U.S. Environmental Protection Agency, Mid-Continent Ecology Division, 6201 Congdon Boulevard, Duluth, Minnesota 55804
Received January 11, 2003; accepted March 25, 2003
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key Words: retinoic acid exposure; amphibian limb development, retinoid signaling pathways; limb malformations.
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Retinoids are metabolic derivatives of vitamin A, which act as signaling molecules that regulate many processes critical to early embryonic development, including axial patterning, cellular differentiation, tissue induction, proliferation, and apoptosis (Sucov and Evans, 1995). Retinoids elicit their effects through binding to members of the steroid-thyroid nuclear receptor superfamily and regulating transcription of specific target genes. Alterations in retinoid levels, whether it is a condition of excess or deficiency, result in developmental anomalies in nearly all vertebrate species studied. The role of retinoids in vertebrate limb development has received considerable attention. When administered to regenerating amphibian limbs, retinoids induce proximo-distal and mirror-image duplications (Maden, 1983
; Scadding and Maden, 1986a
,b
; Thoms and Stocum, 1984
). In the chick, local application of retinoids affects limb pattern regulation by mimicing the effects of the zone of polarizing activity causing, anterior-posterior duplication (Summerbell, 1983
; Tickle et al., 1982
, 1985
). One common feature of these amphibian and chick experiments is the employment of surgical manipulation of the limb tissue. In contrast, systemic pulsed exposure of developing limbs to retinoids does not result in limb duplication. Rather, treatment of developing bud-stage-stage mice (Kochhar, 1977
; Kochhar et al., 1984
) and amphibians (Degitz et al., 2000
; Scadding and Maden, 1986a
,b
), in the absence of physical manipulation, results in concentration and developmental stage-dependent deletions and reduction of hindlimb skeletal segments.
Ideally, native Ranid species would be used to evaluate the potential for environmental retinoid mimics to affect embryonic and larval development. However, conditions for extended culture and breeding of Ranid species are not well defined, and the use of field collected embryos is limited to the natural breeding season. As an alternative, Xenopus laevis is a well-studied model of vertebrate development, for which a great deal of baseline data are available. In addition, previous studies suggest that the response to retinoids between native Ranids and X. laevis are similar (Degitz et al., 2000).
In the current study, we asked the following question. Does chronic exposure to retinoids during amphibian limb development cause malformations characteristic of those observed in the field specimens? RA was used because it is a potent ligand for the retinoic acid receptors and is the standard retinoid to which nearly all others are compared. We exposed X. laevis to RA continuously, either from stage 8 (Nieuwkoop and Faber 1994) to completion of metamorphosis, or for 1-week periods at critical stages of limb development. We found that RA produces obvious malformations when administered during embryogenesis, as would be expected. However, exposure during embryogenesis or limb development does not produce limb malformations in tadpoles not killed by the exposure to RA.
![]() |
MATERIALS AND METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Experimental Design
Embryonic exposure.
Starting at stage 8, Xenopus embryos were exposed to a range of nominal all-trans retinoic acid (Sigma Chemical Co., St Louis, MO) concentrations (0.012, 0.031, 0.052, 0.086, 0.144, 0.24, 0.6, and 2 µg/l) or Lake Superior water (two controls). Each treatment group consisted of 60 organisms. After 3 days of exposure 10 organisms from each treatment were removed and placed in fixative for later characterization of structural abnormalities by comparing exposed animal to controls. We sampled at 3 days because we found, in preliminary experiments, that several of the treatments resulted in mortality by 3 days, making it impossible to characterize the RA-induced malformations. After 9 days (when controls reached stage 48), 20 organisms for each treatment were removed and maintained in continuous flow of Lake Superior until tail resorption was complete (stage 65). The remaining organisms were continuously exposed to RA until stage 65. When tadpoles reached stage 65, they were removed from the test system, weighed, measured (snout-vent length), examined for gross malformations (controls were used for reference), and preserved in 95% ethanol.
Larval exposure.
At stage 48, X. laevis tadpoles were exposed to a range of nominal RA concentrations (0.031, 0.052, 0.086, 0.114, 0.24, 0.6, 2, and 3 µg/l) or Lake Superior water (controls). Each treatment group consisted of 60 organisms. After 7 or 14 days of exposure, 20 organisms from each treatement were removed and maintained in continuous flow of Lake Superior until stage 65. The remaining organisms were continuously exposed to RA until stage 65. Upon removal from the test system, organisms were weighed, measured (snout-vent length), examined for gross malformations (controls were used for reference), and preserved in 95% ethanol. The limb segments and digits of exposed animals were compared to controls
Exposure system.
Glass wool columns were constructed and as described by Kahl et al., 1999. Five hundred mg of RA was dissolved into ~70 ml of high-pressure liquid chromatography (HPLC)-grade acetone and stirred on a magnetic stir plate until dissolved. One half of the dissolved chemical was poured into one leg of a U-shaped column packed with glass wool. Columns were maintained under a nitrogen atmosphere while the chemical was added to the column leg and while the acetone was being evaporated. The solution containing the RA was pulled up and down the column leg (using a vacuum pump) until the acetone was completely evaporated and the RA coated onto the glass wool. This process was repeated using the remainder of the acetone/chemical solution on the other column leg. After evaporation was complete, neoprene stoppers were inserted in both ends of the column, and the stoppers were taped in place with duct tape. The columns were wrapped in aluminum foil (to limit photodegradation), and placed in a freezer until used for testing. Prior to testing, columns were flushed (using Lake Superior water that had been deoxygenated using nitrogen) at 11.0 ml/minute for 1 hour and at 5.0 ml per minute for an additional 23 hours to remove residual compound.
The delivery system consisted of two columns in series, with a new column being added to replace the oldest column every third day. Deoxygenated water was pumped through the columns at a rate of 1.5 ml/minute into a 30-ml pre-dilution cell. Of this volume, 0.36 ml/minute was pumped to the toxicant dilution cell of a computerized electronic delivery system. In this system, four high-volume Fluid Metering, Incorporated (FMI) pumps supplied water (two supplied control water, and two supplied water containing RA) to four sets of electronic solenoids. The computer fired each solenoid for a preprogramed amount of time to provide the amount of diluent water (control tanks) and diluent plus toxicant-laden water necessary to achieve the desired concentration in each exposure tank. All delivery lines were made of Teflon and kept in the dark to minimize the breakdown of the RA. Exposure tanks were glass aquaria (22.5 x 14.0 x 16.5 cm deep) equipped with 13-cm standpipes, which resulted in an actual tank volume of 4.0 l. The continuous flow rate to each tank was 85 ml/minute. Fluorescent lamps provided a light intensity that ranged from 61 to 139 lumens at the water surface. The photoperiod was 12 h light : 12 h dark.
Water analysis.
All laboratory procedures were conducted under amber lighting. Water samples (10 to 75 ml), were collected from the embryo and larval exposures three times the first week of the test and then once a week until test completion. The samples were placed in amber bottles; acetic acid was added to a strength of 20 mM; acetonitrile was added to a final volume of 45%. Samples were concentrated by processing through solid phase extraction (SPE). Briefly, 3-ml C18 SPE columns (J. T. Baker, Phillipsburg, NJ), fitted with 50-ml reservoirs, were precleaned with 10 ml dichloromethane, activated with 25 ml methanol, and rinsed with 5 ml of deionized water. The samples were then applied to the columns under mild negative pressure (5 mm Hg). Following sample addition, each column was washed with 3 ml aqueous acetonitrile (60:40 v/v) and lightly dried. The adsorbed RA was slowly eluted under vacuum (1 mm Hg) with four 500-µl aliquots of methanol. The eluate was evaporated under nitrogen to a known volume (e.g., 100 µl), transferred to an amber vial containing a microinsert, capped, and immediately analyzed for RA.
The SPE methanol concentrates and methanol-diluted samples were analyzed by reversed-phase HPLC. The HPLC system consisted of a Hewlett-Packard (HP) model 1100 pump and automatic liquid sampler (Hewlett-Packard, Avondale, PA) and a BAS model LC-4B amperometric electrochemical detector (Bioanalytical Systems, Lafayette, IN). An aliquot of each sample (30 µl) was injected directly onto an Adsorbosphere HS C18 5-µm column (Alltech, Deerfield, IL) and eluted isocratically with 100 mM sodium acetate in 65% acetonitrile and 2.5 % methanol (pH = 5, acetic acid) at a flow rate of 1.5 ml/min. Data were collected with HP ChemStation software, and RA concentrations were determined using the external standard method of quantification with a five point linear calibration curve. Routine (i.e., ~10 %) quality assurance analyses (blanks, duplicates, and matrix spikes) were conducted with each SPE sample set. Concentrations of RA in the water samples were corrected based on recovery data. The analytical quantification limit was 0.004 ng/ml. No RA was detected in the blanks from the X. laevis embryo (n = 10) or larval exposures (n = 8). The mean (± SD) percentage agreement among duplicate samples was 91 ± 7.5 % (n = 10) and 93 ± 6.7 % (n = 8) from the X. laevis embryo and larval exposures, respectively. The mean (± SD) recovery of RA in the spiked water samples was 89 ± 11.6 % (n = 23) and 90 ± 8.5 % (n = 19) from the X. laevis embryo and larval exposures, respectively.
Data analysis.
The experimental design of this experiment is unique in that we choose to test a relatively large number of concentrations at the expense of tank replication. This was a necessary step, given the number of different exposure durations tested. With the data available in the literature (stage sensitivity differences) we did not feel that we could test several different life stages with a limited number of test concentrations. The principle objective of these experiments was to determine if continuous RA exposure would result in hindlimb malformations, and the experiments were designed to maximize the potential to do so. Given the steep dose-response curves observed (absence of several treatments with partial effects compared to controls) for mortality and dysmorphogenesis in these experiments, curve fitting models were not applicable in all exposures scenarios. Instead we calculated cumulative probability of survival, and ultimately, mortality as described by (Kaplan, 1958) and present the data as percent mortality in a given treatment. The effects were quite obvious, and presentation would not be enhanced by curve fitting.
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
|
The most important finding of the embryonic exposures is the lack of hindlimb malformations in any of the tadpoles surviving until tail resorption from either the 9-day pulse exposure or the chronic exposure. Both exposure paradigms had at least on experimental concentrations in which partial mortality was observed. At these concentrations we did not observe malformations in the survivors. Further, we monitored the test system several times a day and examined all dead animals for evidence of malformations. We did not observe hindlimb malformations in any of these dead organisms removed from the test system.
Larval Exposure
The larval exposures were initiated at stage 48, tadpoles were exposed for 1 week or 2 weeks and were moved to clean water grow out, or were exposed in the test system until tail resorption was complete. We observed high control mortality (29%) in the chronic larval control 4 and larval chronic control 3 was terminate early due to high mortality (Table 3). However, because the controls for the chronic embryonic and chronic larval exposure were maintained under identical conditions and were collected from that same spawn we used embryonic chronic controls for making comparison in the treatment group exposed until tail resorption. In this group exposure to concentrations of 0.24 µg/l and below did not result in mortality greater than in controls, however 0.6, 2, and 3 µg/l RA caused 100% mortality. Time to 100 % mortality in these three treatments was 35, 18, and 11 days, respectively.
|
Similar to the embryonic exposures, we examined the effects of RA on growth (length and weight) and developmental rate (time to stage 64) (Table 2). We did not observe dramatic effects on these endpoints at concentrations that did not produce mortality above control levels. Thus, these endpoints appear to be only slightly more sensitive to RA treatment.
In contrast to the embryonic exposures, only one of the exposure durations that were tested resulted in partial mortality (7-day exposure). In both the 14-day and chronic exposures, mortality appeared to be an all-or-none situation. Similar to embryonic exposures, RA treatment, regardless of duration, did not produce hindlimb malformations in surviving or dead animals removed from the test system.
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
In the current experiments, we have compared the effects of both life-stage sensitivity and duration of RA exposure on development. The developmental toxicity of RA has been the subject of considerable experimentation in a number of vertebrate species. In mice and rats, exposure to retinoids during gastrulation results in a broad spectrum of craniofacial defects (Sulik et al., 1988, 1995
; Vickers, 1985
). With progression of the mammalian embryo into organogenesis, there is a decrease in the sensitivity to retinoids (Kochhar et al., 1984
). In amphibians there are considerable data on the effects of retinoids on embryonic development (Creech Kraft and Juchau, 1995
; Creech Kraft et al., 1995
; Dawid et al., 1993
; Leroy and De Robertis, 1992
) and on hindlimb development and regeneration (Maden, 1983
, 1985
). Synthesis of these studies suggests a significant stage-specific difference in sensitivity. Previously, we reported a difference in sensitivity of several orders of magnitude when comparing early embryos and tadpoles (Degitz et al., 2000
). In those experiments, stage 48 and 52 tadpoles tolerated concentrations as high as 1250 µg/l RA, while stage 8 embryos showed severe dysmorphogenesis and, ultimately, mortality in nearly all of the animals exposed at 25 µg/l RA. However, all of the studies reported in the literature, including our previous work, were conducted using 24-h static exposure paradigms. In our studies, we showed that RA has a short half-life in aqueous solution and that target concentrations were only achieved for a few hours. In the current study, in which a constant flow-through exposure was used, effects were observed at much lower RA concentrations. This result demonstrates the significance of exposure duration in the ultimate outcome. Further, stage sensitivity differences were not as great as one may have predicted from the pulse exposure paradigm. We chose this approach because it more accurately represents potential environmental exposure to retinoid mimics (i.e., chronic or long-term pulse) rather than a few-hour "pulse."
Yet another unanticipated finding was that RA did not result in malformations in animals surviving until stage 65. Exposure starting at stage 8 resulted in a concentration-dependent increase in the frequency and severity of dysmorphogenesis. This was determined by collecting a subsample three days after the exposure was initiated (stages 4042). However, all organisms in these treatment groups died very early in the study. This mortality was apparently caused by gross abnormalities induced by RA. The spectrum of effects included microphthalmia and reductions in the prosencephalon. With increasing concentration, the effects were more severe, ranging from holoprosencephaly and anophthalmia to complete absence of procephalic tissue. At these higher concentrations, there were also dramatic effects on posterior development. These results are consistent with those previously described for amphibians (Creech Kraft and Juchau, 1995; Creech Kraft et al., 1995
; Dawid et al., 1993
; Degitz et al 2000
; Leroy and De Robertis, 1992
) and mammalian species exposed to retinoids (Sulik et al., 1988
, 1995
). Interestingly, we observed high levels of mortality in treatment groups that did not show obvious signs of dysmorphogenesis following three days of embryonic exposure. There are several possible explanations for these observations: developmental events which follow embryogenesis may be more sensitive to RA; lengthier exposure duration may have accumulative impact; or effects which were not externally obvious caused the mortality. We observed partial mortality in both the 9-day (at 0.24 µg/l) and the chronic embryo (at 0.144 µg/l) exposure paradigms. However, we did not observe externally obvious skeletal malformations in the surviving tadpoles. These results suggest that other aspects of growth and development are more sensitive to excessive retinoid concentrations than skeletal development.
It has been proposed that hindlimb malformations observed in amphibians across North America may be attributed to xenobiotic disruption of retinoid signaling pathways (Gardiner and Hoppe, 1999). This proposal was based on the phenotypes observed in Rana septentrionalis specimens collected from an unnamed lake in central Minnesota. Malformations included reductions, deletions, and supernumerary structures. The current studies were designed with a primary objective of examining the impact of chronic RA exposure on limb development. In mice, duplications of posterior structures, including the pelvis and hindlimb, have been reported following retinoid exposure at the blastula stage of development (Rutledge et al., 1994
). Retinoids have been shown to have the ability to mimic the effects of the zone of polarizing activity in developing limbs, causing anterior-posterior duplication (Summerbell, 1983
; Tickle et al., 1982
, 1985
). When administered to regenerating amphibian limbs, retinoids induce proximo-distal and mirror-image duplications (Maden, 1983
; Scadding and Maden, 1986a
,b
; Thoms and Stocum, 1984
). Given the ability of retinoids to alter normal limb patterning, it is not surprising that systemic exposure of developing limb-bud-stage mice (Kochhar, 1977
; Kochhar et al., 1984
), X. laevis (Scadding and Maden, 1986b
), and axolotl (Scadding and Maden, 1986a
) results in concentration- and developmental stage-dependent deletions and reductions of hindlimb skeletal segments. Given this knowledge base, we proposed that there may be two critical windows for producing limb defects: early embryogenesis and during the period of limb pattern formation. Exposure starting at stage 8, regardless of duration, did not result in limb malformation in surviving organisms. To assess the impact during the period of pattern formation, we exposed tadpoles starting at stage 48 for 1 week, 2 weeks, or until stage 65. Similar to the finding of the embryonic exposures, RA did not induce limb malformation in any surviving tadpole. We are confident that the concentrations used were high enough to produce malformations, given that the highest concentration used resulted in 100% mortality within 2 weeks of initiating the exposure. Further, in the 1-week exposure group, we observed partial mortality at 2 µg/l, but we did not observe limb malformations in surviving tadpoles at stage 65. Again, these results suggest that other aspects of growth and development, which are not externally obvious, are more sensitive to retinoids than skeletal development.
From these experiments and our previous work (Degitz et al., 2000), we conclude that it is unlikely that retinoid mimics are playing a significant role in the recently observed hindlimb malformations in Ranids from the field. We cannot rule out the possibility that short-term exposures at high concentrations could cause limb malformation. However, it is difficult to put forth a pulse exposure scenario that would be at a high enough concentration to produce skeletal malformations yet short enough in duration not to cause mortality. Further, there have been no published reports of retinoid mimics being found at any affected site. If, however, at some point retinoid mimics are found, data from the present study and our previous work will provide a basis for determining if exposure patterns and concentrations are consistent with those necessary to produce hindlimb malformations in amphibians.
![]() |
ACKNOWLEDGMENTS |
---|
![]() |
NOTES |
---|
1 To whom correspondence should be addressed at U.S. EPA, Mid-Continent Ecology Division, Duluth, MN 55804-2595.
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Ankley, G. T., Tietge, J. E., DeFoe, D. L., Jensen, K. M., Holcombe, G. W., Durhan, E. J., and Diamond, S. A. (1998). Effects of ultraviolet light and methoprene on survival and development of Rana pipiens. Environ. Toxicol. Chem. 17, 25302542.[CrossRef][ISI]
Ankley, G. T., Tietge, J. E., Holcombe, G. W., DeFoe, D. L., Diamond, S. A., Jensen, K. M., and Degitz, S. J. (2000). Effects of laboratory ultraviolet radiation and natural sunlight on survival and development of Rana pipiens. Can. J. Zool. /Rev. Can. Zool. 78, 10921100.[ISI]
Creech Kraft, J., and Juchau, M. R. (1995). Xenopus laevis: A model system for the study of embryonic retinoid metabolism. III. Isomerization and metabolism of all-trans retinoic acid and 9-cis-retinoic acid and their dysmorphogenic effects in embryos during neurulation. Drug. Metab. Dispos. 23, 10581071.[Abstract]
Creech Kraft, J., Kimelman, D., and Juchau, M. R. (1995). Xenopus laevis: a model system for the study of embryonic retinoid metabolism. I. Embryonic metabolism of 9-cis-and all-trans-retinals and retinols to their corresponding acid forms. Drug Metab. Dispos. 23, 7282.[Abstract]
Dawid, I. B., Otani, H., Curtiss, P., and Taira, M. (1993). Regulatory interactions during embryogenesis in Xenopus laevis. C. R. Acad. Sci. III 316, 945958.[Medline]
Degitz, S. J., Kosian, P. A., Makynen, E. A., Jensen, K. M., and Ankley, G. T. (2000). Stage- and species-specific developmental toxicity of all-trans retinoic acid in four native North American ranids and Xenopus laevis. Toxicol. Sci. 57, 264274.
Gardiner, D. M., and Hoppe, D. M. (1999). Environmentally induced limb malformation in mink frogs (Rana septentrionalis). J. Exp. Zool. 284, 207216.[CrossRef][ISI][Medline]
Johnson, P. T. J., Lunde, K. B., Haight, R. W., Bowerman, J., and Blaustein, A. R. (2001). Ribeiroia ondatrae (Trematoda: Digenea) infection induces severe limb malformations in western toads (Bufo boreas) Can. J. Zool. 79, 370379.[CrossRef][ISI]
Johnson, P. T. J., Lunde, K. B., Ritche, E. G., and Launer, A. E. (1999). The effect of trematode infection on amphibian limb development and survivorship. Science 284, 802804.
Johnson, P. T. J., Lunde, K. B., Thurman, E. M., Ritchie, E. G., Wray, S. N., Sutherland, D. R., Kapfer, J. M., Frest, T. J., Bowerman, J., and Blaustein, A. R. (2002). Parasite (Ribeiroia ondatrae) infection linked to amphibian malformations in the western United States. Ecol. Monogr. 72, 151168.[ISI]
Kahl, M. D., Russom, C. L., DeFoe, D. L., and Hammermeister, D. E. (1999). Saturation units for use in aquatic bioassays. Chemosphere 39, 539551.[CrossRef][ISI][Medline]
Kaplan, E. L. (1958). Nonparametric estimation from incomplete observations. J. Am. Stat. Assoc. 53, 457481.[ISI]
Kiesecker, J. M. (2002) Synergism between trematode infection and pesticide exposure: A link to amphibian limb deformities in nature? Proc. Natl. Acad. Sci. 99, 99009904.
Kochhar, D. M. (1977). Cellular basis of congenital limb deformity induced in mice by vitamin A. In Morphogenesis and Malformation of the Limb (D. Bergsma and W. Lenz, Eds.), pp. 111154. A. R. Liss, New York, NY.
Kochhar, D. M., Penner, J. D., and Tellone, C. I. (1984). Comparative teratogenic activity of two retinoids: Effects on palate and limb development. Terat. Carcin. Mutag. 4, 377387.[ISI]
Leroy, P., and De Robertis, E. M. (1992). Effects of lithium chloride and retinoic acid on the expression of genes from the Xenopus laevis Hox 2 complex. Dev. Dyn. 194, 2132.[ISI][Medline]
Maden, M. (1983). The effects of vitamin A on limb regeneration in Rana temporaria. Dev. Biol. 98, 409416.[ISI][Medline]
Maden, M. (1985). Retinoids and the control of pattern in regenerating limbs. Ciba. Found. Symp. 113, 132155.[ISI][Medline]
Nieuwkoop, P. D., and Faber, F. (1994). Normal table of Xenopus laevis. Garland Publishing, New York.
Rutledge, J. C., Shourbaji, A. G., Hughes, L. A., Polifka, J. E., Cruz, Y. P., and Bishop, J. B. (1994). Limb and lower-body duplications induced by retinoic acid in mice. Proc. Natl. Acad. Sci. 91, 54365440.[Abstract]
Scadding, S. R., and Maden, M. (1986a). Comparison of the effects of vitamin A on limb development and regeneration in the axolotl, Ambystoma mexicanum. J. Embryol. Exp. Morph. 91, 1934.[ISI][Medline]
Scadding, S. R., and Maden, M. (1986b). Comparison of the effects on vitamin A on limb development and regeneration in Xenopus laevis tadpoles. J. Embryol. Exp. Morphol. 91, 3553.[ISI][Medline]
Sulik, K. K., Cook, C. S., and Webster, W. S. (1988). Teratogens and craniofacial malformation: relationships to cell death. Development 103(Suppl.), 213232.[ISI][Medline]
Sulik, K. K., Dehart, D. B., Rogers, J. M., and Chernoff, N. (1995). Teratogenicity of low doses of all-trans retinoic acid in presomite mouse embryos. Teratology 51, 398403.[ISI][Medline]
Summerbell, D. (1983). The effect of local application of retinoic acid to the anterior margin of the developing chick limb. J. Embryol. Exp. Morph. 78, 269289.[ISI][Medline]
Sucov, H. M., and Evans, R. M. (1995). Retinoic acid and retinoic acid receptors in development. Mol. Neurobiol. 10, 16984.[ISI][Medline]
Thoms, S. D., and Stocum, D. L. (1984). Retinoic acid induced pattern duplication in regenerating urodele limbs. Dev. Biol. 103, 319328.[ISI][Medline]
Tickle, C., Alberts, B., Wolpert, L., and Lee, J. (1982). Local application of retinoic acid to the limb bud mimics the action of the polarizing region. Nature 296, 564566.[ISI][Medline]
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. Dev. Biol. 109, 8295.[ISI][Medline]
Vickers, T. H. (1985). Embryo lethality in rats caused by retinoic acid. Teratology 31, 1933.[ISI][Medline]
|