SEARCH

SEARCH BY CITATION

Keywords:

  • Sonic hedgehog;
  • limb development;
  • digit patterning

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES
  8. Supporting Information

Sonic hedgehog (Shh) controls the number and type of digits formed. Using a conditional genetic approach for timed removal of Shh, we previously proposed a biphasic model of Shh function: a transient patterning phase, during which digit progenitors are specified, and an extended proliferative phase, during which expansion of progenitor pools enables digit formation. Other models favor a close integration of digit patterning and expansion, with sequential promotion to more posterior identity over time, apparently supported by some mutants with selective posterior digit loss. To further test these models, we analyzed the dynamics of Shh activity in several oligodactylous mutants with different types of digit loss. The profile of Shh activity and phenotypic outcome in these mutants supports a biphasic over an integrated temporal model. Eomesodermin expression, as an independent marker of posterior digit identity, confirmed that proper digit 4 specification requires only the transient phase of Shh activity. Developmental Dynamics 240:1303–1310, 2011. © 2011 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES
  8. Supporting Information

Sonic hedgehog (Shh) is expressed in the posterior margin of the early limb bud called the zone of polarizing activity (ZPA), and signals to regulate the different types and numbers of digits formed (from anterior-to-posterior, or AP, thumb to pinky, digit 1–5). Based on extensive, predominantly gain of function work in chick, there is strong support for the concept that Shh regulates digit pattern in a dose-dependent manner (highest doses producing the most posterior identity), consistent with its action as a “morphogen” directing AP digit pattern by means of a spatial gradient (rev. by Zeller,2004; Towers and Tickle,2009). Genetic lineage tracing studies have revealed that ZPA descendants contribute substantially to the digits formed, including all of digits 4, 5 and part of digit 3 (Harfe et al.,2004), leading to an expansion-based temporal model, in which Shh activity is integrated over time and cells remaining within the Shh-producing ZPA for the longest time become most posterior (Harfe et al.,2004; Scherz et al.,2007). In a recent variation of this model, based on pharmacologic manipulation of Shh signaling and independent modulation of growth in the chick limb bud, the dual roles of Shh are integrated, so that digit number expansion occurs jointly with digit specification and “promotion” to more posterior identities over time by Shh exposure (Towers et al.,2008). All models for Shh function in patterning digits share, as a common feature, the key element that the highest cumulative Shh activity (either dose or duration) is required for specifying the most posterior digit (digit 5).

Using a conditional genetic approach in mice (ShhFl/−; Hoxb6Cre-ER) to remove Shh at different defined times during the course of early limb development, we previously found that digit loss following Shh removal at different times occurred in the order digit 3 first, then digit 5, digit 2, and lastly digit 4, which did not fit with the expected posterior–anterior loss predicted by Shh morphogen function (Zhu et al.,2008). However, in the mouse, digit condensations normally first appear in an alternating AP sequence starting with posterior digit 4, then digit 2, digit 5, and lastly digit 3, which is precisely the reverse order of digit loss observed following Shh removal. Based on these results, together with the large body of work implicating Shh as a morphogen in digit patterning, a biphasic model of Shh function was proposed in which Shh acts early and transiently as a morphogen to specify and pattern digit progenitors (Phase 1), and acts over an extended period to expand the progenitors and enable realization of the final digits (Phase 2). Consequently, posterior digits would only be the most sensitive to perturbed Shh activity in Phase 1. However, the latest forming digit (3) would be the most sensitive to removal of Shh activity after the end of the specification phase, as seen with the conditional genetic ablation of Shh (Zhu et al.,2008) and similarly reported following the late attenuation of Shh expression by genetic removal of Fgfr1 (Verheyden et al.,2005). An alternate explanation for the order of digit loss observed in mouse has been raised based on the growth-promotion model proposed for chick limb development (Towers and Tickle,2009). As applied to mouse, this model is reconciled with the observed alternating AP sequence of digit appearance by proposing that there are two zones of promotion operating in parallel, a posterior zone including digits 4 and 5 and an anterior zone including digits 2 and 3. In this scenario, digit 4 forms and is promoted to digit 5; digit 2 forms and is promoted to digit 3.

There are at least two different classes of oligodactylous mutants with different types of digit loss resulting from modulated Shh activity. In one class, Shh reduction is associated with loss of posterior digits, which is entirely consistent with classical morphogen gradient or temporal integration models (e.g., Parr and McMahon,1995; Lewis et al.,2001). In another class, mutants have a distinct phenotype with selective loss of either digit 2 or 3 (e.g., Verheyden et al.,2005; Scherz et al.,2007; Zhu et al.,2008). Because the identity based on morphology has been disputed and there is as yet no definitive marker for distinguishing digit 2 vs. 3 identity, we refer to this class as “central” digit loss, as a less committal designation. Although Scherz and coworkers (2007) suggested that the “central” digit lost in the ShhFl/ShhCre mutant is digit 2 based on fate mapping showing that Shh+ descendants contribute to each of the remaining three triphalangeal digits in the mutant, they also noted that the Shh expression domain expands anteriorly in the mutant in response to loss of negative feedback regulation at early stages. Consequently, it is possible that some of the digits marked by Shh expression in this mutant are not necessarily the same as those normally arising (digit 3–5) from Shh+ descendants.

To test the different models for Shh function, the dynamics of Shh activity were compared in these mutant classes. Two separate markers (Tbx2 and Eomesodermin) were used to ascertain the presence of posterior digits, particularly digit 4. Overall, the profiles of Shh activity compared with digit phenotype in both classes of mutant are most compatible with the biphasic model, confirming that posterior digit identities can be specified even in the context of severe oligodactyly as long as the transient patterning phase of Shh activity is intact, and furthermore, that digit 5 specification depends on high Shh activity during the transient early phase, rather than the cumulative level of activity over time. Our results also identify Eomesodermin as a potentially useful second marker for assessing posterior digit identity.

RESULTS AND DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES
  8. Supporting Information

To determine how Shh activity is altered in oligodactylous mutants with different orders of digit loss and whether or not these are consistent with a biphasic model of Shh function, we selected three different mouse mutants with reduced Shh expression and different skeletal phenotypes: loss of posterior digits (Wnt7a−/−, and ShhFl/−;PrxCre); or loss of “central” digits (digit 2 or 3, ShhFl/ShhCre) similar to that occurring when the duration of Shh expression is shortened conditionally (ShhFl/−;Hoxb6CreER). Wnt7a has been shown to regulate Shh expression in both gain and loss of function models in chick and mouse (Yang and Niswander,1995; Parr and McMahon,1995). Because the Wnt7a−/− mutant has a much more highly penetrant phenotype of digit loss in the forelimb (Parr and McMahon,1995; and our unpublished observations, approximately 65%), and PrxCre recombination in the early hindlimb bud is mosaic (Logan et al.,2002), we focused our analysis and comparisons on the forelimb buds for each mutant analyzed. We re-assessed the skeletal phenotypes at embryonic day (E) 17.5 in the genetic backgrounds used in our crosses. Wnt7a−/− mutants typically had 4 digits, with digit 5 being invariably absent (Fig. 1B; 65%), and the remainder were predominantly pentadactylous. Rarely, only three digits formed in Wnt7a−/− mutants (<10%; the additional missing digit was likely digit 4 based on morphology, but the low frequency precluded further analysis to confirm this). ShhFl/ShhCre mutants usually had four digits, with a “central” digit other than digit 5 absent (Fig. 1C), as previously reported (Scherz et al.,2007). Occasionally (∼20%), ShhFl/ShhCre mutants with 3 digits were observed, in which case digit 5 was also absent. However, ShhFl/−;PrxCre embryos very consistently had a more severe phenotype than reported previously (Lewis et al.,2001), and in most mutants, only two digits formed with a single long, triphalangeal digit; very rarely three digits were observed (<5%), with an apparent transformation resulting in two biphalangeal digits and one triphalangeal digit (Fig. 1F). These mutants fall into two different classes with varying degrees of digit loss: “central” digit loss first (ShhFl/ShhCre), or posterior digit loss first (mild, Wnt7a−/−; or severe with transformations, ShhFl/−;PrxCre).

thumbnail image

Figure 1. Skeletal phenotypes and Sonic hedgehog (Shh) activity (Ptch1 expression) in several mutants with altered Shh expression and oligodactyly. A–F: Forelimb skeletal phenotypes in embryonic day (E) 16.5–E17.5 embryos of wild-type control (A), Wnt7a−/− (B), ShhFl/ShhCre (C), ShhFl/−; Hoxb6CreER following Tamoxifen (Tam) injection at E10 or at E9.5 (D, E, respectively), and ShhFl/−;PrxCre (F). Digit identity assignments were made based on previously described criteria (Zhu et al.,2008). Uncertain or disputed assignments (see e.g., Lewis et al.,2001; Scherz et al.,2007) are labeled with question marks. A′–F′″: Ptch1 expression in the different mutants near the onset (E9.5, A′–F′), after approximately 6–8 hr (E9.75, A″–F″), and at later stages of Shh activity (E10.5, A′″–F′″). Whereas Shh activity begins at near normal levels in ShhFl/ShhCre and then declines (C′–C′″), it is initially very low in Wnt7a−/− but recovers rapidly to near control levels (B′–B′″). Note that because of the timing of Tam treatment (E10 or E9.5) of ShhFl/−;Hoxb6CreER embryos, the embryos are essentially wild-type in Ptch1 expression at E9.5 and E9.75 (pretreatment).

Download figure to PowerPoint

We evaluated Shh activity during early (specification/patterning) and late (expansion) phases in each of these mutants, by monitoring the expression of the direct Shh target Ptch1 as a reporter for Shh activity. Interestingly, Ptch1 expression was very weak early near the onset of Shh expression in Wnt7a−/− embryos when the forelimb bud has just formed (3/4 at E9.5, Fig. 1B′), but subsequently recovered rapidly (within 6–8 hr) and was comparable to Ptch1 expression in control embryos (4/4 at E9.75, Fig. 1B″), apparently by means of homeostatic regulatory mechanisms governing Shh activity (see for e.g., Scherz et al.,2007; Bastida et al.,2009; Bouldin et al.,2010). At a later stage, E10.5, Ptch1 expression in Wnt7a−/− mutants ranged from equivalent to (3/5) or slightly lower than (2/5, Fig. 1B′″) control embryos. In contrast, Ptch1 expression in ShhFl/ShhCre embryos was similar to control embryos near the onset of Shh expression (3/3 at E9.5, Fig. 1C′), but subsequently the expression level became progressively reduced (3/3 at E9.75, 4/4 at E10.5, Fig. 1C″–C′″) as previously reported for E10.5 (Scherz et al.,2007). This expression profile was more comparable to the rapid loss of Ptch1 expression following conditional Cre activation with Tamoxifen (Tam) in ShhFl/−; Hoxb6CreER embryos (2 embryos analyzed at each stage, Fig. 1D′″, E″–E′″; see also Zhu et al.,2008). In contrast, Ptch1 expression was markedly reduced at very early as well as later stages in ShhFl/−; PrxCre embryos (2/2 at E9.5, 3/3 at E9.75, 3/3 at E10.5, Fig. 1F′–F′″), in agreement with previous results for E10.5 (Lewis et al.,2001; Scherz et al.,2007).

Because limb bud morphology was also altered in some mutants (in particular Wnt7a−/− embryos had smaller limb buds at very early stages; eg. Fig. 1B′) and in situ hybridization only gives a qualitative estimate of relative expression levels, we also examined Ptch1 RNA expression by real-time quantitative polymerase chain reaction (qPCR) analysis of control and mutant forelimb bud RNAs. The results graphed in Figure 2A (see also Supp. Table S1, which is available online), confirm that in Wnt7a−/− embryos, Ptch1 expression was initially very low (16% of control at E9.5), but recovered to near control levels by E9.75 (77%), presumably due to homeostatic regulatory mechanisms as discussed above. Of interest, at the later stage, Ptch1 expression was reduced compared with controls (32% at E10.5), consistent with decreased Shh expression previously reported (Parr and McMahon,1995), and suggesting that the duration of Shh expression/activity could also be affected to some degree in this mutant. In contrast, in ShhFl/ShhCre embryos Ptch1 expression was reasonably high initially (55% of control at E9.5), but became progressively reduced over time (28% of control at E9.75; 18% of control at E10.5). As can be seen from the Shh activity curves over time (Fig. 2B), Wnt7a−/− and ShhFl/ShhCre mutants (both of which form 4 digits) each have moderately reduced average Shh activity between E9.5 and E10.5 compared with wild-type controls, yet Wnt7a−/− mutants do not specify digit 5, but ShhFl/ShhCre mutants do. The primary difference between the two mutants is the much higher level of Shh activity present in ShhFl/ShhCre limb buds initially (E9.5). Even though the Wnt7a−/− limb has a considerably higher level of activity than ShhFl/ShhCre by E9.75, this is apparently too late for digit 5 specification to occur. Likewise, the duration of moderate Shh activity is actually longer (to E10.5) in Wnt7a−/− compared with ShhFl/ShhCre, yet digit 5 is specified in ShhFl/ShhCre but not Wnt7a−/− mutants. Scherz et al. (2007) reported consistent loss of a phalanx in digit 5 of ShhFl/ShhCre mutants, suggesting altered digit 5 specification in this mutant. However, we observed this only inconsistently and predominantly in younger embryos, and believe that delayed differentiation in ShhFl/ShhCre mutants may obscure identification of late forming, and poorly stained phalanges (see for example, Supp. Fig. S1). In ShhFl/−; PrxCre embryos, Ptch1 expression was markedly reduced compared with controls at all stages evaluated, which was consistent with its severe oligodactyly phenotype and loss of posterior digits.

thumbnail image

Figure 2. Quantitation of Sonic hedgehog (Shh) activity (Ptch1 RNA expression) in mutant forelimb buds at different times during early stages of limb development. A: Bar graphs of Ptch1 RNA levels in different mutant forelimb buds at embryonic day (E) 9.5 (25 somites), E9.75 (28 somites) and E10.5 (38 somites) compared with stage-matched sibling controls (raw data shown in Supp. Table S1). Error bars represent 2 standard deviations from the mean. P values were determined from 2-tailed t-test with ***indicating significance at P < 0.01, and *at P < 0.05 relative to controls for the same stage. B: Schematic curves of Shh activity profiles for different mutants over time compared with wild-type, and with the abrupt loss of Shh in conditional mutant [ShhFl/−;Hoxb6CreER] following Tamoxifen treatment at E10 (dashed line). Flat dashed line shows approximate background Ptch1 level in Shh−/− limb buds (estimated from Supp. Table S1). Note that both Wnt7a−/− and ShhFl/ShhCre mutants have attenuated Shh activity at different times, and both mutants develop four digits. The profiles differ primarily in that for the Wnt7a−/− mutant, where digit 5 is lost, the most marked reduction in Shh activity occurs very early and transiently, whereas for ShhFl/ShhCre, where a central digit is lost, activity is reduced mainly at later times.

Download figure to PowerPoint

The observed Shh activity profiles and phenotypic outcome of digit loss in these mutants are compatible with proposed biphasic roles for Shh, required transiently at an early stage for digit specification-patterning (Phase 1) and over an extended time for expansion of specified digit progenitors (Phase 2). In Wnt7a−/− mutants, loss of a single, posterior-most digit 5 occurs, following a very transient reduction in Shh activity during Phase 1 alone. In ShhFl/ShhCre mutants, loss of a single “central” digit occurs, following reduction of Shh activity during the expansion Phase 2, similar to the phenotype in ShhFl/−; Hoxb6CreER mutants, where the duration of Shh expression is shortened slightly by late activation of Cre with Tamoxifen. In ShhFl/−; PrxCre, a severe phenotype including both posterior digit loss and accompanying anterior transformations follows the marked reduction of Shh activity during both Phase 1 and Phase 2.

A key point that the biphasic model hinges on is whether a posterior digit (4) is still specified properly following Shh removal shortly after the onset of its activity. Because the middle digits, digit 2, 3, and 4 have rather similar morphologies in mouse, we also evaluated markers to confirm whether posterior digit identity was altered in the different mutants. Unfortunately, good markers selective for digit 2 have yet to be identified. However the presence of digit 4 can be evaluated using Tbx2, the late expression of which extends over condensations of digits 4 and 5 in mouse (Fig. 3A; Zhu et al.,2008). Note that, although there is also an anterior expression domain of Tbx2 in the limb bud, this is restricted to the zeugopod region, and does not extend over digit 1. Consequently, expression over distal digit condensations is restricted to digits 4 and 5 in wild-type embryos. Tbx2 has also been implicated in regulating posterior digit identity in the chick hindlimb, where its expression encompasses only the posterior-most digit 4 (Suzuki et al.,2004), suggesting that Tbx2 expression marks digits 4 and 5 selectively, rather than limb bud “width”. In Wnt7a−/− mutants, digit 5 is clearly absent but the remaining digits appear normal morphologically. As expected, Tbx2 expression was present in, and limited to, the now posterior-most remaining digit 4 in Wnt7a−/− embryos (4/4, Fig. 3B). In mutants with shortened Shh activity duration in Phase 2 alone, in which a “central” digit is lost first, Tbx2 expression was present in the two remaining posterior-most digits, as anticipated (ShhFl/−;Hoxb6CreER following Tam injection at E10, 6/6, Fig. 3C; and ShhFl/ShhCre, 4/4, data not shown). As shown previously, following the removal of Shh activity very early during Phase 2 (ShhFl/−;Hoxb6CreER, Tam injection at E9.5, 2/2, Fig. 3D; Zhu et al.,2008), the single elongated (triphalangeal) digit formed still retained Tbx2 expression, consistent with digit 4 identity. In contrast, ShhFl/−;PrxCre embryos did not have any expression of Tbx2 over the single long digit condensation formed, indicating that neither digit 4 nor 5 were specified (4/4, Fig. 3E).

thumbnail image

Figure 3. Tbx2 expression in mutants with altered Sonic hedgehog (Shh) activity and oligodactyly. A–E: Tbx2 RNA expression in embryonic day (E) 11.75 forelimb buds of wild-type control (A), Wnt7a−/− (B), ShhFl/−;Hoxb6CreER Tam treated at E10 (C), or at E9.5 (D), and ShhFl/−;PrxCre (E) embryos. Only the embryos shown in panels A, and B were also stained for Noggin/LacZ expression before in situ hybridization, to demonstrate the Tbx2 expression pattern relative to the different digit condensations.

Download figure to PowerPoint

It has been argued that Tbx2 may not be an accurate marker of digit identity (see Tabin and McMahon,2008). Although Tbx2 expression invariably correlated with expected digit identity based on morphologic and other criteria in the several mutants we have examined, this does not prove that it is invariably a reliable reporter of digit identity. Therefore, we evaluated Eomesodermin (Eomes) as a second marker for digit 4. Previous work has shown that Eomes is expressed selectively at the base of digit 4 by E12.5 and at later stages in wild-type embryos (Hancock et al.,1999; Kwon and Hadjantonakis,2007; see also Fig. 4A and Supp. Fig. S2A,A′). Eomes could be a valuable complementary marker for digit 4, as its expression occurs at a later stage when morphogenesis is already under way (possibly in tendon/ligament precursors surrounding digit 4) and its expression is likely regulated very differently from that of Tbx2. Because Eomes expression overlaps and is obscured by concomitant staining for digit condensations, unstained embryo limb buds at late E12.5 were analyzed, when condensation numbers are clearly evident by trans-illumination. As a control, we examined Eomes expression in Wnt7a−/− mutant embryos, which have reduced Shh activity but still specify digit 4. Specific expression of Eomes over the base of digit 4 was maintained in the Wnt7a−/− mutant, as anticipated (5/5, Fig. 4B and Supp. Fig. S2B,B′). We also evaluated Eomes expression in ShhFl/−;Hoxb6CreER embryos with mild (Tam E10) to severe (Tam E9.5) phenotypes, which were previously interpreted to retain normal digit 4 specification (Zhu et al.,2008). In embryos with mild (four digit condensations; 4/4, Fig. 4C, Supp. Fig. S2C, C′), moderate (three digit condensations; 4/4, Fig. 4D, Supp. Fig. S2D, D′), and severe (single long digit condensation; 1/1, Fig. 4E) phenotypes, specific Eomes expression was still apparent at the base of one condensation. In contrast, in ShhFl/−;PrxCre embryos, in which Shh activity was markedly reduced from its onset, Eomes expression was absent at the base of the single remaining long digit condensation (4/4, Fig. 4F). These results identify Eomes as a potentially useful marker for digit identity, and confirm that even posterior digits (digit 4), require only transient Shh exposure for specification (short Phase 1 duration). Because digit 4 progenitors expand first during Phase 2, a short total duration of Shh activity still enables formation of digit 4 as long as Shh activity during Phase 1 remains intact (see Fig. 5).

thumbnail image

Figure 4. Eomesodermin expression in mutants with altered Sonic hedgehog (Shh) activity and oligodactyly. A–F: Eomesodermin RNA (Eomes) expression in embryonic day (E) 12.5 forelimb buds of wild-type control (A), Wnt7a−/− (B), ShhFl/−;Hoxb6CreER Tam treated at E10 (C), at E9.75 (D), or at E9.5 (E), and in ShhFl/−;PrxCre (F) embryos. Digit condensations at positions shown were clearly visible by trans-illumination. Putative digit identities were assigned to condensations that formed in each mutant based on skeletal morphologies at E17.5 and Tbx2 expression patterns (see Zhu et al.,2008 and Figs. 1, 3); note that mutants proposed to retain digit 4, including ShhFl/−;Hoxb6CreER after very early Tam treatment, form a digit condensation with expression of Eomes at its base (corresponding to digit 4 in wild-type embryos).

Download figure to PowerPoint

thumbnail image

Figure 5. Biphasic model of Sonic hedgehog (Shh) function during limb development, and proposed digit phenotypes generated by altered Shh activity. Shh functions as a morphogen very early and transiently during Phase 1, and as a mitogen to expand digit progenitors over an extended period during Phase 2. Color-coded numbers in the left represent early digit progenitors specified by Shh during Phase 1 (digit 2 is specified first by Shh signaling). Color-coded numbers in the right represent the final realized digits. Intermediate stages show order of expansion of progenitors in the sequence digit 4, 2, 5, 3, to ultimately form digit condensations (Zhu et al.,2008). See text for discussion of phase-specific alterations of Shh activity and digit phenotypes in mutants.

Download figure to PowerPoint

These results indicate that both the initial specification and patterning (or promotion) of digits by Shh occur during a very early and transient time frame (Phase 1 in Fig. 5). In contrast, other recent models based on Shh functional analysis in the chick (Scherz et al.,2007, Towers et al.,2008) propose that digits are specified and promoted to more “posterior” identities over time (integrated with expansion). Digit transformations (to a more anterior identity) are infrequently observed in mouse mutants with altered Shh activity, compared with chick, where such transformations invariably accompany digit loss as Shh activity is reduced. This has been attributed to difficulties in distinguishing digits 2–4 in the mouse. However, when the quantity (ShhFl/−;PrxCre), or efficacy (e.g., Shh-N, Lewis et al.,2001; Li et al.,2006) of Shh are modulated in the very early limb bud, transformations to biphalangeal digits do become apparent, suggesting that in fact the time frame for producing anterior transformations is very compressed in the mouse. Aside from the issue of what duration is required for the integrated specification and promotion to occur, other features of a promotion model in mouse are difficult to fit with experimental observations. To incorporate the observed alternating sequence of digit appearance in the wild-type mouse embryo, it has been suggested that expansion-promotion of digit 2 to 3, and of digit 4 to 5, occur in parallel in anterior and posterior zones of the limb bud (Towers and Tickle,2009). This division into zones is difficult to reconcile with cell tracing studies of Shh descendants, which indicate that digit 3 derives in part from both of these zones (Harfe et al.,2004). One would also expect that the temporal requirements for the integrated specification-promotion in each of the two zones should be invariant. Yet experimentally, Shh reduction phenotypes with “central” digit loss require that this time frame is longer for digit 2–3 than for digit 4–5, whereas those with posterior digit loss would require the time frame to be longer for digit 4–5.

For assessing the time window required for patterning by Shh, the differences between Wnt7a−/− and ShhFl/ShhCre limbs are particularly revealing. In Wnt7a−/− mutants, the profile of Shh activity during Phase 1 is briefly altered, but recovers very rapidly (Figs. 1B′,B″, 2A,B) and only the most posterior digit 5 is affected. Conversely, in ShhFl/ShhCre mutants, Shh activity begins to decrease soon after expression begins (Figs. 1C′,C″, 2A,B), but digit 5 is still specified correctly (as discussed above, see also Supp. Fig. S1). This short time window agrees very well with that previously estimated from the conditional genetic removal of Shh (Zhu et al.,2008). Comparing Shh activity in all of the mutants in Phase 1 vs. Phase 2 (E9.5 vs. E10.5 in Figs. 1, 2), it appears that the quantitative level of Shh signaling present in Phase 2 is less critical for accomplishing its mitogenic function in progenitor expansion, whereas correct specification and patterning in Phase 1 is much more sensitive to the timing of onset and the quantitative level of Shh activity. A high degree of sensitivity to dose may reflect Shh behavior as a morphogen during a very early and transient time window. Such transient, presteady state morphogen effects operating predominantly when a signaling gradient is first being established have been proposed previously, and experimentally validated for Hedgehog signaling in the drosophila wing disc (Nahmad and Stathopoulos,2009; Saunders and Howard,2009).

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES
  8. Supporting Information

Mouse Mutant Strains

The Wnt7a−/− mutant (Parr and McMahon, 1995), ShhFl/Fl mutant (Lewis et al.,2001), Shh−/− mutant (Chiang et al.,1996), NogginLacZ knock-in (Brunet et al.,1998), Hoxb6CreER (Nguyen et al.,2009), ShhCre (Harfe et al.,2004), and Prx1Cre (Logan et al.,2002) mouse lines were all described previously. All alleles were maintained on a predominantly FVB/n background, and expression profiles shown are typical for this background. Noon on the date of the vaginal plug was defined as E0.5. Tamoxifen treatment (single IP dose of 3 mg), embryo collection, and analyses were described previously (Nakamura et al.,2006).

Embryo Analysis

Procedures for whole mount in situ hybridization, LacZ staining, skeletal staining and criteria used for assignment of digit identities were all as previously described (Zhu et al.,2008). For valid comparisons, all Ptch1 hybridizations of controls and different mutant embryos for a particular stage were performed and developed together in a single vial with embryos tagged for identification.

qPCR Analysis of Ptch1 RNA Levels

Mouse forelimb buds were dissected from single control or mutant embryos at the stages (somite number) indicated (see Supp. Table S1), and left and right forelimb buds were pooled for analysis. RNA isolation was performed using RNAqueous-Micro kit from Ambion (catalog no. AM1931). Reverse transcription was carried out using SuperScript III First-Strand kit from Invitrogen (catalog no. 18080-051). qPCR was performed on a Bio-Rad iCycler iQ detection system with Bio-Rad SsoFast EvaGreen Supermix (catalog no. 172-5201) using a three-step amplification (94°C 10 sec, 55°C 10 sec, 72°C 30 sec) for 49 cycles. Exon primers used (each flanking introns) were as follows:

Ptch1-F: CTGGCAGCCGAGACAAGCCC, Ptch1-R: TGGCCTGGGAGGCAGCGTAA, β-Actin-F: CCCTAAGGCCAACCGTGAA, β-Actin-R: CAGCCTGGATGGCTACGTACA.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES
  8. Supporting Information

We thank Naiche Adler and Kat Hadjantonakis for suggesting the use of Eomesodermin as a potential digit identity marker, and Heinz Arnheiter for Ptch1, Benoit Bruneau for Tbx2, and Janet Rossant for Eomesodermin (TBr2) probes. We are grateful to Heiner Westphal and Chin Chiang for providing the Shh−/− mouse line, and to Andy McMahon and Cliff Tabin for making genetically engineered lines generated in their labs (Wnt7a−/−, ShhFl/Fl, NogginLacZ, PrxCre, ShhCre) broadly available through the Jackson Laboratory. This research was supported by the Center for Cancer Research, National Cancer Institute, NIH.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES
  8. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
DVDY_22637_sm_SuppFig1.tif2334KSupporting Information Figure 1. Skeletal staining shows that digit 5 is correctly specified in ShhFl/ShhCre embryos. Embryos were harvested at embryonic day (E) 17.5. Note that three phalanges are present in digit 5 in ShhFl/ShhCre embryos, although cartilage staining is somewhat delayed and irregular compared with sibling controls. mc, metacarpals; mt, metatarsals; P1, the first phalanx; P2, the second phalanx; P3, the third phalanx.
DVDY_22637_sm_SuppFig2.tif2761KSupporting Information Figure 2. Transected limb buds demonstrate Eomesodermin expression around the base of digit 4 condensation. A–D: As in Figure 4, Eomesodermin RNA (Eomes) expression is shown in embryonic day (E) 12.5 forelimb buds of wild-type control (A,A′), Wnt7a−/− (B,B′), and ShhFl/−;Hoxb6CreER Tam treated embryos at E10 (C,C′), or at E9.75 (D,D′). Digit condensations at positions shown were clearly visible by trans-illumination. Putative digit identities were assigned to condensations that formed in each mutant based on skeletal morphologies at E17.5 and Tbx2 expression patterns (see Zhu et al., 2008 and Figs. 1, 3). Transection of hybridized limb buds (A′–D′) shows that in mutants proposed to retain digit 4, this digit condensation is surrounded at the base by cells expressing Eomes (similarly to digit 4 in wild-type embryos).
DVDY_22637_sm_SuppTable1.doc119KSupporting Information Table 1.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.