Cell Differentiation and Development, 32 (1990) 17-26

Elsevier Scientific Publishers Ireland, Ltd. 17

CELDIF 00692

Synthetic retinoids, retinobenzoic acids, Am80, Am580 and Ch55 regulate morphogenesis in chick limb bud

Koji Tamura ‘, Hiroyuki Kagechika 2, Yuichi Hashimoto 2, Koichi Shudo 2, Kojune Ohsugi 1

and Hiroyuki Ide ’

’ Biological Institute, Tohoku University, Sendai and ’ Faculty of Pharmaceutical Sciences, University of Tokyo, Tokyo, Japan

(Accepted 28 June 1990)

The retinobenzoic acids Am80, Am580 and Ch55 are synthetic stable analogs of retinoic acid @A), and show very strong differentiation-inducing activity in human myelogeneous leukemia cell line HL&O. To examine the effects of these synthetic retinoids on limb pattern formation, AGl-X2 beads containing these retinoids were applied to the anterior margin of stage 19-20 chick wing buds. By implanting the beads with 1 pg /ml retinoids, normal wings were formed and extra digits 2 or 32 were rarely formed. As the retinoid concentrations increased from 10 pg/ml to 100 pg/ml, duplicatedTmbs 3234 43234 432234 4334 were -‘LL_ progressively produced. At higher concentrations, 1 mg/ml, the wings often truncated, although duplication occurred in some embryos. These synthetic analogs seem to have the same degree of morphogenetic potential as RA, since the activity index of these retioids was similar to that of RA. Since these synthetic retinoids hardly bind to CRABP (cellular retinoic acid-binding protein), it may be possible that the retinoids and RA may affect limb-pattern formation without the interaction with CRABP.

It is known that limb buds cannot develop distal structures when the posterior region including all ZPA (zone of polarizing activity) is removed. When beads containing the above mentioned retioids were implanted to the anterior margin of wing buds from which the posterior one third region including all ZPA had been removed, distal growth of the wing buds and the formation of digit elements were observed. Some of the wing buds produced a completely reverse digit pattern 432. From these results, we discussed the roles - of RA in limb development and pattern formation.

Synthetic retinoids; Retinobenzoic acid; Morphogenesis; Chicken

Introduction

Pattern formation is one of the most interesting problems in developmental biology and has been studied in various systems. In the developing chick

Correspondence address: K. Tamura, Biological Institute, Tohoku University, Aoba-Yama, Sendai 980, Japan.

limb bud, it is well known that a mirror-image symmetrical digit pattern develops after grafting tissue from the posterior margin (ZPA, zone of polarizing activity) (Saunders and Gasseling, 1968). Limb pattern formation can be explained by the positional information theory (Wolpert, 1969). According to this theory, each cell is in- formed on its position, by getting ‘positional val- ues’ from the concentration of morphogen around

0922-3371/90/$03.50 0 1990 Elsevier Scientific Publishers Ireland, Ltd.

18

it, which are produced in ZPA and forms a con- centration gradient across the antero-posterior axis of the limb bud, after which the cells differentiate into the appropriate tissues decided by the ‘posi- tional values’. In many attempts to identify the nature of morphogen, all-truns-retinoic acid (RA) has been found to be a candidate for ZPA factor, since it mimics the ZPA activity (Tickle et al., 1982). RA exists 2.5-times more abundantly in the posterior one-third part, which comprises a quarter of the total limb bud mass, than in the remaining part (Thaller and Eichele, 1987). We have reported the distribution of retinoids containing RA in chick limb bud with monoclonal antibody (Tamura et al., 1990). But little is known about the detailed roles that RA plays in the formation of the limb pattern.

RA induces cell differentiation in some em- bryonal carcinoma cells, in which RA is consid- ered to affect at a transcriptional level (Roberts and Sporn, 1984). Some intracellular binding pro- teins of RA (CRABP, cellular retinoic acid bind- ing protein; Chytil and Ong, 1984) and RA recep- tors in the nuclei (RAR, retinoic acid receptor; Giguere et al., 1987; Petkovich et al., 1987; Hashimoto et al., 1988) have been reported. In order to understand the function of RA at the gene level, it is important to know which part these binding proteins of RA play a role. Chick limb buds have CRABP and the distribution of CRABP has been recently reported (Maden et al., 1988). But the roles of CRABP in limb pattern formation remains uncertain. R4R gene is also expressed in the developing chick limb (Noji et al., 1989) but it is not quite clear which genes are regulated by RAR in the limb bud cells.

The retinobenzoic acids Am80, Am580 and Ch55 are synthetic analogs of RA and potent inducers of HL-60 cell differentiation into mature granulocytes (Kagechika et al., 1984; Shudo et al., 1985). These retinoids bind to RAR strongly (K, = 10” M-‘) (Hashimoto et al., 1988) but bind less or hardly at all to CRABP (Jetten et al., 1987; Takagi et al., 1988). Therefore, we attempted to induce duplicate formation with these three reti- noids to examine whether R4 needs CRABP for the determination of the A-P axis in the limb morphogenesis.

Grafting of ZPA into the anterior margin causes an increase in the width of the host limb bud by stimulating DNA synthesis in the distal meso- dermal cells (Cooke and Summerbell, 1980; Sum- merbell, 1981). On the other hand, removal of the posterior ZPA-containing region results in the de- letion or reduction of the distal structure by the failure of sufficient outgrowth (Summerbell, 1979; Hinchliffe and Gumpel-Pinot, 1981). These results suggest that ZPA forms a growth-promoting sig- nal as well as a pattern-controlling signal. Though RA mimics these functions of ZPA on duplicate formation, there is no evidence to deny the possi- bility that RA enhances the ZPA signals above normal level in the anterior region. So, in this study it was also examined whether retinobenzoic acid could replace ZPA both in the promotion of growth and in the function of pattern control by applying retinoids to the anterior margin of wing buds from which the posterior one-third region containing all ZPA had been removed.

Materials and Methods

Chemicals All-truns-retinoic acid (Lot No. 98F-0778) was

obtained from Sigma. Am80, Am580 and Ch55 (Fig. 1) were prepared by the method described previously (Kagechika et al., 1984; Shudo et al., 1985).

Implantation of bead containing retinoidrr into limb tissue

The beads were soaked in retinoid solution and implanted according to Eichele et al. (1984). Briefly, AGl-X2 beads (Bio-Rad) of 150-200 pm in diameter were soaked in retinoids dissolved in dimethyl sulfoxide (DMSO) at various concentra- tions (1, 10, 100 pg/ml and 1 mg/ml) for 1 h at room temperature and then rinsed in a drop of the culture medium (Dulbecco’s modified Eagle’s MEM plus 10% fetal calf serum). The beads were never soaked in the culture medium for more than 1 h to prevent the retinoids from pouring out of the beads. Implants were placed opposite somite 16-17 into a slit made beneath the apical ecto-

19

dermal ridge of stage 19-20 chick wing bud (Hamburger and Hamilton, 1951).

In the ZPA-removal experiment, about one third of the posterior region of the stage 19-20 chick wing bud was removed (Fig. 4) with shar- pened tungsten needles, and beads containing reti- noids were implanted into the anterior margin.

TABLE I

Duplicate formation following the application of synthetic retinoids a,b

Digit pattern Concentration of retinoids (mg/ml)

0.001 0.01 0.1 1

normal 2234 32 (2) 34 432 (2) 34 4334 434 34 4 No digits

CM5 normal 2234 32 (2) 34 432 (2) 34 4334 434 34 4 No digits

Am580 normal 2234 32 (2) 34 432 (2) 34 4334 434 34 4 No digits

16 2 4 -

1 - 6 - 5 - 2 2

1 _ 12

3 - 2 - 4 1 2 5 _ 1

1 2

1 3

Fixation and staining of limb tissue After the operation, the embryos were in-

cubated for a further 7 days at 38” C. Then the embryos were dessected out, fixed overnight in 10% formalin in Tyrode’s solution, stained (6 h at 37°C) with 0.1% Alcian blue in 70% ethanol with 0.1 N HCl, dehydrated and cleared in methyl salicylate.

ReSUlts

Induction of duplicate formation with synthetic reti- noids

Normal wing bud developed the digit pattern 234 from anterior to posterior after implanting the - beads soaked in DMSO only (Fig. 2a). When the

RA

Am580

Am80 c Implantation was made at stage 20. Skeletal pattern was examined 7 days after the operation.

beads soaked in 100 pg/ml of R4 were implanted to the anterior margin of the wing buds, typical duplications with digit pattern 432234 were formed (Fig. 2b) as reported previomickle et al., 1985). Table I summarizes the results of experi- ments in which beads had been soaked in various concentrations of Am80, Am580 and Ch55, re- spectively, and then implanted. With low con- centrations of retinoids, only extra digit 2 was produced (Fig. 2c; 1 pg/ml Ch55). As the retinoid concentration increased, digit patterns 3234 (Fig. 2d; 10 pg/ml Am580) and 432234,43234 (Fig. 2e; --

kcooH Ch55

0

Fig. 1. Structure of RA (a), Am80 (b), Am580 (c) and Ch55 (d).

- d

Fig. 2. Skeletal pattern of wings developed from the wing buds into which beads soaked in various concentrations of retinoids were implanted. (a) Normal wing with digit pattern 234; bead was soaked in DMSO. (b) Typical duplicated wing with digit pattern 432234; bead was soaked in 100 ug/ml RA. (c) %ng with one additional digit 2234; bead was soaked in 1 ag/ml Ch55. (d) Digit

7 pattern 32234; bead was soaked in 10 pg/ml Am580. (e) Wing with symmetrical drgt pattern 43234; bead was soaked in 100 pg/ml Am80. (f)ng with short digit 4; 1 mg/ml Am580. (g) T nmcated wing containing humerus and a knob of cartilage: 1 mg/ml

Am80. Bars = 1 mm.

21

100 pg/ml Am80) were observed. In some cases, the radius diverged to the middle. At the highest dose, 1 mg/ml, the digit area was reduced, to 34 or 4, and sometimes digitless wings were forma (Fig. If; 1 mg/ml AmgO), although duplications 4334 and 434 were also observed. Beak defects were oftenqserved at this concentration. At 1 mg/ml of Ch55, all the treated embryos died. From these results we estimated the activity index of these retinoids after Summerbell (1983) to de- fine the strength of effect of these retinoids on pattern formation. The degrees of duplication in Table 1 were classified into seven groups and a score, Si, was assigned to each group as follows: 234, Si = 0; 2234, Si = 1; 3[22]34, Si = 2; 4[322]34, Si = 3; 234, Si = 4; 234, Si = 5; 234, Si = 6 (where the digits withinsquare bra&% were of variable incidence, and those crossed out were missing). The activity index (AI) was then

ml)

Fig. 3. Activity index (AI) of retinoids for the duplicate forma-

tion, which were calculated as described in the text. The concentrations of retinoids applied into anterior margin of wing bud are shown in abscissa. AI of retinoic acid was calculated on the basis of the data of Eichele et al. (1985). AI

at 1 mg/ml of Ch55 is not shown since all the treated embryos died.

cont.

Fig. 4. A diagram of ZPA-removal experiment. Stage 19-20 wing buds were cut from the center of apex to the posterior

one-fourth point of the proximal line, and about one-third of

the posterior regions (striped parts) had been removed. (a) Control embryos without retinoids. (b) Beads soaked in 10, 100

ag/ml and 1 mg/ml of retinoid solution were applied into the

anterior margin beneath the apical ectodermal ridge of the

wing buds from which the posterior one-third region had been

removed.

calculated as:

Si * 100% AI= c 4n

n

According to this index, 0% shows that normal wing is formed in every case, 100% shows forma- tion of an extra digit 4 in every case, and 200% indicates that truncation of digit 4 occurs in every case. Fig. 3 is a schematic representation of AI. All synthetic retinoids used in this study showed as high AI as RA, and AI increased dose-depen- dently in all cases. (The data about RA were drawn from the study of Tickle et al. (1985)).

Production of reverse digit pattern by synthetic reti- noids

Since it was found that above three the reti- noids had the same potency of duplication as PA, we examined the effects of these retinoids on ZPA-free limb buds. When the posterior one-third region of stage 19-20 wing bud was removed, as shown in Fig. 4, limb bud developed zero, one (humerus, Fig. 5al) or two skeletal elements (humerus and radius, fig. 5a2), and in all cases no digit formation was observed (see Table II).

b

Fig. 5. Skeletal pattern of wings showing the effects of retinoids on the ZPA-free limb buds. (al) One control wing; only humerus is developed. (a2) Other control wing; two skeletal elements (humerus and radius) are formed. (b) ZPA-free wing received a bead soaked in 100 pg/ml Am80; the digit pattern is completely reversed, 432. (c) ZPA-free wing treated with 100 ag/ml Am80; digit pattern is incompletely reversed, 43. (d) Treated with 10 pg/ml Am80;wo elements of digits 22 are formed. (e) Treated with 10

- ag/ml Am80; digit 2 is formed. Bars = 1 mm.

After the implantation of beads treated with 100 pg/ml Am80 into the ZPA-free wing buds, some elongated wings developed, in which carti- lage elements were twisted forward at the middle of the radius and ulna, and a completely reversed digit pattern, 432, was formed (Fig. 5b). Others twisted backward at the elbow joints and formed digits 43 from anterior to posterior (Fig. 5~). 100 rg/ml<f Am580, Ch55 and RA also produced similar structures (not shown). These reversed wings were smaller than normal wings with regard to distance from elbow to the apex of the digit

(about 60-708 of the normal length, compare Fig. 5b and c with Fig. 2a). One specimen treated with 10 pg/rnl of Am80 is shown in Fig. 5d. In this case, one element of zeugopod twisted forward in an extreme manner when compared to the control and two bifurcated elements of digits 22 had been formed. One digit element was observed at the apex of a zeugopod element in some other sample treated with 10 pg/ml AmSO, and the digit pat- tern seemed to be 2 (Fig. 5e). The total length of the wings developed from the retinoid-treated ZPA-free buds was greater than that of non-treated

23

TABLE II

Skeletal pattern of ZPA-free wing following implants of the beads soaked in retinoids

Skeletal pattern Cont. Concentration of retinoids (ag/mI)

Am580 Ch55 Am80 100 100 10 100 500

No element 6 (20.0) 1 (5.6) 2 (10.5) 3 (12.5) 2 (8.7) 9 (50.0) One element a 13 (43.3) 7 (38.8) 8 (42.1) 8 (33.3) 4 (17.4) 5 (27.8) Two elements b 11 (36.7) 5 (27.8) 2 (10.5) 7 (29.2) 7 (30.4) 1 (5.5) Unidentified digit ’ 2 (11.1) 1 (5.3) 2 (8.3) 4 (17.4) 3 (16.7) digit 2 or 22 ’ 4 (16.7) digits 43 c 1 (5.6) 3 (15.8) 2 (8.7) digits 432 ’ 2 (11.1) 3 (15.8) 4 (17.4)

Numbers in parentheses refer to percentage for total number. a Humerus. b Humerus + radius. ’ Humerus + zeugopodium(s) + digit(s).

ZPA-free buds. At 500 pg/ml of Am80, a limb rarely grew out distal and formed only undefined small cartilage elements.

When wing buds were treated with 100 pg/rnl of Arn580, Ch55 and Am80, distal outgrowth and digit formation were observed in 27.8, 36.9 and 43.5% wings, respectively (Table II). Furthermore, completely reversed digit pattern 432 developed in 11.1, 15.8 and 17.4% of Am580, ch55 and Am80, respectively. Some embryos (25.0%) treated with 10 pg/ml of Am80 showed progressive skeletal development with digit structure, and 16.7% of embryos had one or two elements with a joint, which were identified as digit 2. When the wing buds were treated with 500 pg/ml of Am80, 16.7% of the specimen elongated into the distal direction and undefined digit elements were formed.

Discussion

Evaluation of effects of three synthetic retinoitis Am80, Am580 and Ch55 were able to produce

mirror-image duplicates as RA. These effects oc- curred dose-dependently at concentrations be- tween 1 pg/ml and 1 mg/ml, and the effective doses were almost similar to that of RA (Tickle et al., 1985). Additional digits formed by the treat- ment with these retinoids at low doses were 2 or 2. At 100 pg/ml, a complete set of duplication

432234 was frequently observed. Treatment with a higher dose, 1 mg/ml, induced digit defects to various degrees. The activity index of these three retinoids, which was estimated from the effects on pattern formation, increased dose-dependently and the dose-response curves, was similar to that of RA (Fig. 3). These results show that the three synthetic retinoids and RA seem to have the potency to affect limb pattern formation to the same extent.

These synthetic retinoids induced terminal dif- ferentiation of HL-60 cell and their activities were lo-lOO-fold higher than that of RA (Kagechika et al., 1984; Shudo et al,, 1985). However, in our experiments they are effective at similar doses to RA. This discrepancy may be due to the dif- ference between in vitro and in vivo experiments, since these analogs affect the differentiation of dissociated distal limb mesenchymal cells in vitro (Ide and Aono, 1988) about lOO-fold more strongly than RA (data not shown). These retinoids are more hydrophilic than RA, thus the release of retinoid may be different to that of RA from beads in vivo.

Maden and Summerbell (1986) have identified CRABP in the developing chick limb bud and further reported the distribution pattern in the limb bud region (Maden et al., 1988). They have proposed a hypothesis, in which CRABP may reduce the effective concentration of RA that re-

24

aches the nucleus, and amplify the gradient of the RA concentration across the anterior-posterior axis (Maden et al., 1988).

The retinoids used in this study bind to CRABP less efficiently than RA; in particular, the affini- ties of the strongest Ch55 for CRABP is estimated to be less than l/2000 of that of RA (Jetten et al., 1987; Takagi et al., 1988). Nevertheless, these reti- noids could mimic the ZPA activity in duplicate formation. These results suggest that these reti- noids can affect the limb pattern formation and hence the possibility can be considered that RA may function in normal limb morphogenesis without interaction with CRABP. These retinoids also induced limb outgrowth in the ZPA-free wing bud with the occasional formation of completely reversed digits. Thus, RA may play a role only not in pattern formation but also in the distal growth of limb bud, as will be mentioned below. CRABP might not be concerned with the cell proliferation involved in RA-induced distal outgrowth.

Correlation between ZPA activity and RA We have found that retinoids by themselves can

replace the ZPA in the ZPA-free wing bud. Very recently, similar results with RA have been re- ported by Eichele (1989). For the explanation of these results, we propose some interpretations of the function of RA in limb morphogenesis.

(1) Cell growth. After grafting ZPA into the anterior margin of limb bud, the host limb bud increased in width along the antero-posterior axis with the stimulation of DNA synthesis in the mesoderm (Cooke and Summerbell, 1980). In the culture system, posterior fragments containing ZPA showed stronger growth-promoting effects on pre-axial cells than the anterior fragment (Aono and Ide, 1988). When the posterior region includ- ing all ZPA was removed, the remaining part developed poorly and the digit structures were deleted (Hinchliffe and Gumpel-Pinot, 1981).

The AER (apical ectodermal ridge) is also known to be essential for limb growth. If the AER is removed at each stage, limb bud fails to develop further and distal components yet to be laid down are not formed (Summerbell, 1974). The AER is not maintained autonomously, but maintained by mesodermal factor(s), AER maintenance factor

(AERMF). MacCabe and Parker (1975) have sug- gested that AERMF may be secreted from ZPA. Growth-promoting activity of ZPA might be equal to the AER-maintenance activity.

These results suggest that ZPA possesses a growth-promoting signal which functions directly or indirectly. In this study the fact that retinoids that have RA activity could regulate limb growth without ZPA suggests that RA participates in distal outgrowth of limb buds. Then, how does RA control cell growth? RA itself may be a cell growth-promoting signal released from ZPA. It has been shown that RA promotes proliferation directly in the distal mesodermal cells of chick limb but in vitro (Ide and Aono, 1988). However, it is also possible that some molecules whose pro- duction is regulated by RA promote cell growth indirectly; that is, RA may convert anterior mesodermal cells to ZPA which secretes growth- promoting signals as described above. At any rate, one of the genes regulated by RA in the limb bud may be concerned with a growth factor or a sub- stance which controls growth-promoting cascade. Alternatively, RA may operate as the trigger of AERMF production. As regards the AER, Tickle et al. (1989) have shown recently that RA applica- tion leads to reorganization of the AER, which must be mediated by irreversible changes in the mesoderm. This suggests that the gene products which are formed under the regulation of RA in the limb mesenchymal cells may control the AER function.

(2) Pattern formation. About the determination of antero-posterior axis in the limb bud, which is the other function of ZPA, Wolpert (1969, 1971) has proposed a model in which a gradient of diffusible substance, morphogen, released from ZPA provides cells with recognition of their posi- tion (positional value) and then, each cell subse- quently differentiates according to the positional value. Though RA mimics this activity of dupli- cate formation, there is the possibility that RA simply enhances the ZPA function by changing the pattern-controlling signal released by ZPA into its active form. Our result that RA can determine the antero-posterior axis of the limb bud under the condition that there is no ZPA signal suggests

25

that RA does not merely modify the activity of the ZPA signal in the anterior region but participates in the normal pattern-determination mechanisms of limb bud. In ZPA-free wing buds, RA applica- tion into the anterior side could induce digit pat- tern 432 at a high dose (100 pg/ml) and 22 or 2 at a lowdose (10 pg/rnl) (Fig. 5). We canexplain these results by the presence of a RA gradient along the antero-posterior axis. The other possibil- ity, however, cannot be excluded, which is that RA converts cells of the anterior margin into ZPA cells, and that the new ZPA provide the distal cells with positional values. This idea is supported by the study that tissue adjacent to the RA-carrier implanted into anterior margin could form the duplication when implanted into another host limb (Summerbell and Harvey, 1983; Tickle et al, 1985). Our data can be interpreted also by this idea: Dose-dependent digit formation induced by these retinoids may be due to the number of cells con- verted into ZPA cells by RA at various concentra- tions.

Our results lead to the conclusion that RA may actually regulate limb morphogenesis. However, it is uncertain whether RA acts in the endogenous limb field directly, since we have observed only the effects of IZA exogenously added and we have little knowledge with regard to the endogenous RA in the limb bud, except for regional difference of RA concentrations (Thaller and Eichele, 1987). It will be necessary to examine the changes of limb pattern when the endogenous activity of RA in the limb bud is completely blocked.

Acknowledgments

This work was supported in part by grants from the Ministry of Education, Science, and Cul- ture, Japan, and the Mitsubishi Foundation and the Naito Foundation.

References

Aono, H. and H. Ide: A gradient of responsiveness to the growth-promoting activity of ZPA (zone of polarizing activ- ity) in the chick limb bud. Dev. Biol. 128, 136-141 (1988).

Bell, K.M. (1986) The preliminary characterization of mitogens secreted by embryonic chick wing bud tissues in vitro. J. Embryol. Exp. Morphol. 93, 257-265.

Chytil, F. and D.E. Ong: Cellular retinoid-binding proteins. In: The Retinoids, eds, M.B. Spom, A.B. Roberts and D.S. Goodman, Vol. 2 (Academic Press, New York) pp. 89-123 (1984).

Cooke, .I. and D. Summerbell: Cell cycle and experimental pattern duplication in the chick wing during embryonic development. Nature (Lond.) 287, 697-701 (1980).

Eichele, G.: Retinoic acid induces a pattern of digits in ante- rior half wing buds that lack the zone of polarizing activity. Development 107, 863-867 (1989).

Eichele, G., C. Tickle and B.M. Arberts: Microcontrolled re- lease of biologically active compounds in chick embryos: beads of 200~Pm diameter for the local release of retinoids. Anal. B&hem. 142, 542-555 (1984).

Giguere, V., E.S. Ong, P. Segui and R.M. Evans: Identification of a receptor for the morphogen retinoic acid. Nature (Lond.) 330, 642-629 (1987).

Hamburger, V. and H. Hamilton: A series of normal stage in the development of chick embryo. J. Morphol. 88, 49-92 (1951).

Hashimoto, Y., H. Kageehika, E. Kawachi and R. Shudo: Specific uptake of retinoids into human promyelocytic leukemia cells HL-60 by retinoid-specific binding protein: possibly the true retinoid receptor. Jpn. J. Cancer Res. (Gann) 79,473-483 (1988).

Hinchliffe, J.R. and M. Gumpel-Pinot: Control ‘of mainte- nance and anteroposterior skeletal differentiation of the anterior mesenchyme of the chick wing bud by its posterior margin (the ZPA). J. Embryol. Exp. Morphol. 62, 63-82 (1981).

Ide, H. and H. Aono: Retinoic acid promotes proliferation and chondrogenesis in the distal mesodermal cells of chick limb bud. Dev. Biol. 130, 767-773 (1988).

Jetten, A.M., K. Anderson, M.A. Deas, H. Kagechika, R. Lotan, J.I. Rearick and K. Shudo: New benzoic acid de- rivatives with retinoid activity: lack of direct correlation between biological activity and binding to cellular retinoic acid binding protein. Cancer Res. 47, 3523-3527 (1987).

Kagechika, H., E. Kawachi, Y. Hashimoto and K. Shudo: New type inducers of differentiation of human HL-60 pro- myelocytic leukemia cells. Chem. Pharm. Bull. 32, 4209- 4212 (1984).

MacCabe, J.A. and B.W. Parker: The in vitro maintenance of the apical ectodermal ridge of the chick embryo wing bud: an assay for polarizing activity. Dev. Biol. 45, 349-357 (1975).

Maden, M. and D. Summerbell: Retinoic acid-binding protein in chick limb bud: identification at developmental stages and binding affinities of various retinoids. J. Embryol. Exp. Morphol. 97, 239-250 (1986).

Maden, M., D.E. Ong, D. Summerbell and F. Chytil: Spatial distribution of cellular protein binding to retinoic acid in the chick limb bud. Nature (London) 335, 733-735 (1988).

Noji, S., E. Koyama, T. Yamaai, T. Nohno, M. Petkovich, P. Chambon and T. Taniguchi: Developmental expression of the retinoic acid receptor genes, as revealed by in situ hybridization. Cell Differ. and Develop. (Supple) 27, S169 (1989).

Petkovich, M., N.J. Brand, A. Krust and P. Chambon: A human retinoic acid receptor which belongs to the family of nuclear receptors. Nature (Lond.) 330, 444-450 (1987).

Roberts, A.B. and M.B. Spom: Cellular biology and biochem- istry of the retinoids. In: ‘The Retinoids’ eds., M.B. Spom, A.B. Roberts and D.S. Goodman, Vol. 2 (Academic Press, New York) pp. 209-286 (1984).

Saunders, J.W., Jr. and M.T. Gassehng: Ectodermal- mesenchymal interactions in the origin of limb symmetry. In: Epithehal-Mesenchymal Interactions, (eds., R. Fleish- majer and R.E. Billingham (Williams and Wilkins, Bal- timore) pp. 78-97 (1968).

Shudo, K., H. Kagechika, E. Kawachi and Y. Hashimoto: Chalcone carboxylic acids. Potent differentiation inducers of human promyelocytic cells HL-60. Chem. Pharm. Bull. 33, 404-407 (1985).

Summerbell, D.: A quantitative analysis of the effect of exci- sion of the AER from the chick limb bud. J. Embryol. Exp. Morphol. 32, 651-660 (1974).

Summerbell, D.: The zone of polarising activity: evidence for a role in normal chick limb morphogenesis. J. Embryol. exp. Morph. 50, 217-233 (1979).

Summerbell, D.: The control of growth and the development of pattern across the anterc-posterior axis of the chick limb bud. J. Embryol. Exp. Morphol. 63, 161-180 (1981).

Summerbell, D.: The effect of local application of retinoic acid to the anterior margin of the developing chick limb. J. Embryol. Exp. Morphol. 78, 269-289 (1983).

Summerbell, D. and F. Harvey: Vitamin A and the control of

pattern in developing limbs. In: Limb Development and Regeneration, eds., J.F. Fallon and A. I. Caplan, Part A (Alan R. Liss, New York) pp. 109-118 (1983).

Takagi, K., M. Suginuma, H. Kagechika, K. Shudo, M. Ninomiya, Y. Muto and H. Fujiki: Inhibition of omithine decarboxylase induction by retinobenzoic acids in relation to their binding affinities to cellular retinoid-binding pro- teins. J. Cancer Res. Clin. Oncol. 114, 221-224 (1988).

Tamura, K., K. Ohsugi and H. Ide: Distribution of retinoids in the chick limb bud: analysis with monoclonal antibody. Dev. Biol., in press (1990).

Thaller, C. and G. Eichele: Identification and spatial distribu- tion of retinoids in the developing chick limb bud. Nature (Lond.) 327, 625-628 (1987).

Tickle, C., D. Summerbell and L. Wolpert: Positional signal- ling and specification of digits in chick limb morphogene- sis. Nature (Lond.) 254, 199-202 (1975).

Tickle, C., B.M. Alberts, L. Wolpert and J. Lee: Local applica- tion of retinoic acid to the limb bud mimics the action of the polarizing region. Nature (Lond.) 296, 564-566 (1982).

Tickle, C., J. Lee and G. Eichele: A quantitative analysis of the effect of all-truns-retinoic acid on the pattern of chick wing development. Dev. Biol. 109, 82-95 (1985).

Tickle, C., A. Crawley and J. Farrar: Retinoic acid application to chick wing buds leads to a dose-dependent reorganiza- tion of the apical ectodermal ridge that is mediated by the mesenchyme, Development 106, 691-705 (1989).

Wilson, D.J. and J.R. Hinchhffe: The effect of the zone of polarizing activity (ZPA) on the anterior half of the chick wing bud. Development 99, 99-108.

Wolpert, L.: Positional information and the spatial pattern of cellular differentiation. J. Theor. Biol. 25, l-47 (1969).

Wolpert, L.: Positional information and pattern formation. Curr. Top. Dev. Biol. 6, 183-224 (1971).