SEARCH

SEARCH BY CITATION

Keywords:

  • development;
  • limb;
  • marsupial;
  • AER;
  • opossum

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. LITERATURE CITED

Marsupials give birth after short gestation times to neonates that have an intriguing combination of precocial and altricial features, based on their functional necessity after birth. Perhaps most noticeably, marsupial newborns have highly developed forelimbs, which provide the propulsion necessary for the newborn's crawl to the teat. To achieve their advanced state at birth, the development of marsupial forelimbs is accelerated. The development of the newborn's hind limb, which plays no part in the crawl, is not accelerated, and is likely even delayed. Given the large differences in the rate of limb outgrowth among marsupials and placentals, we hypothesize that the pathways underlying the early development and outgrowth of marsupial limbs, especially that of their forelimbs, will also be divergent. As a first step toward testing this, we examine the development of one of the two major signaling centers of the developing limb, the apical ectodermal ridge (AER), in a marsupial, Monodelphis domestica. We found that, while both opossum limbs have reduced physical AER's, in the opossum forelimb this reduction has been taken to the extreme. Where the M. domestica forelimb should have an AER, it instead has only a few patches of disorganized cells. These results make the marsupial, M. domestica, the only known amniote (without reduced limbs) to exhibit no morphological AER. However, both M. domestica limbs normally express Fgf8, a molecular marker of the AER. Anat Rec 293:1325–1332, 2010. © 2010 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. LITERATURE CITED

When marsupials are born, they are at a stage of development that appears premature to that of newborn placentals. For example, the overall state of development of a marsupial newborn resembles an 11– or 12-day-old embryonic mouse or a 10- to 12-week-old embryonic human (Hughes and Hall, 1984; Vaglia and Smith, 2003). Despite their generally altricial state, marsupial newborns immediately complete a life or death crawl from the birth canal to the teat where they attach and complete much of their development (Sharman, 1973; Gemmell et al., 2002). To make this crawl, the newborns pull their bodies along using their large, precocially developed forelimbs (Fig. 1). The hind limb, which is still at a very rudimentary stage of development, is not involved in this crawl. The maturation rate of marsupial forelimbs is accelerated, and that of the hind limb delayed, to create these divergent limb morphologies at birth (Bininda-Emonds et al., 2003; Bininda-Emonds et al., 2007; Harrison and Larsson, 2008; Weisbecker et al., 2008; Sears, 2009; Werneburg and Sánchez-Villagra, in press).

thumbnail image

Figure 1. Scanning electron microscope (SEM) images of M. musculus (Mus; mouse) and M. domestica (Md; opossum) limbs. Both the mouse forelimb (E) and the opossum hind limb (F) display a well-developed, protruding apical ectodermal ridge running along the DV boundary of the developing limb bud at limb Stage (L) 5. In contrast, the opossum forelimb has scattered clumpings of cells along the DV boundary, but no protruding ridge at any examined limb stage (limb Stage 4–A and close-up in G, limb Stage 5–B and close-up in H, limb Stage 6–C, limb Stage 7–D). D = dorsal; V = ventral. Arrows indicate the DV boundary of the limb, and the AER (if present). Scale-bars represent 50 microns.

Download figure to PowerPoint

The acceleration of opossum forelimb development begs the question–are changes in early limb development responsible for this phenomenon? To answer this question, we have to explore the mechanisms controlling limb development. There are two major signaling centers responsible for limb outgrowth and patterning in amniotes–the zone of polarizing activity (ZPA) and the apical ectodermal ridge (AER). The ZPA is found in the posterior mesenchyme of the developing limb bud and provides important molecular signals for positioning the anteroposterior axis of the limb. In contrast, the AER is a visible ridge at the distal tip of the limb bud that protrudes along the dorsoventral (DV) boundary (Capdevila and Belmonte, 2001). The physical manifestation of the AER forms through a thickening of the epithelium along the tip of the limb (Niswander, 2003). The AER has several important functions during limb development, and plays a large role in both distal outgrowth of the limb and DV patterning (for reviews see Niswander, 2003; Fernandez-Teran and Ros, 2008). Surgical removal of the AER from the developing limb bud results in progressive truncation of the limb and loss of distal elements, depending on when the AER is removed (Saunders, 1948; Summerbell, 1974). Furthermore, tetrapods with reduced limbs (e.g., whales and snakes) are thought to have achieved their specialized limb morphology through, at least in part, reductions in their physical AER's (Cohn and Tickle, 1999; Thewissen et al., 2006). However, although the molecules produced by the AER are essential to limb outgrowth (Niswander et al., 1993; Fallon et al., 1994), experimental manipulation suggests that its physical manifestation is not (Errick and Saunders, 1974, 1976). Therefore, the actual role of the physical AER in tetrapod limb development remains unknown.

Because of the integral role of the AER in limb outgrowth, we hypothesize that, over the course of evolution, development of the marsupial AER was altered to help achieve or in response to the rapid outgrowth of their forelimbs. To test this hypothesis, we compare the morphological (at the gross and cellular level) and molecular development of the AER in the fore- and hind limbs of the marsupial M. domestica. Among marsupials, we have targeted M. domestica because of its ease of use as a laboratory mammal (Keyte and Smith, 2008), and its placement among the most basal of living marsupials (Horovitz et al., 2009). In terms of the molecular development of the AER, fibroblast growth factors (FGF's) are the key molecular factors required for AER function. This was conclusively established through physical manipulations in which Fgf soaked beads were found to have the capability of rescuing proximo-distal development of the limb after AER removal (Niswander et al., 1993; Fallon et al., 1994), and genetic manipulations, in which it was found that in the complete absence of Fgf8 and Fgf4 limbs fail to form (Sun et al., 2002). Four Fgf's are expressed in the AER: Fgf8 which is expressed throughout the AER, and Fgf4, Fgf9, and Fgf17 which are restricted to the posterior and distal AER (Niswander, 2003). Of these, the expression of Fgf8 is considered a synonymous marker for the AER, given that its temporal and spatial expression matches the AER's entire duration in mouse and chick (Fernandez-Teran and Ros, 2008). Therefore, because of its importance in AER function, we focus our molecular research on the levels and patterns of expression of Fgf8 in the developing marsupial AER. We also compare AER development in marsupial limbs to that in a placental mammal, the mouse (Mus musculus).

By taking this approach, this study has the potential to provide insights into not only the evolutionary changes that led to the highly derived development of marsupial limbs, but also the function of the morphological AER, a prominent outstanding question in the field of limb developmental biology.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. LITERATURE CITED

Specimens

M. domestica (opossum) and Mus musculus (mouse) embryos were obtained from in-house colonies maintained by the Sears laboratory. To obtain timed mating, male and female opossums were placed together and filmed using an infrared camera from 7 to 11 pm daily, and the time of mating noted. Mice were checked for vaginal plugs every morning, and if plugs were discovered, mating was assumed to have taken place at midnight the night before.

Opossum embryos were staged according to criteria originally developed by McCrady (1938) for Didelphis and modified by Mate et al. (1994) and Kathleen Smith (Duke University, personal communication). For the opossum forelimb, multiple embryos from Stages 28, 29, and 30 were examined, and for the opossum hind limb, multiple embryos from Stages 31, 32, and 33 were examined. Limbs from these embryos were staged according to the guide established for mice by Wanek et al. (1989)–hereafter called “limb stages.” For the forelimb, Stage 28 encompassed limb Stages 3 and 4, Stage 29 encompassed limb Stages 5 and 6, and Stage 30 encompassed limb Stages 7 and 8. For the hind limb, Stage 31 encompassed limb Stage 4, Stage 32 encompassed limb Stages 5 and 6, and Stage 33 encompassed limb Stages 7 and 8 (Table 1). As described by Wanek et al. (1989), limb Stages 4 through 7 captures the transition from a bud-like to paddle-like limb. These stages also encompass the entire duration of a prominent, physical AER in mice. In mice, the future AER begins to thicken in limb Stage 3, and is prominent during limb Stages 4 through 6. During limb Stage 7, the prominence of the AER begins to reduce (Table 1; Wanek et al., 1989; Fernandez-Teran and Ros, 2008). For comparative purposes, mice from embryonic days 10.5 to 12 post-fertilization were also examined. Limb stages for these mouse embryos were also determined by comparison with Wanek et al. (1989) (Table 1).

Table 1. AER development in opossum and mouse
Limb StageMd FLMd FL AER morphologyMd HLMd HL AER morphologyMouse FLMouse FL AER morphologyMouse HLMouse HL AER morphology
Stage 4Late Stage 28Isolated cell clumps at DV boundaryStage 31Nearly continuous cell clumps at DV boundaryE11AER prominentE12AER prominent
Stage 5Early Stage 29Isolated cell clumps at DV boundaryEarly Stage 32AER prominentE12AER prominentE12AER prominent
Stage 6Late Stage 29Isolated cell clumps at DV boundaryLate Stage 32AER prominentE12AER prominentE12AER prominent
Stage 7Early Stage 30Fewer isolated cell clumps at the DV boundaryEarly Stage 33AER reducingE12AER reducingE13AER reducing

Histology and Electron Microscopy

For histological analysis, embryos were fixed in 4% paraformaldehyde, then rinsed stepwise into 100% methanol and stored at −20°C until used. For histological preparation, embryos were rinsed stepwise into 100% PBS, then left in 30% sucrose in 4°C overnight. Embryos were then soaked in OCT compound at room temperature for 1 h. Limbs were then flash-frozen in OCT compound after being aligned for sectioning in the appropriate plane (in this case sagittal). Sagittal histological sections were taken at 10 um using a cryosectioner. Samples were stained with hematoxylin and eosin to visualize cellular morphology (Nagy et al., 2002).

For scanning electron microscopy (SEM), embryos were fixed in 2% paraformaldehyde and 2.5% gluteraldehyde, then stained with 1% osmium tetroxide before being brought stepwise into 100% ethanol for storage at −20°C until used (Nagy et al., 2002). To prepare for SEM, embryos were dried in a critical point drier, then sputter coated with gold palladium. Embryo morphology was then visualized using an Environmental Scanning Electron Microscope (ESEM; Phillips Xl30 ESEM-FEG manufactured by FEI Company) housed in the Imaging Technology Group of the Beckman Institute (University of Illinois).

In Situ Hybridization and Semi-Quantitative RT-PCR

Embryos for in situ hybridization were dissected in 1X DEPC PBS, fixed in 4% paraformaldehyde, then rinsed stepwise into 100% methanol and stored at −20°C until used. In situ hybridization for Fgf8 was performed on whole-mount specimens using digoxygenin-labeled RNA probes derived from mouse sequences (Tanaka et al., 1992), which also recognize the opossum transcript (Nagy et al., 2002). Fgf8 expression was assayed in opossum forelimbs of Stages 28 through 30, and opossum hind limbs of Stages 30 through 32.

To quantify Fgf8 transcript levels in opossum and mouse fore- and hind limbs, we used a semi-quantitative RT-PCR assay with 18S rRNA as a control using the Quantum RNA Universal 18S Internal Standard Kit as per the manufacturer's instructions (Applied Biosystems #AM 1718) with a primer/competimer ratio of 1:9 and 30 PCR cycles. Limb buds were dissected from comparable stages of AER development (limb Stage 4) for the opossum forelimb (from Stage 28), opossum hind limb (from Stage 31), and mouse fore- and hind limbs (11 and 12 days post-fertilization, respectively). Limb Stage 4 was targeted as it captures the early maturation stage of AER development. For each sample (e.g., opossum forelimb), all the limb buds in a single litter were combined for analysis to insure sufficient RNA levels. RNA was extracted (RNeasy kit; Qiagen #74104) and cDNA generated for each sample (SuperScript III First-Strand Synthesis system for RT-PCR; Invitrogen #18080-051). To generate conserved Fgf8 primers for opossum and mouse, we downloaded and aligned Fgf8 mRNA sequences from Ensembl. The following primers (which span a 288 bp region and are homologous in opossum and mouse) were used: GAGCAGAGCCTGGTGACGGATC (sense) and TAGTTGTTCTCCAGCACGATC (anti-sense). We used the following PCR conditions: 94°C for 20 sec, 55°C for 30 sec, and 72°C for 45 sec. PCR products were run on 3% agarose gels, and the resulting Fgf8 and 18S bands were quantified using Quantity One 1D analysis software (Bio-Rad). Significance of differences between levels of Fgf8 expression in opossum and mouse fore- and hind limbs were evaluated using Mann-Whitney U tests (Sokal and Rohlf, 1995).

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. LITERATURE CITED

Opossum Limbs Display Divergent AER Morphologies at the Gross and Cellular Level

SEM results revealed that mouse fore- and hind limbs during limb Stage 5 have fully developed and mature AER's, as expected (Table 1; Fig. 1E; Wanek et al., 1989). A similar fully developed and mature physical AER was observed in the opossum hind limb during limb Stages 5 and 6 (Stage 32; Table 1; Fig. 1F). However, the AER of the marsupial hind limb does not achieve the prominence of the AER of mouse limbs. During limb Stage 4, the opossum hind limb AER was beginning to form through the concatenation of a multitude of cell clumps forming a nearly continuous line along the DV boundary of the developing limb bud. In both the opossum hind limb and mouse limbs, the physical AER begins to reduce during limb Stage 7 (Table 1).

The opossum forelimb did not display a well-formed physical AER at any examined limb stage. Where the AER should have been, SEM results demonstrated that the opossum forelimb had only disorganized clumps of cells scattered along the DV limb border for limb Stages 4 (Fig. 1A), 5 (Fig. 1B), and 6 (Fig. 1C). During limb Stage 7, the number of cell clumps at the opossum forelimb DV boundary seemed to decrease (Fig. 1D). The location of the scattered “cell clumpings” along the DV boundary of the limb suggests that even though no organized, physical AER is present, the marsupial forelimb still has clear dorsal and ventral polarity.

Histological analyses of cellular morphology confirm that both mouse fore- and hind limbs and the opossum hind limbs possess a distinct, protruding AER separating the dorsal and ventral sides of the limb, whereas the opossum forelimb does not (Fig. 2). In mouse limbs of limb Stage 5, there is obvious thickening and compaction of the ectoderm at the distal tip of the limb bud (Fig. 2a). Similarly, the opossum hind limb at limb Stage 5 displays a significant thickening of the distal ectoderm, where the AER is located (Fig. 2c). Again, the AER of the opossum hind limb is not as pronounced as that of the mouse. In contrast, in the opossum forelimb at limb Stage 5, the thickness of the ectoderm is uniform along the distal edge of the limb bud, and there is no noticeable compaction or protrusion of cells forming an AER.

thumbnail image

Figure 2. Histological sections of M. musculus (mouse) and M. domestica (opossum) limbs during limb Stage 5, stained with hematoxylin and eosin. The AER can be seen protruding from the distal tip of the limb bud in the mouse forelimb (A), and in the opossum hind limb (C), although the AER of the opossum hind limb is relatively less prominent. However, even at the cellular level no physical AER ridge can be detected along the DV boundary of the opossum forelimb (B).

Download figure to PowerPoint

Fgf8 Expression is Conserved in Opossum Limbs

Analyses of patterns and levels of Fgf8 expression indicate that Fgf8 expression is conserved in opossum limbs. As revealed by in situ hybridization, Fgf8 transcripts are expressed in a solid line along the DV boundary of both the developing fore- and hind limb in opossums, in a similar manner to that previously documented in other tetrapods (Fig. 3; Crossley and Martin, 1995; Crossley et al., 1996; Cooper et al., in press). Moreover, semi-quantitative RT-PCR analysis of the levels of Fgf8 transcript revealed that the level of Fgf8 transcript is indistinguishable (p = 0.8873) in mouse and opossum fore- and hind limbs of comparable developmental stages (Fig. 3). Taken together, these results indicate that Fgf8 is expressed in the same levels and patterns in opossum and mouse limbs, despite the opossum forelimb's lack of a physical AER, and the opossum hind limb's less prominent AER.

thumbnail image

Figure 3. Fgf8 expression in M. domestica (opossum) fore- and hind limbs. In situ hybridization reveals that Fgf8 is expressed (indicated by black staining) normally in opossum forelimbs (A and B - Stage 28; limb Stage 3) and hind limbs (C and D - Stage 30; limb Stage 2). Arrows indicate Fgf8 expression in the limb buds. (c) Semi-quantitative RT-PCR reveals that Fgf8 is also expressed the same relative level in the fore- (FL) and hind limbs (HL) of opossums (Mono.) and mouse, Mus musculus (Mus) at limb Stage 4.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. LITERATURE CITED

Several conclusions can be drawn from the results of this study. First, and perhaps most interestingly, this study suggests that the marsupial, M. domestica, is the only studied amniote (without reduced limbs) to not develop a mature, physical AER in all of its limbs. It is, of course, possible, that a physical AER appears only very briefly in the opossum forelimb, for a fraction of a limb stage. This is unlikely, however, as limb stages happen over very short spans of time during opossum forelimb development, each lasting only a couple of hours (personal observation). At the very least, results of this study allow us to conclude that the opossum forelimb either never forms a mature, physical AER, or forms a physical AER for just a small fraction of limb development relative to opossum hind limbs and mouse limbs. The physical AER of the opossum hind limb is also reduced relative to mouse, but not to the severe extent of that of the opossum forelimb. The variation in the degree of AER protrusion among opossum hind limbs and mouse limbs is not altogether unexpected, as AER prominence differs slightly in other tetrapods. For example, the chick AER is notably more pronounced that that of the mouse (Niswander, 2003; Fernandez-Teran and Ros, 2008).

Second, despite not forming a mature, physical AER, the M. domestica forelimb normally expresses Fgf8, a well-known genetic marker for the AER. Furthermore, despite its relatively reduced prominence, the opossum hind limb also normally expresses Fgf8.

Taken together, these results suggest that the physical AER of both opossum limbs has undergone an evolutionary reduction, but that in the opossum forelimb this reduction was taken to the extreme. Additionally, results suggest that although the opossum forelimb does not have a physical AER, both it and the opossum hind limb retain a molecular AER. These findings are consistent with our hypothesis that the marsupial AER was altered to help achieve or as a result of the rapid outgrowth of their forelimbs. However, further testing is needed to determine if the reduction of the AER played a role in the acceleration of opossum forelimb development, or if it was just a byproduct of it. Although the opossum represents the only marsupial species in which AER morphology has been investigated, the opossum's basal placement in the group (Horovitz et al., 2009), and the general similarity of limb development among marsupials (e.g., Sears, 2004; Bininda-Emonds et al., 2007; Sears, 2009), leads us to predict that the AER morphology observed in M. domestica will be found to represent the general marsupial condition.

With these results many new questions arise. First, it is surprising that the physical AER of the opossum forelimb is so dramatically reduced, as a current hypothesis is that the physical AER functions to condense cells and cellular products and thereby accelerate growth (Fernandez-Teran and Ros, 2008). As marsupial forelimbs develop fastest of any examined amniote (Sears, 2009), to be consistent with this hypothesis, the AER of the opossum forelimb should actually be enlarged, rather than reduced. This suggests that size of the AER, or even its physical presence, are not directly linked to rate of limb outgrowth. Results of this study are also not consistent with another current model of AER function, namely that the AER provides the mechanical support necessary for proper DV patterning (Dahmann and Basler, 1999), as opossum forelimbs successfully establish their DV polarity despite not having a physical AER.

However, the question remains; why is the physical AER of the opossum forelimb so dramatically reduced? Perhaps as a result of the extreme shortening of its development, opossum forelimbs just do not have the time or the resources to build a fully mature, protruding AER. If this is the case, the disorganized clumps of cells observed along the DV boundary of the developing opossum forelimb could be the vestigial remnants of an ancestral, morphological AER. In this scenario, despite not forming a morphological AER, opossums are able to maintain limb outgrowth through the presence of a molecular AER (as indicated by Fgf8 expression). A similar situation exists in other amniotes that lack a ridge-like AER but form a fully functional limb [e.g., urodeles (e.g., Han et al., 2001)]. As a note, Fgf8 expression has not yet been investigated in the direct developing frog, Eleutherodactylus coqui. These findings suggest that the formation of a fully developed, protruding physical AER is not essential to limb development. Given that the function of a physical AER is not known, despite years of study (Fernandez-Teran and Ros, 2008), and that the actual morphology of the AER can be experimentally manipulated without affecting limb outgrowth (Errick and Saunders, 1974, 1976), this is a strong possibility. However, the strong conservation of a well-developed, protruding AER across tetrapods belies the assertion that the morphological AER has no function.

Physical AER's have been observed in most tetrapods, in all amniotes without reduced limbs, and in all mammals (even those with reduced limbs) that have been studied (Stopper and Wagner, 2005; Cooper et al., in press). In addition to the findings of this study, researchers have found only two tetrapod groups with fully developed adult limbs that do not form physical AER's, and both are non-amniotes with specialized development. The first is the urodele amphibians, which retain a capacity for limb regeneration into adult-hood (Sturdee and Connock, 1975), and the second is Eleutherodactylus coqui, a frog that develops directly with no larval stage (Richardson et al., 1998). The actual morphology of the AER varies between tetrapod species, with some AER's being more prominent and/or stratified than others, but the condition of having a physical AER appears to be ancestral for tetrapods. This argument is strengthened by the fact that several fish (e.g., zebrafish, killfish, and the Australian lungfish) also possess morphological AER's (Wood, 1982; Grandel and Schulte-Merker, 1998; Hodgkinson et al., 2007; Mercader, 2007). If a morphological AER truly played no essential role in limb development, we would expect to have seen it reduced or lost by chance in more amniote lineages than in just marsupials. In any event, the situation in opossums indicates that more study of the role that the physical AER plays in limb development is needed.

Another outstanding question is–how have opossums modified their limb development to achieve this highly divergent AER morphology? Results of this study suggest that pathways other than those involving Fgf8 (e.g., interactions between the BMP's and engrailed, or between the Wnt's and other FGF's) have been modified in opossum forelimbs relative to the limbs of other tetrapods to generate the opossum's divergent morphology. The AER goes through four phases during its duration (Fernandez-Teran and Ros, 2008). In the first phase, AER induction, AER precursor cells are specified, and migrate to the distal tip of the developing limb bud. AER maturation, the second phase, is characterized by the establishment of a mature, protruding AER through the compaction of the precursor cells. In the third stage, AER maintenance, the AER is maintained as a functional structure. Finally, in the fourth stage, AER regression, the AER flattens and regresses. Given the morphology of the cells in the region where the opossum forelimb AER would be (and their similarity to the early stages of AER development in the opossum hind limb), we propose that induction of the opossum forelimb AER (Phase 1) begins normally, and that the AER precursor cells are specified and migrate to the distal tip of the developing limb bud. However, we propose that the compaction of these precursor cells (Phase 2) to make a mature, protruding AER is disrupted in opossum forelimbs. With this in mind, it is possible that the expression of genes associated with this compaction have also been disrupted (e.g., Wnt's, Dkk1, En1, etc.; Fernandez-Teran and Ros, 2008). Testing of these hypotheses is currently underway.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. LITERATURE CITED

The authors would like first and foremost to acknowledge Kathleen Smith for conversations that provided the impetus for this study. The authors also thank the members of the Sears and Marcot Laboratories at the University of Illinois for stimulating discussions on this topic that improved the study and manuscript. The authors also thank two anonymous reviewers whose comments strengthened the manuscript. Finally, the authors thank Lisa Powers for help with experimental implementation and training.

LITERATURE CITED

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. LITERATURE CITED