counterparts. The articulations and types of joint also differ, and one of the most striking differences between the forelimb and the hindlimb lies in the fact that, while the hand flexes posteriorly at the wrist, the foot extends directly from the ankle. There are also elements that are unique to each limb, such as the patella and calcaneous (heel bone) of the hindlimb, and the pisiform element in the forelimb. Besides these skeletal differences, the muscle and tendon patterns in each limb are also distinct, and in avians there are ectodermal differences, in that the
forelimb forms feathers whereas the distal hindlimb is covered with scales. The genetic basis of the differences between the forelimb and hindlimb has been less clear, however. For a gene to control the hindlimb-specific or forelimb-specific pattern, it must not only be expressed in one limb type, but it must also be expressed from early stages throughout the limb field. Some genes, such as HoxC10 and HoxC11, are expressed differentially in limbs, but at late stages and in restricted patterns [1] (Figure 1). More recently, genes that may be involved in expressing limb identity have been described. The first such genes to be identified, Tbx4 and Tbx5, encode members of the T-box family of transcription factors; later work identified Pitx1 (or Ptx1), which encodes a bicoid-like transcription factor [2–6] (Figure 1). Tbx5 is expressed in the forelimb bud, whereas Tbx4 and Pitx1 are expressed in the early hindlimb; importantly, embryonic manipulations in avian embryos have also demonstrated a tight correlation between the expression of these genes and limb identity. Classical experiments have shown that limb identity is specified early in development, and that this is independent of the limb ectoderm. Thus, when hindlimb mesenchyme is grafted into the forelimb it still develops as hindlimb. Consistent with the view that Tbx4, Tbx5 and Pitx1 play a role in the elucidation of limb identity, it has been shown that when forelimb mesenchyme is grafted into the hindlimb it maintains Tbx5 expression, and that when hindlimb mesenchyme is grafted into the forelimb the expression of both Tbx4 and Pitx1 persists [2–5]. Of these genes, Pitx1 seems to be particularly important, as it is expressed in the hindlimb primordium before Tbx4 is expressed [5]. Moreover, three recent papers have provided strong genetic evidence that Pitx1 plays a role in the elaboration of hindlimb identity [7–9]. Two groups have analysed the function of Pitx1 by using retroviral vectors to express the gene in the developing forelimb bud [8,9], and in both cases the consequence of this misexpression was startling. The infected forelimbs no longer displayed the expected posterior flexure of the ‘autopod’ (hand and digits) at the wrist with respect to the ‘zeugopod’ (lower arm). Rather, in these limbs the autopod remained in the same anteroposterior plane with respect to the zeugopod, a situation that one normally associates with the hindlimb. Expression of Pitx1 in the developing forelimb also had consequences for the skeletal elements. In the normal
Dispatch
R369
Figure 1 Schematic representation of the developing limb buds, indicating expression of limbspecific genes (left) and the skeletal elements of the forelimb and hindlimb (right).
Humerus Radius Tbx5
II
Forelimb Ulna
IV III Phalanges
Tibia Hindlimb
Femur Fibula
Pitx1 Tbx4 HoxC10/11
I
II III IV
Phalanges Current Biology
avian wing there are three digits, II, III and IV, along the anteroposterior axis. Of these, digit III is the longest, whereas digit II is characteristically short. In the leg, by contrast, there are four digits, but leg digits II, III and IV are of similar size, whereas the anterior-most digit I is much smaller. Interestingly, in Pitx1-expressing forelimbs, digits II, III and IV are of similar length, and in some cases infected wings also produced an ectopic small anterior digit, similar to digit I of the leg [8,9].
shorter. The femur and the tibia, in particular, also become narrower, and in these mutants the tibia and fibula become similarly sized, resembling the relationship between the radius and ulna of the forelimb. The reduced diameter of these bones is due to impaired ossification and calcification, and there is no articular cartilage proximally. These animals also display alterations in the joints of the hindlimb. The femur, which normally articulates only with the tibia, now also articulates with the fibula, a situation similar to that seen in the articulation between humerus, radius and ulna at the elbow.
Another indication that Pitx1 can promote a hindlimb developmental programme has come from the analysis of the muscles and tendon patterns in the Pitx1-expressing limbs. The leg has a different number of muscles from the wing, both dorsally and ventrally, and each has characteristic shape and points of origin and insertion. Pitx1-expressing wings were found to have muscle and tendon patterns characteristic of a leg, particularly in dorsal regions [8]. Indeed, this may in part explain the changed flexure and orientation of the wing autopod mentioned above. It should, however, also be noted that Pitx1 overexpression in the forelimb did not cause a complete transformation to a hindlimb phenotype. Although the shape of the phalanges of the infected forelimb resembled those of a hindlimb, the number of phalanges was what one would expect of forelimb.
The Pitx1 mutant mice also display alterations in limbspecific elements. The calcaneous — the hindlimb-specific heelbone — is much reduced and altered in shape. The patella — another hindlimb-specific element — is missing, as are the zucker nodes. Szeto et al. [9] also found an ectopic element on the proximal tarsus resembling a pisiform-like element, which is normally a forelimbspecific condensation. Another point of note was that most of the structures affected were dorsally derived, suggesting that Pitx1 is more concerned with patterning this region of the limb bud. Furthermore, as with the avian studies, analysis of the digits in the Pitx1 mutant mice also revealed that there was not a complete transformation of hindlimb to forelimb.
Complementing these avian studies, two groups have also looked at the effects of inactivating the Pitx1 gene in knockout mice [7,9]. The result of these studies also indicate that Pitx1 has an important role in imparting hindlimb identity. In the Pitx1 mutant mice, the pelvic girdle is smaller and the long bones — femur, tibia and fibula — significantly
Importantly, these studies on the function of Pitx1 have given us some insight into the genetic pathway controlling hindlimb identity. In the chick experiments, it was found that overexpression of Pitx1 in the forelimb resulted in the stimulation of Tbx4 expression [8,9], which is normally only expressed in the hindlimb at
R370
Current Biology, Vol 9 No 10
these stages. Pitx1 expression in the forelimb also resulted in the expression of other hindlimb-specific genes, such as HoxC10 and HoxC11 [8]. Driving expression of Pitx1 in the forelimb did not, however, suppress the forelimb-specific gene Tbx5 [8,9]. This is an important result suggesting, firstly that Tbx5 is under independent control, and secondly that the incomplete forelimb-to-hindlimb transformation may be due to persistent Tbx5 expression in the infected limb. Supporting these results, the Pitx1 mutant mice were found to show a severe reduction, but not complete absence, of Tbx4 expression in the hindlimb [7,8]. Again, this demonstrates a link between Pitx1 and Tbx4 expression, and also provides an explanation for the partial transformation observed in these animals. The new results discussed here are important because, for the first time, they have uncovered the genetic basis of how identity is imparted to a developing limb. Pitx1 acts to promote a hindlimb programme of development and, in part, does so by promoting Tbx4 expression. Clearly, the expression of Pitx1 itself must be controlled by earlier, more globally acting genes, and the prime candidates for fulfilling such a role are the Hox genes. Previous work has shown that genes of the Hox9 group are expressed in the lateral plate mesoderm before and during limb outgrowth, and importantly that specific complements of these genes pick out forelimb, flank and hindlimb territories [10]. Consequently, it will be of great interest to determine how the link between the axial patterning role of the Hox genes feeds in to Pitx1 expression, and thus to the specification of limb identity. Acknowledgements We would like to thank Cathie McMahon for critical input.
References 1. Nelson CE, Morgan BA, Burke AC, Laufer E, DiMambro E, Murtaugh LC, Gonzales E, Tessarollo L, Parada LF, Tabin C: Analysis of Hox gene expression in the chick limb bud. Development 1996, 122:1449-1466. 2. Ohuchi H, Takeuchi J, Yoshoika H, Ishimaru Y, Ogura K, Takahasi N, Ogura T, Noji S: Correlation of wing-leg identity in ectopic FGF-induced chimeric limbs with the differential expression of chick Tbx5 and Tbx4. Development 1998, 125:51-60. 3. Issac A, Rodriguez-Esteban C, Ryan A, Altabef M, Tsukui T, Patel K, Tickle C, Izpisua-Belmonte JC: Tbx genes and limb identity in chick embryo development. Development 1998, 125:1867-1875. 4. Gibson-Brown JJ, Agulnik SI, Silver LM, Niswander L, Papaioannou VE: Involvement in T-box genes Tbx2–Tbx5 in vertebrate limb specification and development. Development 1998, 125:2499-2509. 5. Logan M, Simon H, Tabin C: Differential regulation of T-box and homeobox transcription factors suggest roles in controlling chick limb-type identity. Development 1988, 125:2825-2835. 6. Lanctot C, Lamelot B, Drouin J: The bicoid-related homeoprotein Ptx-1 defines the most anterior domain of the embryo and differentiates posterior from anterior lateral mesoderm. Development 1997, 124:2807-2817. 7. Lanctot C, Moreau A, Chamberland M, Tremblay ML, Drouin J: Hindlimb patterning and mandible development require the Ptx1 gene. Development 1999, 126:1805-1810. 8. Logan M, Tabin CJ: Role of Pitx1 upstream of Tbx4 in specification of hindlimb Identity. Science 1999, 283:1736-1739.
9. Szeto DP, Rodriguez-Esteban C, Ryan AK, O'Connell SM, Liu F, Kioussi C, Gleiberman AS, Izpisua-Belmonte JC, Rosenfeld MG: Role of the bicoid-related homeodomain factor Pitx1 in specifying hindlimb morphogenesis and pituitary development. Genes Dev 1999, 13:484-494. 10. Cohn MJ, Patel K, Krumlauf R, Wilkinson DG, Clark JDW, Tickle C: Hox9 genes and vertebrate limb specification. Nature 1997, 387:97-101.