SEARCH

SEARCH BY CITATION

Keywords:

  • heterochrony;
  • marsupial limb;
  • species-specific differences

ABSTRACT

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. ACKNOWLEDGMENTS
  7. REFERENCES

Background: At birth, marsupial neonates have precociously developed forelimbs. The development of the tammar wallaby (Macropus eugenii) hindlimbs lags significantly behind that of the forelimbs. This differs from the grey short-tailed opossum, Monodelphis domestica, which has relatively similar fore- and hindlimbs at birth. This study examines the expression of the key patterning genes TBX4, TBX5, PITX1, FGF8, and SHH in developing limb buds in the tammar wallaby. Results: All genes examined were highly conserved with orthologues from opossum and mouse. TBX4 expression appeared earlier in development than in the mouse, but later than in the opossum. SHH expression is restricted to the zone of polarising activity, while TBX5 (forelimb) and PITX1 (hindlimb) showed diffuse mRNA expression. FGF8 is specifically localised to the apical ectodermal ridge, which is more prominent than in the opossum. Conclusions: The most marked divergence in limb size in marsupials occurs in the kangaroos and wallabies. The faster development of the fore limb compared to that of the hind limb correlates with the early timing of the expression of the key patterning genes in these limbs. Developmental Dynamics 243:324–338, 2014. © 2013 Wiley Periodicals, Inc.


INTRODUCTION

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. ACKNOWLEDGMENTS
  7. REFERENCES

In most mammals, there is little difference in the development of the fore- and hind limbs at birth, although there are many varied limb structures amongst vertebrates. In marsupials, the neonate is highly altricial but possesses well-developed forelimbs that are essential for survival. The precocious development of the forelimbs enables the neonate to climb from the urogenital opening to the pouch or mammary area and attach itself to a teat where it completes its development (Renfree et al., 1989; Schneider et al., 2009; Drews et al., 2013). The South American grey short-tailed opossum (Monodelphis domestica), which has relatively similar fore- and hindlimbs in the adult, is the only marsupial so far in which developmental expression of the limb patterning genes has been examined extensively (Keyte and Smith, 2010, 2012; Sears et al., 2012). South American and Australasian marsupials diverged around 78 MYA (Bininda-Emonds et al., 2007). It is of interest to examine the development of the limbs in a marsupial that has very distinctive hind limbs with syndactylous digits, adapted for hopping. The tammar (Macropus eugenii) is a small member of kangaroo and wallaby family, and delivers a neonate that weighs around 400 mg. At birth, it has hindlimbs that are poorly developed with a basic, paddle-like form in which the digits are only just distinguishable, although syndactyly is evident (Fig. 1). It is only during post-natal development that the hindlimb of the young undergoes accelerated growth (Tyndale-Biscoe and Renfree, 1987; Renfree, 1995). To date, expression patterns of only two genes have been examined in tammar limbs, namely HOXA13 and HOXD13, both of which are involved in digit development (Chew et. Al, 2012). Thus, the goal of this investigation is to gain insights into the developmental processes that underpin the heterochronic development of the specialised fore- and hind limbs in the tammar wallaby.

image

Figure 1. A neonatal tammar wallaby. a: A tammar wallaby fetus at day 25 of gestation. Roman numerals indicate digit number. b: A scanning electron microscope image of the day-25 hindlimb.

Download figure to PowerPoint

Limb development and the patterning genes that control it are conserved amongst vertebrates. The initiation of limb development is triggered by signals in the underlying lateral mesenchyme of the developing embryo. The position of limb formation along the rostral-caudal axis is marked by the proteins Paired-like homeodomain 1 (Pitx1), T-box 4 (Tbx4), and T-box 5 (Tbx5) (Logan and Tabin, 1999; Niswander, 2003; Minguillon et al., 2005). TBX5 is expressed in the forelimb but not the hindlimb, whilst PITX1 and TBX4 are expressed in the hindlimb but not the forelimb (Logan and Tabin, 1999; Niswander, 2003; Zeller et al., 2009) (Fig. 2A). However, PITX1, although a useful marker of molecular changes associated with hindlimb expression, is not specific to the hindlimb domain. TBX4 expression is restricted to the region of hindlimb development. Thus, PITX1, TBX4, and TBX5 serve as useful markers to examine the timing and development of fore- and hindlimb bud development.

image

Figure 2. A diagram showing genes involved in the initiation of the forelimb and hindlimb and the feedback loop regulating outgrowth. A: Fore and hindlimb buds are initiated by TBX5 expression in the forelimb and PITX1 and TBX4 in the hindlimb. These genes then initiate FGF10 expression, which induces expression in the ectoderm. B: A feedback loop is initiated between the apical ectoderm and the ZPA, which secretes SHH. An interaction between BMP4 via GREM1 is established to determine limb patterning and outgrowth. FL, forelimb; HL, hindlimb. Roman numerals indicates digit number.

Download figure to PowerPoint

Once the fore and hind-limb fields are determined, a signalling feedback loop is initiated. TBX4 and TBX5 in the lateral plate mesoderm induce fibroblast growth factor 10 (FGF10) expression, which initiates differentiation and outgrowth of the limb buds, and signals to the overlying ectoderm (Ohuchi et al., 1997; Takeuchi et al., 1999, 2003) (Fig. 2B). Another key signalling centre is the zone of polarizing activity (ZPA), located in the posterior region of the limb. The ZPA secretes a morphogen, sonic hedgehog (SHH), that is essential for the establishment of anterior-posterior patterning and digit identity (Echelard et al., 1993). In oligozeugodactyly (ozd) chick mutants that lack limb-specific ZPA/SHH function, there is a loss of anterior-posterior skeletal elements, which include the loss of digits (Ros et al., 2003). The SHH knockout mice show a similar limb phenotype to ozd chickens (Kraus et al., 2001; Ros et al., 2003). In contrast, exogenous addition of SHH via a bead inserted ectopically into the anterior border of a chicken limb bud in culture results in mirror image duplications of digits (Lopez-Martinez et al., 1995). SHH also has several downstream effectors including the protein Patched homologue (PTCH), and members of the glioma-associated oncogene homologue (GLI) family of transcription factors (O'Rourke et al., 2002; Bowers et al., 2012). Interestingly, loss of Gli3 in the limb bud results in gain of digits polydactyly (Zuniga and Zeller, 1999).

The apical ectodermal ridge (AER) is another critical signalling centre for limb development that plays an integral role in limb bud outgrowth (Capdevila and Izpisua Belmonte, 2001). The AER secretes fibroblast growth factor 8 (FGF8) that initiates an epithelium-mesenchyme feedback signalling loop involving SHH in the ZPA (Weatherbee et al., 2006; Hockman et al., 2008). This loop is also regulated by Gremlin (Grem1). In most eutherian mammals, the AER is manifest as a visible ridge protruding from the distal edge of the limb bud in both the fore- and hindlimbs. However, in the opossum (defined as Monodelphis domestica in this report) (Doroba and Sears, 2010) and the Mexican axolotl (Ambystoma mexicanum) (Han et al., 2000), the AER lacks a distinct ridge-like form. The opossum AER expresses FGF8 despite the forelimb AER consisting of only several disorganised cells that protrude above the surface ectoderm. In contrast, the hindlimb has a fully developed AER, but it is not as prominent as that of the mouse (Doroba and Sears, 2010).

In the opossum, establishment of the forelimb field appears to arise earlier when compared to that of the mouse and expression of TBX5 precedes formation of the thoracic somites. In contrast, the mouse forelimb Tbx5 expression starts about the time when thoracic somites form (Keyte and Smith, 2010). Likewise, the opossum hindlimb fields appear to arise slightly earlier and expression of TBX4 precedes formation of the first lumbar somites, whilst in mice these events are synchronous, so the establishment of fore-and hindlimb fields is accelerated in opossum (Keyte and Smith 2010). After induction of the forelimb, the opossum forelimb has earlier initiation of outgrowth, associated with greater myocyte allocation. There is early expression of the limb bud patterning factors SHH and FGF8 suggesting there is an acceleration of the opossum forelimb growth and development, whilst development of the hindlimb is delayed and almost pauses at the bud stage (Keyte and Smith 2010, 2012). Thus, there appear to be differences between marsupials and eutherians in the relative timing of expression of the genes regulating the development of the fore- and hindlimbs.

Differences in the timing of developmental events, or heterochronies, are most marked in marsupials with their mixture of precocial and altricial features along the anterior-posterior axis in the neonate. Heterochrony was first defined to refer to a change in relative timing of developmental events in two related taxa, but the term has become more broadly defined to apply to a wide range of developmental processes or developmental sequence analyses (Smith, 2001). In this study, we use the definition of heterochrony of Smith (2001) as applied to marsupials by Keyte and Smith (2010, 2012) to examine the difference in the timing of the development of the fore- and hindlimbs in the tammar wallaby during embryonic and fetal development.

It is difficult to directly compare the different developmental stages between the mouse, tammar, and opossum. The only complete embryology of a marsupial remains as that of the North American opossum, Didelphis virginiana (McCrady, 1938). McCrady stages were developed for this opossum, in which the gestation period is only 13 days. Keyte and Smith have more recently adapted the Didelphis stages (McCrady) for Monodelphis to fit the 35 stages into 14.5 days. The tammar, which is monovular, undergoes an 11-month diapause followed by a 26.5-day pregnancy so it does not readily conform to McCrady stages. Thus, for ease of comparison we have chosen to use days post-conception for mouse and opossum or days post-reactivation of the diapausing blastocyst for the tammar as the basis of comparison, and when relevant, state the equivalent McCrady stage for Monodelphis. Macropodid marsupials deliver the most well-developed marsupial fetus, so we examined the heterochronous expression of the key limb patterning genes in the fore- and hindlimbs of the developing tammar wallaby and compared them to those of the opossum and mouse.

RESULTS AND DISCUSSION

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. ACKNOWLEDGMENTS
  7. REFERENCES

Whilst there is great diversity in final limb form, all limbs go through a similar sequence of patterning. Marsupials are born with more advanced forelimbs compared with the hindlimb but within the marsupials there is considerable diversity in form. In the tammar, the forelimb field was first apparent as a narrow condensation parallel to the neural tube between somites 6–13, with no hindlimb visible between early to mid-day 17 (Table 1). At the end of day 17, the forelimb ridge became further shortened between somites 8 and 12, but there was still no sign of the hindlimb. A day later (day 18), the forelimb had formed into a small distinct bud while the hindlimb was just visible as a condensation of tissue. By day 19, the forelimb had formed into a paddle-like shape whilst the hindlimb became a slightly raised protrusion. The forelimb between day 20–day 21 had distinct ridges where the prospective digits will form. In contrast, the hindlimb had grown in size and was now a flattened, pointed paddle. A day later at day 23, the first signs of the epitrichial claws in the forelimb were observed and the hindlimb possessed obvious syndactylous digits. A day before birth, day 25, the forelimb was well-formed with clearly demarked epitrichial claws whereas the hindlimb only had the beginning of digits. The heterochrony in the tammar forelimb is dramatically marked by the in utero climbing movements of the forelimb from about 3 days before birth as observed by ultrasound, while the hindlimb remains completely immobile (Drews et al., 2013).

Table 1. Stages of Development of the Limb in the Tammar Wallaby, Macropus eugenii
Days of gestationDescription of tammar limb
7 early-midForelimb regions show little to no protrusions and the hindlimb ridge has not formed (Early: ∼16–17 somites, Mid: ∼19–20 somites)
17 lateSmall protrusion of the ectoderm in the prospective forelimb region and no presence of the hindlimb (∼21–23 somites)
18 EarlyThe forelimb bud is formed and will begin to have a defined shape. The hindlimb is a small protrusion
18 mid
18 late
19Paddle-like structure in the forelimb but the hindlimb is a small protrusion
20–21Forelimbs are paddle-like and digital rays formed. Hindlimb is a small bud but with no distinct ridges
22Forelimb digital rays are pronounced and the hindlimb has formed a flattened pointed structure
23–24Forelimb epitrichial claws are small and digits are pronounced. The hindlimb is a syndactylous paddle with digit separation
25–26 (birth)Defined forelimb form and epitrichial claws are obvious in the forelimb. The hindlimb has distinct digits and there was elongation of the 4th digit.

Induction of the Forelimb

T-box protein 5 is expressed at an early developmental stage in tammar forelimbs.

The pattern of TBX5 expression displayed in the tammar limbs is similar to those published in the mouse and the opossum but the timing of TBX5 expression differed markedly between the species. In the mid day 17 (∼18 somites), the earliest stage examined for this gene (Fig. 3A, Table 2), TBX5 was expressed in the limb mesenchyme from the 5th somite and extended to at least 12th somite. By late day 17 (∼23 somite stage), there were distinct forelimb buds marked by TBX5 expression between somite 8 and 12 (Fig. 3B). In the embryo early on day 18 of gestation (Fig. 3C), TBX5 expression became restricted to the underlying mesenchyme and the buds became more elliptical in shape. This pattern persisted from late day 18 to day 19 (Fig. 3D and E). At day 19 to 21 (4–5 days before birth) (Fig. 3F, G and H), TBX5 staining became restricted to the hand plate and was not expressed at the distal tip of the forelimb. By day 23 (3–4 days before birth), TBX5 was no longer detected in the developing forelimb (Fig. 3I).

image

Figure 3. mRNA in situ hybridisation analysis of TBX5 on tammar embryo limbs in whole mounts and dissected fetal limbs. TBX5 is first detected at 20 somites (arrow) and the staining marks the formation of the forelimb ridges in the (A) 20 somite embryo (d18 of gestation) and (B) 23-somite (d18 of gestation). (C) On day 18 (early) gestation to day 19 of gestation, the ridge has developed into a bud. TBX5 is also detected in the heart. (D-I) TBX5 staining of forelimbs becomes more diffuse in the older stage limb buds, day 19+ fetus to day 23 fetus. There was no staining in the hindlimb region of any specimens examined. All limb bud images are orientated to point in a distal direction. FL-forelimb.

Download figure to PowerPoint

Table 2. Numbers of Limbs Used for Each Gene at Each Developmental Stagea
 Gene and limb
 TBX5PITX1TBX4SHHFGF8
Days of gestationForelimbHindlimbHindlimbForelimbHindlimbForelimbHindlimb
  1. a

    n/d, not done.

16n/dn/dn/dn/dn/dn/dn/d
17 early2n/dn/dn/dn/dn/dn/d
17 mid2n/dn/d22n/dn/d
17 late22222n/dn/d
18 early - mid32232n/dn/d
18 late32322n/dn/d
193333344
203333322
y 213333222
223333322
2333n/d3334
25n/dn/dn/dn/dn/d34
26 (birth)n/dn/dn/dn/dn/dn/dn/d

We then compared the timing and spatial pattern of expression of tammar TBX5 with patterns published for mouse and opossum. The expression patterns were similar but there were marked differences in timing. In the mouse, limb-associated Tbx5 is first detected at E8.5 in the forelimb (Fig. 4A), which is considerably later than the opossum that is first detected in the presumptive limb field before late day 10.5 (McCrady stage 24, ∼10 somites) (Keyte and Smith, 2010; Gibson-Brown, 1996) (Fig. 4B and C). At this time, the opossum neural tube has not closed, the heart and otic pit are not fully formed, but in the mouse these structures are already evident (Keyte and Smith, 2010). Interestingly, mouse TBX5 KO prevented outgrowth at E9 and presumably TBX5 is present before day 9 (Agarwal et al., 2003). Thus, forelimb development in the opossum occurs at a much earlier relative developmental stage than in the mouse (Keyte and Smith, 2010). We find that a similar heterochronic shift in the timing of forelimb induction is also present from the earliest stages examined in the tammar. However, TBX5 expression in the tammar was only marginally earlier when compared with that of the mouse and relative to the development of other structures such as the heart and head (Table 1, Fig. 4). At birth, the morphological difference between the tammar fore- and hindlimb is more dramatic when compared with that of the mouse.

image

Figure 4. Comparison of TBX5 expression in the limbs of the (A) mouse, (B) opossum, and (C) tammar. Patterns of mRNA expression from the mouse and opossum were traced from previously published data (Keyte and Smith, 2010; Minguillon et al., 2012). The purple represents the region of mRNA expression in the limbs. Unless indicated, all specimens are orientated to point distally.

Download figure to PowerPoint

These observations in the shifts in developmental timing in two divergent marsupials, the opossum and tammar, which last shared a common ancestor around 78 million years ago, suggest that early forelimb development is a conserved marsupial feature. In most species, there is advanced development of the forelimb but marsupials are considered to show an extreme example of heterochrony (Bininda-Emonds et al., 2007; Sears, 2009). This advanced development of the forelimb relative to the hindlimb in marsupial species could be achieved through an acceleration of the forelimb elements, a delay in the hindlimb elements, or a combination of both (Keyte and Smith, 2010; Sears, 2009, Bininda-Emonds et al., 2007). Marsupials have advanced the developmental timing of the formation of the forelimb and shoulder girdles when compared to those of eutherian mammals (Weisbecker et al., 2008).

The precocious development of the marsupial limb is determined during somitogeneis, well before the initiation of the limb buds. It appears that the forelimb heterochrony is at least partially achieved by a differential rate of somitogenesis, since in the opossum somites initially form a pair about every hour and the posterior somite formation is slowed gradually to form approximately one pair every 4.5 hr (Keyte and Smith, 2012). In addition, somitogenesis is initiated earlier in the opossum relative to other developmental events compared to the mouse (Keyte and Smith, 2012). This appears to also be the case in the tammar, in which somites are present before neural tube closure (Fig. 4A). Between days 16 and 18 of gestation, caudal somites are generated approximately every 4.5 hr, a very similar rate to that of the opossum. Thus, it appears that somite maturation and the allocation of tissues to derivative structures could also be playing a role in the heterochrony in both marsupial species' forelimbs. What first induces TBX5 expression is not yet clear, although HOX signalling may be involved in this process as well as somitogenesis (Minguillon et al., 2012).

Hindlimb Patterning and Outgrowth

PITX and TBX4 are expressed later in development than in opossum but earlier than in mouse.

A delay in the formation of hindlimb elements could also contribute the extreme heterochrony in the marsupial phenotype (Keyte and Smith, 2010; Sears, 2009). Thus, we also examined the expression of the two hindlimb marker genes PITX1 and TBX4 using whole-mount in situ hybridisation on tammar embryos. Interestingly, PITX1 mRNA was not detected in the day-18 embryo (mid) (Fig. 5A, Table 2), but was first detected in the hindlimb a day later (Fig. 5B). At no stage was PITX1 detected in the forelimb bud. However, PITX1 was present in the pharyngeal folds and head region of the day-18 (late) tammar embryo, (Fig. 5B). The PITX1 signal became more diffuse as the hindlimb bud developed and was expressed in the underlying limb bud mesenchyme (Fig. 5C–E). As the future foot plate (autopod) formed into an arrow-shaped bud, the staining in the wallaby hindlimb was isolated in the early zeugopod region of the limb (Fig. 5F).

image

Figure 5. Expression of PITX1 in tammar embryos. A: PITX1 had no detectable expression in either the forelimb or hindlimb at day-18 embryo but PITX1 mRNA is detected from day 19 to 24 in tammar embryo hindlimbs (B–F). The expression is detected in the pharangeal arch of the early embryo and the prospective hindlimb bud. B: Early day-19 fetal hindlimb staining. C: Late day-20 fetal hindlimb. D: Day-21 fetal hindlimb. E: PITX1 mRNA encompasses most of the limb bud but becomes restricted to the “trunk” of the hindlimb arrow-shaped paddle at (F) day 23. FL, forelimb; HL, hindlimb.

Download figure to PowerPoint

PITX1 is expressed in the hindlimb bud at E9.5 in the mouse (Fig. 6A) (Keyte and Smith, 2010). PITX1 expression in the opossum is also hindlimb specific (Fig. 6B) (Keyte and Smith, 2010). Similarly, PITX1 was detected in the tammar hindlimb in a pattern similar to that in the mouse and opossum (Fig. 6C).

image

Figure 6. Comparison of PITX1 expression in the limbs of the (A) mouse, (B) opossum, and (C) tammar Patterns of mRNA expression from the mouse and opossum were traced form previously published data (Keyte and Smith, 2010; Duboc and Logan, 2011). The purple represents the region of mRNA expression in the limbs. Unless indicated all specimens are orientated to point distally.

Download figure to PowerPoint

Whilst PITX1 is upstream of TBX4, its expression is less informative as it does not only mark the hindlimb domain. Thus, we also examined TBX4, which was specifically detected in the hindlimb and not in the forelimb of the tammar (Fig. 7, Table 2). Notably, in some specimens, there was strong expression in the developing heart (Fig. 7A, Table 2), which is consistent with TBX4 expression in other species (Rallis et al., 2003; Naiche and Papaioannou, 2007; Keyte and Smith, 2010). The earliest expression of TBX4 occurred in the hindlimb of the early day-18 embryo in the wallaby (Fig. 7B and C). The intensity of the staining increased from late day-18 to late day-19 fetal hindlimb (Fig. 7D and E) but decreased at day 20 (Fig. 7F and G), and by day 22 TBX4 was not detected (Fig. 7H).

image

Figure 7. Expression of TBX4 in tammar embryos and limb buds (A–H) TBX4 marks the beginning of the hindlimb bud formation in a late day-18 embryo. In (A) arrow indicates faint staining. However, a distinct expression is detected in the heart at day 19 early. (B) Staining appears strongest towards the posterior end of the embryo (towards the tail) (C–E). Top panels are dorsal perspective and bottom panels ventral. TBX4 staining was detected in the entire hindlimb bud at day 20 early (F), but by the end of day 20 (G), the staining appears to be weaker and by day 22 (H), there is no staining. Whole mount embryos show no staining in the prospective forelimb region in any of the specimens examined. Arrow indicates staining the heart tube region. FL, forelimb bud; HL, hindlimb region.

Download figure to PowerPoint

The tammar wallaby staining here corresponds with the patterns observed in the mouse and chicken. In the mouse, TBX4 expression is strongly detected at E9.5 (Fig. 8A) but this is developmentally much later compared with expression in the opossum, which is first detected before day 10.7 (McCrady stage 26) (Fig. 8B; see also Fig. 11). In the tammar, the timing of TBX4 expression was similarly shifted and was developmentally detected earlier than in the mouse but was later than in the opossum (Fig. 8C). Sears (2009) suggested that the earliest development of the marsupial hindlimb is also accelerated, which appears to be the case in the opossum but not the tammar. Interestingly, mouse Tbx4 expression continues to be detected at least until E13.5 restricted to the autopod region, but in the tammar hindlimb it was detected only until the bud stage.

image

Figure 8. Comparison of TBX4 in the limbs of the (A) mouse, (B) opossum, and (C) tammar. Patterns of mRNA expression from the mouse and opossum were traced from previously published data (Rallis et al., 2003; Keyte and Smith, 2010). The purple represents the region of mRNA expression in the limbs. Unless indicated, all specimens are orientated to point distally.

Download figure to PowerPoint

Both TBX4 and PITX1 expression showed hindlimb-specific patterning, which is consistent with expression reported in the mouse and chicken (Minguillon et al., 2005; Duboc and Logan, 2011). Pitx1 in later development has a role independent of Tbx4 in shaping bone and muscle morphology in the hindlimb. Mice null for Pitx1 have hindlimb defects in this region that are not rescued by Tbx4 expression (Duboc and Logan, 2011).

Downstream Limb Patterning and Outgrowth

Heterochrony of SHH expression between fore- and hindlimb.

We observed heterochrony in tammar limb field initiation using the fore and hindlimb markers TBX4, TBX5, and PITX1 but it was unclear whether downstream growth and patterning genes also differed between the fore- and hindlimb in the tammar. Thus, we examined a key patterning gene, SHH, that marks the zone of polarising activity, using whole-mount hybridisation on limb buds from day 18 to day 23 of gestation, a time when the syndactylous digits in the hindlimb develop.

SHH mRNA was absent in the tammar forelimb before day 18 (Fig. 9A) and very faintly detected at mid-day 18 (Fig. 9B). The expression intensified by late day 18 (Fig. 9C) at a time when the limb begins to form into a paddle-like shape (Fig. 9D). However, there was no staining detected in the hindlimb early on day 18 of gestation (Fig. 9I–K). At day 19 (late) (Fig. 8E), the forelimb has become a distinct paddle with the presence of future zeugopod elements (radius and ulna) but SHH mRNA was still detected in the autopod (hand plate). A day later, SHH was no longer detected in the forelimb and the distinct ridges of the digital rays were visible (Fig. 9F–H). SHH expression was first detected in the hindlimb on late day 19, almost two days after SHH was first detected in the forelimb (Fig. 9L), clearly demonstrating the marked heterochrony between the fore- and hind limbs. Expression of SHH continued in the hindlimb until day 22 (Fig. 9M and N) and at this stage the staining was most strongly detected. However, by day 23, expression in the hindlimb was no longer detected (Fig. 9O).

image

Figure 9. SHH mRNA in situ hybridization on tammar forelimbs (A–H) and hindlimbs (I–O). A: SHH is not detected in the early day-18 embryo. B: However, in the forelimb SHH is first detected at in the posterior end of the early day-18 fetus (C,D) but the strongest localisation in the day 18 (mid). E: SHH is still detected when the forelimb bud has elongated at day-19 fetus. F–H: There was no expression detected in the forelimb at day-20, −22, or −23 fetus. I–K: No expression is detected in the hindlimb until (L) at day-19 fetus. SHH then remains detectable at (M) day 20-21 until (N) day -22 fetus but (O) at day-23 expression in the fetus is no longer detected.

Download figure to PowerPoint

The first onset of SHH expression in the tammar forelimb was at early day 18, when the forelimb is a small protrusion with the beginnings of a hindlimb bud. This is at an earlier developmental stage than in the mouse at E9.75 (Echelard et al., 1993), which has a better defined forelimb bud, hindlimb and the posterior neuropore is closing (Figs. 10A, C, 11, Table 2). However, in the opossum forelimb, the first reported expression appears to be much earlier at approximately day 10.6 (McCrady stage 25) when the embryo does not have a pronounced limb-bud but has fused heart tubes and between 12–13 somites. This is approximately equivalent to 15–16 days of gestation in the tammar (Figs. 10B, C, 11) and much earlier than in the mouse (Keyte and Smith, 2010).

image

Figure 10. Comparison of SHH expression in the limbs of the (A) mouse, (B) opossum, and (C) tammar. Patterns of mRNA expression from the mouse and opossum were traced form previously published data (Echelard et al., 1993; Kruger et al., 2001; Keyte and Smith, 2010). The purple represents the region of mRNA expression in the limbs. Unless indicated all specimens are orientated to point distally.

Download figure to PowerPoint

image

Figure 11. The comparison of limb events in the mouse, opossum, and tammar. The tammar events are based on a figure from Keyte and Smith (2010) and all stages are listed in days (see Table 1). We show the differing morphologies of each species at relevant time points as line diagrams on this figure. There appears to be greater shift in forelimb marker TBX5 in the opossum compared to the mouse and tammar. Note that TBX5 and TBX4 expression is only slightly earlier in the tammar but SHH expression is much earlier. FL, forelimb; HL, hindlimb.

Download figure to PowerPoint

There is marked heterochrony in SHH expression between the fore- and hindlimbs in the tammar, and at day 19, SHH was detected in the hindlimb (Fig. 10D), which is a just a bud compared to the forelimb that has already elongated to form a paddle-like structure. SHH is not expressed in the equivalent stage in the mouse (∼E10.5, hindlimb). This heterochrony could reflect a difference in timing of expression of the upstream genes.

In the opossum, both the fore- and hindlimbs develop much earlier compared to the mouse (Keyte and Smith, 2010). In contrast, we found that tammar TBX4 and TBX5 was expressed only slightly earlier when compared to the mouse (Fig. 11), although SHH in both the tammar fore- and hindlimb was detected much earlier than in the mouse. Together this suggests that acceleration of the tammar forelimb is achieved through rapid advancement of the initiation of development of the limb field (TBX5) and the later limb-controlling elements such as SHH. It is also possible that the tammar heterochrony is achieved through both the acceleration of the forelimb program and the slowing of the hindlimb program. Whilst the slowing of the hindlimb is a possibility we have no evidence in the current study to suggest if this is the case. Here we highlight differences and similarities between these two marsupials, which both show forelimb heterochrony.

FGF8 and the Apical Ectodermal Ridge (AER)

The AER is more prominent in the forelimb.

The formation of the AER and FGF8 expression delineating the AER is thought to be essential for the correct development of the limb. In the tammar at day 19, FGF8 mRNA showed a clear and restricted expression pattern in both the hindlimb and forelimb Fig. 12A–C, Table 2). The staining was detected in the distal ectodermal region of the embryo and not in the underlying mesenchyme (Fig. 12B). FGF8 staining was also detected in the tail bud region (Fig. 12C). This staining confirms that expression pattern of FGF8 in the wallaby in both fore- and hindlimb is consistent with that of the mouse, opossum, and chicken (Lewandoski et al., 2000; Keyte and Smith, 2010).

image

Figure 12. The expression of FGF8 in the tammar AER. A: An FGF8 wholemount in situ hybridisation of a day-19 embryo showing staining in the fore- and hindlimb buds and pharyngeal folds. B: A close-up examination of the forelimb shows staining on the distal region of the outer ectoderm and on the (C) hindlimb and tail bud. Arrowhead points to tail bud. FL, forelimb; HL, hindlimb.

Download figure to PowerPoint

At day 19 of gestation, there was a distinct AER in the forelimb but not the hindlimb of the tammar embryo (Fig. 13A, C, E). The AER was formed by a ridge of thickened ectoderm in the forelimb (Fig. 13G) but not in the hindlimb at this stage (Fig. 13B and D). By day 20–21, the tammar hindlimb had developed an AER (Fig. 13H). The delayed development of the hindlimb AER correlates with the heterochrony that is observed between the tammar fore- and hindlimb.

image

Figure 13. Apical ectodermal ridge in the tammar limb. Whole mount images of the normal (A) forelimb and (B) hindlimb of a day 19 of gestation embryo. Representative scanning electron microscope (SEM) images of a (C) forelimb. Inset (E): High power of apical ectodermal ridge (E) magnified forelimb and (D) hindlimb bud at day 19 of gestation. Representative histological sections stained with haematoxylin and eosin of a day-19 of gestation (F) hindlimb and (G) forelimb and a representative hindlimb at (H) day 20–21 of gestation. For all images, dorsal and ventral aspect is labelled with a black arrow indicating the position of the AER, and all sections were embedded in a sagittal perspective.

Download figure to PowerPoint

There is known variability in the expression of FGF8 and in the morphology of the AER (Han et al., 2000; Cretekos et al., 2007; Doroba and Sears, 2010). Both the tammar and opossum have less pronounced AERs compared with the mouse AER. In the tammar, the forelimb had a more pronounced AER than the hindlimb, in contrast to that observed in the opossum. The opossum does not show a pronounced AER in the forelimb at any stage (Doroba and Sears, 2010). These observations support the suggestion that a distinct morphological AER is not required for correct proximal distal (PD) limb patterning, or that it is a very transient structure (Doroba and Sears, 2010). The fact that removal of the AER (Richardson et al., 1998) still results in some outgrowth supports this idea. If in fact the physical presence of an AER is needed for proximal-distal outgrowth, then we could speculate that the difference in marsupial AERs causes the differences in their relative limb development at birth. The tammar forelimb is much better developed than the opossum forelimb at birth and would thus require a more distinct AER to accelerate growth.

Species-Specific Differences

Whilst investigations into heterochrony have typically examined the differences between species, only a few have focused on the differences between fore- and hindlimb regulation (Keyte and Smith, 2010, 2012). We observed changes in the timing and regulation of the genes patterning the tammar limb compared to the mouse, but surprisingly some of these were expressed later than in the opossum. Marsupials possess the most extreme differences in the timing of development of their fore- and hind-limbs amongst mammals, but there is also variability between species in the state of development at birth.

Marsupials have been classified into three grades of development at birth. Grade1, the least advanced, is seen in the dasyurids (Dasyurus viverrinus); Grade2 is seen in the bandicoots; and Grade 3, the most advanced, is seen in the macropodid marsupials (Hughes and Hall, 1988). However, the most dramatic differences between fore- and hindlimb development occur in the honey possum (Tarsipes rostratus), which weighs less than 5 mg at birth and yet it still has to climb to the well-developed and deep pouch (Cummins et. al., 1986; Renfree unpublished observations). The developmental grading is based on the development of critical structural adaptations at birth, including olfactory structures, touch, digestive tract, excretory system, respiration, external form and limb development, and recognises that the forelimbs are particularly well advanced in macropodid marsupials. Kangaroo and wallaby young have the greatest distance to climb to reach the teat, in contrast to the dasyurids that are virtually deposited almost directly onto the mammary area. Allowing the hind limb to lag behind the precocious development of the forelimb may free resources for anterior development including forelimbs, heart, lungs, mouth, and gut (Hughes and Hall, 1988). Different periods of gestation are related to the grade of development at birth. The energy-trade off hypothesis posits that the shorter the gestation, the greater the degree of heterochrony in anterior versus posterior development. This attributes the small size of the marsupial hindlimb to the metabolic costs of the well-developed forelimbs (Muller 1967; Hughes and Hall, 1988; Smith, 2006; Weisbecker et al., 2008; Sears 2009; Keyte and Smith, 2012). These differences at birth are, in turn, explained by the differing heterochronies in gene expression that enable the precocious development of the marsupial forelimb.

Syndactyly.

The tammar hindlimb also differs in form from the more generalised marsupials and possesses only four digits (Weisbecker and Nilsson, 2008; Chew et al. 2012). Digit 1 is never present and digit 4 becomes greatly elongated. SHH pathway mutations often create polydactyl (gain of digits) digits or oligodactly (loss of digits) but in some cases result in syndactylous conditions (Suzuki, 2013) that are similar in appearance to the tammar wallaby hindlimb. Interestingly, we saw no differences in the expression of SHH expressions that could account for the marsupial syndactyly. Another potential cause of syndactyly could be modification to Homeobox (HOX) genes in the developing limb. In particular, HOXA13 or HOXD13 mutations result in syndactylous fusion and digit loss in mice (Kawakami et al., 2009). However, HOX genes D13 and A13 are highly conserved and expressed throughout limb and digit formation in the tammar (Chew et al., 2012), so the digit loss in the wallaby is as yet unexplained. In addition, as Weisbecker and Nilsson (2008) point out, the syndactylous condition of marsupials is not strictly analogous to the soft tissue syndactyly as seen in human mutations.

Conclusions

Pre-natally, the tammar forelimb is precociously developed. Post-natally, the rate of forelimb growth appears to slow while that of the hind limb rapidly accelerates. However, although the key genes that appear to control these processes pre-natally are conserved with those of other mammals, the timing of their expression could underpin the disparate development of the fore- versus hindlimb. The energy trade-off may be the driving force in this heterochrony, but the questions remain as to whether the forelimb is accelerated, or the hindlimb delayed, or both. In the case of the tammar wallaby, it seems that there is both an acceleration of forelimb development in the developing embryo and fetus and a delay of hindlimb development until after birth when the situation changes and hindlimb growth exceeds that of the forelimb.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. ACKNOWLEDGMENTS
  7. REFERENCES

Specimen Collection and Staging of Specimens

Tammar wallaby specimens (Macropus eugenii) were collected from Kangaroo Island, South Australia. Embryos were fixed whole in 4% paraformaldehyde overnight 4°C and washed twice in 1 × phosphate buffered saline (PBS) and stored at −20°C in methanol ready for gene expression analysis (n = 47). Each tammar normally has only a single embryo per year, conceived at a post-partum oestrus. Removal of the sucking stimulus of the young in the pouch results in reactivation of the diapausing blastocyst and birth 26.5 days later (Renfree et al., 1989). Earlier stages are more difficult to sample because the exact time of reactivation is variable between animals. We ensured that for the rare and precious early somite stages, each staining was performed at least twice (Table 2). In older stages, left and right fore- and hindlimbs were dissected away from the body before staining and in situs performed at least in triplicate. Additional specimens for scanning electron microscopy and histology collected at days 18 (n=3), 19 (n=3), 20 (n=3),and 21 (n=4) were fixed in Superfix (0.2% glutaraldehyde in cacodylate buffer) and washed twice in cacodylate buffer for 60 min and stored until ready for SEM processing or in 4% paraformaldehyde (PFA) until staining in haemotoxylin and eosin. Fetal limbs were dissected from the main body trunk of an additional 10 embryos and snap frozen for RT-PCR (n = 2 at each of 5 stages). All embryo ages were estimated by measurements of head-length, crown-rump-length, weight using developmental growth curves and images, (Renfree, 1973, Tyndale-Biscoe and Renfree, 1987) (Table 1). However, the age of some specimens has been estimated to within half a day as there is a poor correlation between age and stage in individual animals after reactivation from embryonic diapause and there is some variation of limb phenotypes relative to other features between fetuses. All sampling techniques and collection of tissues conformed to Australian National Health and Medical Research Council (2004) guidelines and were approved by The University of Melbourne Animal Experimentation and Ethics Committees.

Extraction and Isolation of Tammar PITX1, TBX4, TBX5, FGF8, and SHH

Primers for PITX1, TBX4, TBX5, FGF8, and SHH were designed based on the NCBI tammar genome (Table 3). Total RNA was extracted from tammar fetal limbs ranging between day 21 fetus and day 26 and cDNA synthesis was performed according to the manufacturer's protocol (Invitrogen, Thornton, NSW, Australia). RT-PCR for each stage (n=3) was performed using GoTaq Green Master Mix (Promega, Madison, WI) under the following conditions: 33 cycles of 30 s, 95°C; 30s, 60°C; 60 s, 72°C, in a 25-μl reaction. The subsequent product was confirmed to be the target gene of interest via gel extraction (Qiagen, Valencia, CA) and sequencing (Melbourne University Diagnostic and Sequencing Unit). The RNA control was a pooled template before reverse transcription into cDNA of forelimb or hindlimb.

Table 3. Primers Used During This Study
 Forward primerReverse primer
TBX5CCGGAATTCGGCTGAAGTTCCACGAAGTGGCCCAAGCTTCCAGTGGGTATGG
PITX1CAGTGAGGACGGAGGTGGTGAAACTTGGCTCGGCGGTTCTTGA
TBX4GTGGCAGGAAAGGCAGAGCCAGTTCTCCACAGTCCC
SHHCTGCTGGTATGCTCGGGGACTCAGGAGCCAGGTGCCTATTTG
FGF8GAGCAGAGCCTGGTACrGATCCYTTGCGSGTRAAAGCCCATGTACC

Probe Construction and Whole-Mount In Situ Hybridisation

Tammar gene–specific primers were designed and used to amplify fragments by polymerase chain reaction (PCR) from tammar cDNA at day-23, day-24, or day-25 fetal forelimb or hindlimb. PCR products were gel extracted (Qiagen) and cloned into pGEM-T easy vector (Promega) and sequenced to ensure specificity of target gene fragments and orientation of the insert. Template used for generation of digoxigenen-labelled RNA probes (Roche Diagnostics, Indianapolis, IN) was amplified by PCR using the plasmid as template (diluted 1:250) and a combination of M13 forward or M13 reverse sequencing primer (vector specific) with the corresponding gene-specific primer and synthesized using T7 and SP6 polymerase (Promega). Whole mount in situ hybridisation on tammar embryos for all genes (Table 2) was performed as described in Chew et al. (2012). However, FGF8 in situ staining was performed as described in Keyte and Smith (2010). Photographs of staining from the resulting whole mount in situ hybridisations were taken using an Olympus DP25 camera mounted on an Olympus (ZX-9) dissection microscope with either a black felt or clear background. Whole-mount images are a montage of several images with 15% overlap merged together.

Scanning Electron Microscopy and Histology

Tammar specimens for SEM were dehydrated through ethanol series to 100% ethanol, dried in a critical point dryer, and mounted on their dorsal side using carbon rods. Gold sputter coating followed and images were taken using a Phillips XL30 Field-emission Scanning Electron Microscope. Paraformaldehyde fixed specimens for histology were initially stored in 70% ethanol before they were embedded in paraffin wax. Serial sections were cut at 6 μm and mounted on polylysine slides (Polyesine, Menzel Glaser, Braunschweig, Germany) and stained with haematoxylin and eosin before imaging. Images were taken using an Olympus DP25.

ACKNOWLEDGMENTS

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. ACKNOWLEDGMENTS
  7. REFERENCES

We thank Dr. Kathleen Smith and Dr. Anna Keyte for their invaluable advice on whole-mount in situ hybridisation and for fruitful discussions on marsupial embryos. We also thank members of the marsupial research team for their assistance in animal handling and maintenance.

REFERENCES

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. ACKNOWLEDGMENTS
  7. REFERENCES