Wing buds whose posterior half is excised, develop into wings lacking distal structures. However, such experimentally generated preaxial half wing buds can be rescued by implanting a retinoic-acid-releasing bead at their anterior margin. The polarity of the pattern that originates from preaxial half wing buds is reversed. For example, instead of a 234 digit pattern typical for normal wings, the order of digits is 432. This result implies that retinoic acid has the capacity to reprogram anterior limb bud tissue, and that the resulting change in cell fate does not depend on the presence of posterior tissue regions such as the zone of polarizing activity (ZPA).

Posterior limb bud mesenchyme, known as the zone of polarizing activity (ZPA), has a remarkable effect on the anteroposterior pattern of vertebrate limbs. When a block of ZPA tissue is grafted to an anterior site in a limb bud, the graft will induce a mirror-image duplication pattern (Saunders and Gasseling, 1968; Tickle et al. 1975). One way to explain such pattern duplications is to assume that the grafted and the endogenous ZPA release a morphogen that forms a U-shaped concentration gradient across the limb rudiment which provides limb bud cells with positional information (Wolpert, 1969; Tickle etal. 1975; for an alternative view see e.g. Bryant and Muneoka, 1986).

A promising candidate for the hypothetical ZPA morphogen is all-trans-retinoic acid, because it closely mimics ZPA tissue grafting (Tickle et al. 1982; Summerbell, 1983). Specifically, when a ion-exchange bead impregnated in retinoic acid is implanted at the anterior wing bud margin, that is opposite the ZPA (see Fig. 1, left part), a mirror-image duplication pattern will develop that is morphologically indistinguishable from one generated by ZPA grafting. Both ZPA and retinoic acid cause duplications in a dose-, time-, stage-, and position-dependent manner (Tickle et al. 1982, 1985; Summerbell, 1983; Eichele et al. 1985, reviewed in Eichele, 1989). Furthermore, retinoic acid is endogenously present in limb buds, and it is enriched in the ZPA-containing posterior tissue (Thaller and Eichele, 1987). Limb bud cells contain both nuclear receptors for retinoic acid (Smith and Eichele, in preparation) and as cellular retinoic-acid-binding protein (Maden and Summerbell, 1986; Maden et al. 1988). Taken together, these findings make a good case that retinoic acid or a structurally close congener is the postulated morphogen of the ZPA.

Fig. 1.

Scheme of the rescue experiment. A 250μm diameter bead impregnated in retinoic acid was implanted in ovo at the anterior wing bud margin underneath the apical ectodermal ridge (AER) opposite the zone of polarizing activity (ZPA). 1–2h later, the posterior half of the bud was removed as indicated in the scheme on the right, and the embryos were returned to the incubator. In control experiments the bead was either omitted or soaked in DMSO only (solvenrfor retinoic acid). The prospective tissue regions for humerus (h), radius (r), ulna (u), digit 2 (2), digit 3 (3) and digit 4 (4) are indicated.

Fig. 1.

Scheme of the rescue experiment. A 250μm diameter bead impregnated in retinoic acid was implanted in ovo at the anterior wing bud margin underneath the apical ectodermal ridge (AER) opposite the zone of polarizing activity (ZPA). 1–2h later, the posterior half of the bud was removed as indicated in the scheme on the right, and the embryos were returned to the incubator. In control experiments the bead was either omitted or soaked in DMSO only (solvenrfor retinoic acid). The prospective tissue regions for humerus (h), radius (r), ulna (u), digit 2 (2), digit 3 (3) and digit 4 (4) are indicated.

The unsatisfactory aspect of implanting a retinoic-acid-releasing bead into an intact wing bud is that such buds also contain the endogenous ZPA (see Fig. 1, left). Consequently, the experiment does not reveal whether ZPA tissue must be present for retinoic acid to induce pattern duplications. For example, the ZPA might provide growth factors (Cooke and Summerbell, 1980) in whose absence retinoic acid is not effective. In other words, retinoic acid might be necessary but not sufficient to respecify the pattern. One way to clarify this issue is to excise the posterior half of the wing bud including the ZPA and to apply retinoic acid to the remaining anterior half bud (see Fig. 1, right). It has previously been shown that preaxial half wing buds develop poorly (e.g. Hinchliffe and Gumpel-Pinot, 1980). However, if retinoic acid is supplied, a significant number of anterior halves will develop into an essentially normal extremity. This observation suggests that retinoic acid is not only necessary but also sufficient to specify the pattern developing from preaxial half limb buds.

AG1-X2 ion exchange beads were presoaked in either 12·5 or 20 μgml−1 all-riww-retinoic acid dissolved in DMSO and washed in PBS (Eichele and Thaller, 1987). This routinely gives 43234 duplications and results in concentrations of applied retinoic acid in the limb bud close to that of endogenous retinoic acid (Thaller and Eichele, 1987). A bead was implanted underneath the apical ectodermal ridge at the anterior (Fig. 1) or at the posterior wing bud margin (opposite somite border 19/20) of a Hamburger-Hamilton stage 20 embryo (Hamburger and Hamilton, 1951). The bead was allowed to attach to the tissue by incubating the embryos for 1 to 2h at 37·5°C. Note, treatment of whole limb buds with retinoids for such a short period will never lead to a duplication (Eichele et al. 1985). This mitigates against the objection that, during the time span preceding amputation, posterior tissue could have permanently influenced the anterior tissue. Using sharpened watchmaker forceps, either the posterior or the anterior half portion of the limb was removed by a first incision across the limb bud followed by a cut at the base of the bud parallel to the body wall. Control wing buds had either no bead implant or carried a plain bead that was presoaked in DMSO only. After the operation, the embryos were returned to the incubator. Greater than 80% of the embryos survived to day 10 or later, when they were fixed, stained and cleared in the usual way.

Wing pattern restoration by retinoic acid in ZPA-free anterior half limb buds

Detailed mapping studies of ZPA activity in wing buds by Honig and Summerbell (1985) showed that at stage 20/21 ZPA extends almost along the entire posterior wing bud margin. To be certain that all ZPA is removed, the complete posterior wing bud half was excised (see Fig. 1). Without further treatment, this always resulted in truncated wings consisting of humerus and radius only (first column in Table 1; for a normal pattern see Fig. 2 A). This result agrees with earlier findings by Hinchliffe and Gumpel-Pinot (1981) and is essentially consistent with the fate map of Hinchliffe et al. (1981) that places at least part of the humerus, the ulna and digits 3 and 4 into the excised posterior half of the wing bud (Fig. 1). Control experiments in which a plain bead was implanted into anterior half buds also yielded truncated wings that consist of humerus and radius or a partial radius (Fig. 2 B and second column in Table 1). We conclude that the bead by itself has no effect on pattern formation. Moreover, the formation of such severely truncated limbs in both controls implies that the ZPA has indeed been completely removed. Had some ZPA tissue remained, at least a partial digit pattern should have developed (see e.g. Hinchliffe et al. 1981).

Table 1.

Patterns that developed in anterior half wing buds

Patterns that developed in anterior half wing buds
Patterns that developed in anterior half wing buds
Fig. 2.

Patterns that developed following various types of treatment. (A) Normal wing. (B) Wing that received a plain bead at the anterior margin and whose posterior half was subsequently removed. (C) A wing that received a retinoic acid releasing bead but did not show distal pattern restoration. (D) Example of a wing that formed a ‘generic’ digit following retinoic acid treatment. Abbreviations: h, humerus; r, radius; u, ulna; z, zeugopodal element; 2, 3, 4, digits; d, digit.

Fig. 2.

Patterns that developed following various types of treatment. (A) Normal wing. (B) Wing that received a plain bead at the anterior margin and whose posterior half was subsequently removed. (C) A wing that received a retinoic acid releasing bead but did not show distal pattern restoration. (D) Example of a wing that formed a ‘generic’ digit following retinoic acid treatment. Abbreviations: h, humerus; r, radius; u, ulna; z, zeugopodal element; 2, 3, 4, digits; d, digit.

A very different result is obtained if the implanted bead is impregnated with retinoic acid. In 30% of the cases, an essentially complete extremity was restored (Figs 3 and 4). Specifically, wings that developed following retinoic acid treatment, fell into three groups (last column in Table 1). The first group consisted of 63 wings (46 %, first line in Table 1) that had a humerus and typically one radius-like forearm element (Fig. 2 C), a pattern very close to that seen in the controls (Fig. 2 B). Sometimes the single radius-like forearm element was either shortened, broadened, bent or missing. Such abnormal forearms were never seen in the control wings, hence they presumably reflect a minor effect of retinoic acid on the limb bud tissue. The second group, represented by 32 wings (23%, second line in Table 1), had a digit which often resembled digit2 (Fig. 2 D). We interpret such limbs as a partial restoration of the pattern. The most interesting group, comprising 40 wings (31 %, line 3 to 5 in Table 1), had essentially restored hand plates. 12 wings had a completely restored hand with all three digits present (Fig. 3), 23 wings had digits 3 and 4 (Fig. 4), and the remaining 6 wings had either just a digit 4, a digit 3, or a3 and 2 (not shown).

Fig. 3.

Wing that developed from an preaxial half bud treated with retinoic acid. (A) Skeletal pattern illustrating the presence of humerus and digits 2, 3 and 4. (B) Camera lucida drawing of the same specimen. Note the reversed polarity of the hand plate. Digit 4 is most anterior and digit 2 most posterior (for a normal pattern see Fig. 2 A). (C) Exterior structures of the same specimen. Note the presence of long feather filaments (arrow) at the anterior margin of the hand.

Fig. 3.

Wing that developed from an preaxial half bud treated with retinoic acid. (A) Skeletal pattern illustrating the presence of humerus and digits 2, 3 and 4. (B) Camera lucida drawing of the same specimen. Note the reversed polarity of the hand plate. Digit 4 is most anterior and digit 2 most posterior (for a normal pattern see Fig. 2 A). (C) Exterior structures of the same specimen. Note the presence of long feather filaments (arrow) at the anterior margin of the hand.

Fig. 4.

Photograph (A) and camera lucida drawing (B) of a retinoic acid restored anterior half wing bud that resulted in a wing with digits 3 and 4. Note the splitting of the zeugopod (z).

Fig. 4.

Photograph (A) and camera lucida drawing (B) of a retinoic acid restored anterior half wing bud that resulted in a wing with digits 3 and 4. Note the splitting of the zeugopod (z).

A notable feature of the specimens rescued by retinoic acid is that a restoration of the pattern is accompanied by a reversal of the anteroposterior polarity of the hand plate (Figs 3 and,4). In the retinoic-acid-treated bud, the most anterior digit is a digit 4 and not a digit 2 as is the case in a normal wing (compare Figs 2 A and 3 B). The feather anlagen are reversed as well. Normally, the longest feather papillae are located at the posterior wing margin. By contrast, in the restored limb the longest feather anlagen are situated at the anterior wing margin (Fig. 3 C, arrow). In the normal chick wing there are considerably more feather germs dorsally than ventrally. This dorsoventral asymmetry is maintained in a rescued wing (not shown). Thus rescued limbs have a truly inverted anteroposterior axis and do not result from a 180° rotation around the proximodistal axis, an operation that would, of course, also have reversed the anteroposterior polarity.

The polarity reversal of distal structures of restored wings demonstrates that the hand pattern is specified by the anteriorly located retinoic-acid-releasing bead. Because the proximally located humerus has the same polarity as in the contralateral wings, its axial polarity must have been specified by the posteriorly located ZPA. Hence, there must have been a switch in axial polarity, happening while the forearm elements are specified (stage 21, see Saunders, 1948 and Summerbell, 1974). Evidence for this being the case is provided by the frequently observed dismorphic forearms that often consist of several short cartilage elements, or of a bifurcated piece of cartilage (Figs 3 and,4). Rescued wings are consistently shorter than the contralateral wings. While the humerus of the treated wing is virtually unaffected and amounts to 97±5% of that of the untreated left wing, the length of e.g. the proximal phalange of digit 3 in retinoic-acid-rescued limbs is only 59±14% of that of the same phalange in the contralateral extremity. A similar reduction in length of distal structures was observed by Wilson and Hinchliffe (1987) when rescuing anterior half buds with ZPA tissue grafts (see Discussion).

Application of retinoic acid to posterior half wind buds has no effect on the pattern

The amputation of the anterior half of the wing bud results in wings that usually lack radius and digit 2. To determine whether retinoic acid influences this outcome, beads soaked in 12.5 gml-1 retinoic acid were implanted underneath the apical ectodermal ridge at the posterior side (opposite somite border 19/20) of the wing bud of a stage 20 embryo. Subsequently, the anterior half of the bud was removed as described for the experiments above, and the embryo was left to develop. The wing patterns obtained were as follows. 13 out of 14 cases were lacking the radius. 6 had a full set of digits while 8 were without digit 2. We conclude from this that at the dose applied, retinoic acid has no marked effect on the pattern that develops from postaxial half wing buds.

Retinoic acid when discharged into limb buds will result in pattern duplications along the anteroposterior axis. For example, instead of a normal digit pattern (234, Fig. 2A), a 432234 pattern is generated that has an additional set of digits. This outcome indicates that retinoic acid is involved in the specification of the pattern of additional digits along this axis. However, it is not clear from this type of experiment at what level in the patterning process retinoic acid functions. It could be a very early cue in axis specification, but it could also be operative at a relatively late stage or even just serve as a permissive agent, e.g. one that induces ZPA (Tickle et al. 1985). One reason why the conventional experiment using whole limb buds cannot address this issue, is that the anterior tissue in which the duplicate set of structures arises is connected to posterior tissue that gives rise to the normal set of digits (see fate map in Fig. 1). Linkage allows communication between the two parts and hence postaxial tissue might influence preaxial tissue in a critical way. To avoid this complication, we have excised the posterior half of the bud. It has previously been shown that the resulting anterior half wing bud undergoes extensive cell death soon after the removal of the postaxial half (Hinchliffe and Gumpel-Pinot, 1981). Hence, not surprisingly, preaxial half buds will develop poorly (Hinchliffe and Gumpel-Pinot, 1981; Hinchliffe et al. 1981 and first and second column in Table 1 in this report). We found that treatment with retinoic acid can reprogram preaxial half buds so that they quite often develop into a wing with a complete set of distal structures (Figs 3 and 4). Because this switch in cell fate and the ensuing development occurs in the absence of posterior tissue, we argue that posterior tissue is not required for retinoic acid to exert its rescue function. Thus, it appears that retinoic acid is necessary but also sufficient to specify the wing pattern along the anteroposterior axis in anterior half wing buds. Moreover, since reprogramming of cell fate is a prerequisite for all subsequent morphogenetic processes, it is likely that retinoic acid acts early in patterning. The caveat in this line of reasoning is that posterior tissue might have critically influenced the anterior region prior to the experimental manipulations, and that this influence persists.

Saunders and Gasseling (1968) have reported that ZPA grafted to an anterior site in preaxial half buds leads without fail to complete wings that have a reversed anteroposterior polarity. In fact, the wing shown in Fig. 3 looks exactly like the specimen shown in their study. In a recent study Wilson and Hinchliffe (1987) have examined the development of preaxial half buds to which a piece of ZPA was attached at the distal end of the cut face. Such buds develop into a virtually normal wing in nearly half of the cases (the polarity of the resulting wings is obviously the same as in normal limbs). The chief conclusion is that providing preaxial half wing buds with either locally applied retinoic acid or with ZPA grafts, results in a fair number of cases in a complete limb.

Wing buds lacking the preaxial half result in almost normal wings. We found that implanting a retinoic-acid-releasing bead at the posterior margin of such posterior half limb buds does not affect this outcome. This is reminiscent of the earlier observation that local application of retinoic acid from beads implanted at the posterior margin (opposite somite 19/20) of whole limb buds has little effect on the pattern, except at high doses, where digit 2 will be absent (Tickle et al. 1985). It is possible that posterior cells are already committed at Hamburger-Hamilton stage 20 (the stage when the experiments are conducted) and an exogenously provided stimulus given too late, can therefore no longer influence cell fate. This explanation can be tested by delivering retinoic acid at earlier stages.

Obviously, one is tempted to conclude from the rescue experiment described here that retinoic acid also is the primary inducer for formation of the normal pattern that arises from posterior tissue (see fate map Fig. 1). The fact that retinoic acid is endogenously present in limb buds (Thaller and Eichele, 1987) tends to go along with this extrapolation. However, some degree of caution is appropriate. As long as one has not demonstrated that physiological doses of retinoic acid will alter the normal pattern, one cannot be absolutely certain that the insights gained from experiments such as the one described in this report give an appropriate picture of normal development. Efforts in this direction are underway in our laboratory.

I thank C. Thaller and Dr S. E. Wedden for critical reading of the manuscript. This study was supported by grant HD 20209 from the National Institutes of Health and in part by a gift from the Lucille P. Markey Charitable Trust.

Bryant
,
S. V.
and
Muneoka
,
K.
(
1986
).
Views of limb development and regeneration
.
Trends Genet
.
2
,
153
159
.
Cooke
,
J.
and
Summerbell
,
D.
(
1980
).
Growth control early in embryonic development; the cell cycle during experimental pattern duplication in the chick wing
.
Nature, Lond
.
287
,
697
701
.
Eichele
,
G.
(
1989
).
Retinoids and vertebrate limb pattern formation
.
Trends. Genet
.
5
,
246
251
.
Eichele
,
G.
and
Thaller
,
C.
(
1987
).
Characterization of concentration gradients of a morphogenetically active retinoid in the chick limb bud
.
J. Cell Biol
.
105
,
1917
1923
.
Eichele
,
G.
,
Tickle
,
C.
and
Alberts
,
B. M.
(
1985
).
Studies on the mechanism of retinoid-induced pattern duplications in the early limb bud: temporal and spatial aspects
.
J. Cell Biol
.
101
,
1917
1923
.
Hamburger
,
V.
and
Hamilton
,
H.
(
1951
).
A series of normal stages in the development of the chick embryo
.
J. Morph
.
88
,
49
92
.
Hinchliffe
,
J. R.
,
Garcia-Porrero
,
J. A.
and
Gumpel-Pinot
,
M.
(
1981
).
The role of the zone of polarizing activity in controlling the differentiation of the apical mesenchyme of the chick wing bud: histochemical techniques in the analysis of a developmental problem
.
Histochem. J
.
13
,
643
658
.
Hinchliffe
,
J. R.
and
Gumpel-Pinot
,
M.
(
1981
).
Control of maintenance and anteroposterior skeletal differentiation of the anterior mesenchyme of the chick wing bud by its posterior margin (the ZPA)
.
J. Embryol. exp. Morp
.
62
,
63
82
.
Honig
,
L. S.
and
Summerbell
,
D.
(
1985
).
Maps of the strength of positional signalling activity in the developing chick wing bud
.
J. Embryol. exp. Morph
.
87
,
169
175
.
Maden
,
M.
,
Ong
,
D. E.
,
Summerbell
,
D.
and
Chytil
,
F.
(
1988
).
Spatial distribution of cellular protein binding to retinoic acid in the chick limb bud
.
Nature, Lond
.
335
,
733
735
.
Maden
,
M.
and
Summerbell
,
D.
(
1986
).
Retinoic acid-binding protein in the chick limb bud: identification at developmental stages and binding affinities of various retinoids
.
J. Embryol. exp. Morph
.
97
,
239
250
.
Saunders
,
J. W.
and
Gasseling
,
M. T.
(
1968
).
Ectodermal-mesenchymal interactions in the origins of limb symmetry
.
In Epithelial-Mesenchymal Interactions
(ed.
R.
Fleischmajer
and
R. E.
Billingham
), pp.
78
97
.
Baltimore
:
Williams and Wilkins
.
Saunders
,
J. W.
, Jr
. (
1948
).
The proximo-distal sequence of origin of the parts of the chick wing and the role of the ectoderm
.
J. exp. Zool
.
108
,
363
403
.
Summerbell
,
D.
(
1983
).
The effects of local application of retinoic acid to the anterior margin of the developing chick limb
.
J. Embryol. Exp. Morph
.
78
,
269
289
.
Summerbell
,
D.
(
1974
).
A quantitative analysis of the effect of excision of the AER from the chick limb bud
.
J. Embryol. exp. Morph
.
32
,
651
660
.
Thaller
,
C.
and
Eichele
,
G.
(
1987
).
Identification and spatial distribution of retinoids in the developing chick limb bud
.
Nature, Lond
.
327
,
625
628
.
Tickle
,
C.
,
Lee
,
J.
and
Eichele
,
G.
(
1985
).
A quantitative analysis of the effect of all-trans-retinoic acid on the pattern of chick wing development
.
Devi Biol
.
109
,
82
95
.
Tickle
,
C.
,
Alberts
,
B. M.
,
Wolpert
,
L.
and
Lee
,
J.
(
1982
).
Local application of retinoic acid to the limb bud mimics the action of the polarizing region
.
Nature, Lond
.
296
,
564
565
.
Tickle
,
C.
,
Summerbell
,
D.
and
Wolpert
,
L.
(
1975
).
Positional signalling and specification of digits in chick limb morphogenesis
.
Nature, Lond
.
254
,
199
202
.
Wilson
,
D. J.
and
Hinchliffe
,
J. R.
(
1987
).
The effect of the zone of polarizing activity (ZPA) on the anterior half of the chick wing bud
.
Development
99
,
99
108
.
Wolpert
,
L.
(
1969
).
Positional information and the spatial pattern of cellular differentiation
.
J. theor. Biol
.
25
,
1
47
.