TERATOLOGY 43:217-224 (1991)
Effects of Retinoic Acid on Chick Tail Bud Development C. MAY GRIFFITH
AND
MICHAEL J. WILEY
Department of Anatomy, University of Toronto, Toronto, Ontario, Canada M5S l A 8
ABSTRACT The present study describes the teratogenic effects of retinoic acid (RA) on the development of the chick tail bud. Chick embryos were recovered 48 hours after treatment at HH stages 11 to 16 with various dosages of RA by subblastodermal injection. At the gross level, RA treatment resulted in varying degrees of caudal regression, scoliosis, limb malformations, and open posterior neuropores among the survivors. Histological examination of tail buds from treated embryos revealed defects which included total dysplasia of caudal structures, the presence of accessory neural tube and notochord tissue, and abnormal fusions of the notochord to the neural tube and tailgut. The incidence, severity, and location of the defects were dependent on the dose of the teratogen, and the stage of development a t the time of treatment. The defects resembled those induced in previous studies by treatment with sialic acid binding lectins such as wheat germ agglutinin and limulus polyphemus lectin (Griffth and Wiley, '90b). Recent studies suggest that regional differences exist in the early development of the axial structures of the embryo. While the three germ layers of classical embryology are responsible for the development of the rostra1 levels of the neural tube, notochord, and primitive gut, the tail bud appears to play an important role in the development of the caudal levels of these structures (Schoenwolf, '77, '78, '79; Svajger et al., '85; Gajovic et al., '89). The tail bud is an aggregate of mesenchyma1 cells located caudal to the posterior neuropore, and is derived from the remains of the primitive streak. During secondary neurulation, cells in the tail bud have been shown to undergo a mesenchymal-to-epithelial transformation to form the caudal part of the neural tube in chicks (Schoenwolf, '77, '78, ,791, rodents (Schoenwolf, '84; Shedden and Wiley, '87), and humans (Lemire, '69). However, species differences exist, both in the details of the morphogenetic process and in the extent to which the tail bud contributes to the spinal cord of the adult. In rodents, secondary neurulation involves an even and regular extension of the neurocoele of the primary neural tube into the tail bud, while tail bud mesenchymal cells become radially arranged around the growing lumen and adopt characteristics of 0 1991 WILEY-LISS, INC.
neuroectoderm cells (Schoenwolf, '84; Shedden and Wiley, '87). In chick embryos, however, the tail bud cells initially differentiate into two populations; an outer population of neuroectodermal cells radially arranged around an inner group of cells which retain their mesenchymal appearance. Isolated cavities appear a t the boundary between the two populations. These cavities enlarge at the expense of the group of inner cells, coalesce and establish continuity with the lumen of the primary neural tube. Development of the caudal part of the avian neural tube from the tail bud is further complicated by the presence of an "overlap zone." Within this zone, the dorsal part of the neural tube is derived by primary neurulation from the tapering neural folds of the posterior neuropore, while the ventral part is derived by secondary neurulation from the tapering cranial end of the tail bud (Schoenwolf, '79; Schoenwolf and DeLongo, '80; Dryden, '80). Extirpation experiments and studies of chick-quail chimeras have been used to
Received August 20, 1990; accepted October 23, 1990. Address reprint request6 to Dr. Michael J. Wiley, Department of Anatomy, Medical Sciences Building, University of Toronto, Toronto, Ontario, Canada M5S 1A8.
218
C.M. GRIFFITH AND M.J. WILEY
demonstrate that the tail bud forms the lumbosacral segments of the chick neural tube (Schoenwolf, '77, '78; Schoenwolf et al., '85). However, the extent of the tail bud contribution to the mammalian neuraxis is less certain (Schoenwolf, '84; Shedden and Wiley, '87). Similarly, while there is evidence to suggest a contribution from the tail bud to the caudal part of the primitive gut and notochord, the extent of the contribution is not known (Svajger et al., '85; Gajovic et al., '89). The species differences in the normal development of the tail bud could be expected to lead to differences in response to teratogens which affect its development. Retinoic acid (RA) has been shown to induce a variety of notochordal abnormalities and dysplastic changes in the caudal part of the spinal cord in rodents exposed during the early period organogenesis. Such defects have been linked with the disruptive effects of the teratogen on the morphology of the early tail bud (Tibbles and Wiley, '88).Very little is known, however, of the effects of the teratogen on avian tail bud development. Jelinek and Kistler ('81) reported that administration of RA to chick embryos on day 2 produced a nonspecific syndrome of caudal regression. However, their study utilized only a small number of embryos (10) and no histological examination was reported. The objective of the present investigation was to provide a more detailed account of the effects of RA on the development of the secondary neuraxis and associated tail bud-derived structures in the chick. MATERIALS AND METHODS
Fertile White Leghorn eggs (Glen Fenelon Farms, Stroud, Ontario, Canada) were incubated in a humidified forced draft incubator at 38°C for 48 to 56 hours. Treatment Eggs were windowed and the embryos staged according to the criteria of Hamburger and Hamilton ('51). Embryos at stages 13-early 14 (the tail bud anlagen stage) were used in this study. This stage was chosen based on reports of tail bud abnormalities induced by RA when administered to rodent embryos at comparable stages (Shenefelt, '72; Shedden and Wiley, '87; Tibbles and Wiley, '88). Each embryo was given a subblastoderma1 injection of all-trans-retinoic acid (Sig-
ma Chem. Co., St. Louis, MO) suspended in 5 pl of 0.9% saline. Based on the report of RA teratogenicity by Jelinek and' Kistler (%l), doses of 0.5, 1.0, 2.0, 5.0, and 10.0 pg per embryo were tested. Twenty embryos were used in each group. In addition, control groups of 20 embryos each, that were either windowed only, or windowed and administered saline were prepared. The eggs were then resealed using sellotape and reincubated for another 48 hours. Stage response to retinoic acid exposure The 4 groups of embryos used in this study were a t different stages of tail bud development, namely (1) late primitive streak stage (HH stages 11-12), (2) tail bud anlagen stage (HH stages 13-early 14), (3) early tail bud stage (mid-late HH stage 14), and (4) late tail bud stage (HH stages 1516). Each embryo was given a subblastoderma1 injection of 1.0 pg RA suspended in 5 p1 saline. Eggs were then resealed and reincubated for approximately 48 hours. Forty embryos were used in each group. Stagematched saline-injected control groups were also prepared. Analysis of embryos The proportion of survivors recovered after the 48 hour reincubation period was noted and the embryos were fixed in Carnoy's fluid (Humason, '79). Fixed embryos were examined for gross malformations before processing for histological study. In order to examine embryos for internal malformations of the tail bud, the tissues at the level of, and caudal to the hind limb buds were dissected out and routinely processed for wax embedding (Humason, '79). Transverse serial sections of 6 pm thickness were made for each sample. The sections were then stained with haematoxylin and eosin (H and E; Humason, '79) for histological examination. All windowed and saline control embryos with external malformations and 10 "normal" embryos chosen at random were sectioned for microscopical examination. Statistical analysis The effect of increasing doses of RA on survival and teratogenicity was analyzed by linear regression and chi-square proportional analysis of a 2 x 2 contingency table, with a correction for continuity (Zar, '84). The chi-square analysis was also used to an-
EFFECTS OF RETINOIC ACID TABLE 1 . The effect of increasing dosages of R A on the survival and incidence of caudal axial malformations in stage 13-14 chick embryos after 48 hours exposure to the teratogen Survivors with caudal No. Survivors axial defects Dose (UP) treated (%) (%) Sham 20 95 0 Saline 20 90 0 0.5 1.o 2.0 5.0 10.0
20 20 20 20 20
65 75 20
31 47 75
10 0
-
-
alyze the survival rate and proportion of embryos with caudal defects caused by administration of RA at different stages of development. Statistical significance was set at Pc0.05. RESULTS
Dose-response There were no significant differences in the survival and malformation rates between windowed controls and those injected with saline (Table 1). The proportion of surviving embryos recovered 48 hours after RA exposure at HH stages 13-14 decreased rapidly with increasing doses of the teratogen (Table 1) and at a dose of 10.0 Fg per embryo, mortality was 100%.The proportion of survivors with caudal malformations increased up to a dose of 2 pg per embryo. However, further increases in dosage resulted in an apparent decline in abnormalities. This decline was likely due to the high mortality at higher dosages and the death of malformed embryos. Exposure of embryos to RA at stages 1314 produced a spectrum of caudal malformations ranging from shortened tails to complete caudal regression (Fig. 1).In addition to caudal axial defects, embryos with scoliosis and malpositioned hind limbs were also observed (Table 2). The dose which produced the optimal combination of low mortality and a high incidence of caudal malformations was 1.0 kg per embryo at stages 13-14. This dose was therefore used to test the pattern of response to the teratogen following treatment at other stages of tail bud development. Stage-response The effects of RA administration appeared to be stage dependent (Fig. 2). The
219
earlier stages (HH stages 11-12 and 13early 14) showed a significant decline in survival rates after exposure to 1.0 Fg RA, compared to saline-treated controls. However, embryos that were treated with RA during mid-late stage 14 and stages 15-16 showed no significant decrease in survival in comparison with saline controls. Significant increases in the proportion of survivors with caudal defects were found in RA-exposed embryos from all the stages studied. Caudal regression was the most common gross abnormality observed. Although the defect was found at all stages tested, partial regression was more prevalent in the younger embryos (stages 11-12, 13-early 14) while complete regression was more frequently observed in the more advanced embryos (stages mid-late 14, 15-16). Scoliosis and malpositioned limbs were also seen across all stages tested (Table 2). Transverse sections through the lumbosacral and more caudal regions of embryos showed that saline-treated controls had well-developed neural tubes, notochords, ganglia, and differentiating somites by 48 hours after treatment (Fig. 3). In comparison, RA treatment during the late primitive streak stages (stages 11-12) produced accessory neural tube lumina, accessory fragments of notochord, neural tube-notochord fusions, notochord-gut wall fusions, and ourenteric outgrowths in the lumbosacral region (Figs. 4, 5). In the ourenteric outgrowths, which are outgrowths of tissues into the gut that contain a combination of neural tube, notochord, and/or somites, the neural tube and notochord were continuous with the neural tube and notochord of the trunk. In more severe cases, i.e., when there was caudal regression at the gross level, the structural arrangement within the overlap zone was completely disrupted (Fig. 4F). Embryos treated at the tail bud anlagen stage (stages 13-early 14) showed similar lumbosacral defects, in proportions that were generally not significantly different from those of the pretail bud embryos (Fig. 5). The proportion of surviving embryos with neural tube-notochord fusions was, however, significantly higher than those of the pretail bud stages. Embryos exposed at mid-late stage 14 or stages 15-16 showed a significantly decreased response to the teratogen compared to the younger stages (Fig. 5). In addition to a decline in the severity of the defects, a
220
C.M. GRIFFITH AND M.J. WILEY
Fig. 1. Embryos fixed at 48 hours after exposure to 1.0 pg RA at stage 13-early 14, showing a range of caudal malformations at the gross level. (A) Control.
(B-D) Embryos showing increasing severity of caudal regression (from shortening of the tail t o the absence of any tail structures). Bar, 250 pm.
TABLE 2 . The effect of R A (1 pg per embryo) on the incidence of scoliosis, malpositioned hind limbs, and open posterior neuropore in chick embryos recovered 48 hours after treatment with the teratogen at different stages of development Survivors (%) with Malpositioned Open posterior Stages Treatment Survivors' Scoliosis hind limbs neuropore 11-12 Saline 38 5 0 0 29 3 10 10 RA 13-arly Saline 38 0 3 0 14 RA 26 12 27 8 Mid-late Saline 36 0 0 0 14 RA 29 3 10 3 15-16 Saline 38 0 0 0 35 0 3 0 RA 'Forty embryos were used in each treatment group.
caudal shift in the position of the defects was observed in embryos treated with RA during increasingly advanced stages in development. In embryos that were treated with RA at stages 15-16, the defects were usually observed near the tip of the tail. Vascular lesions (Fig. 4F) were also a common finding among the RA-treated embryos and the proportion of embryos affected also appeared to show a stage dependence. Thirty-one percent of embryos treated at stages 11-12 had vascular lesions as did 12% of embryos treated at stages 13+arly 14, 14% of mid-late stage 14 treated embryos, and 3% of stage 15-16 treated specimens.
DISCUSSION
Treatment of chick embryos by subblastodermal injection of RA resulted in defects of the caudal axis, which at the gross level were manifested as varying degrees of caudal regression. When examined histologically, the defects included total disruption of tail bud development, accessory neural tube lumina, accessory branches of the notochord, and abnormal relationships between the notochord and the neural tube or tail gut. While total disruption of tail structures was most frequently encountered, the presence of accessory neural tube lumina and the abnormal fusion between the neural
EFFECTS OF RETINOIC ACID
221
100 r
80
60
40 L
0
2
20
0 11-12
A
13-early
14
nxd-lafe
14
15-16
S t a g e of Development (n=4o in each group)
Fig. 3. Transverse section through the caudal region of a saline-treated control embryo. n, neural tube; c, notochord; g, ganglion; s, somites. Bar, 100 pm.
N (D
CII
' .?
40
m
C
C
-
11-12
O
Z
B
0sallne
13-early
C
n L
m m
m ID
m
20
14
C
mid-lafe
14
35-16
S t a g e of Development RA
Fig. 2. The effects of 1.0 p,g/embryo RA on (A) the survival and (B) the incidence of malformation in embryos exposed to the teratogen during different stages of caudal axial development. *Significant difference from saline controls (P
Effects of Retinoic Acid on Chick Tail Bud Development C. MAY GRIFFITH
AND
MICHAEL J. WILEY
Department of Anatomy, University of Toronto, Toronto, Ontario, Canada M5S l A 8
ABSTRACT The present study describes the teratogenic effects of retinoic acid (RA) on the development of the chick tail bud. Chick embryos were recovered 48 hours after treatment at HH stages 11 to 16 with various dosages of RA by subblastodermal injection. At the gross level, RA treatment resulted in varying degrees of caudal regression, scoliosis, limb malformations, and open posterior neuropores among the survivors. Histological examination of tail buds from treated embryos revealed defects which included total dysplasia of caudal structures, the presence of accessory neural tube and notochord tissue, and abnormal fusions of the notochord to the neural tube and tailgut. The incidence, severity, and location of the defects were dependent on the dose of the teratogen, and the stage of development a t the time of treatment. The defects resembled those induced in previous studies by treatment with sialic acid binding lectins such as wheat germ agglutinin and limulus polyphemus lectin (Griffth and Wiley, '90b). Recent studies suggest that regional differences exist in the early development of the axial structures of the embryo. While the three germ layers of classical embryology are responsible for the development of the rostra1 levels of the neural tube, notochord, and primitive gut, the tail bud appears to play an important role in the development of the caudal levels of these structures (Schoenwolf, '77, '78, '79; Svajger et al., '85; Gajovic et al., '89). The tail bud is an aggregate of mesenchyma1 cells located caudal to the posterior neuropore, and is derived from the remains of the primitive streak. During secondary neurulation, cells in the tail bud have been shown to undergo a mesenchymal-to-epithelial transformation to form the caudal part of the neural tube in chicks (Schoenwolf, '77, '78, ,791, rodents (Schoenwolf, '84; Shedden and Wiley, '87), and humans (Lemire, '69). However, species differences exist, both in the details of the morphogenetic process and in the extent to which the tail bud contributes to the spinal cord of the adult. In rodents, secondary neurulation involves an even and regular extension of the neurocoele of the primary neural tube into the tail bud, while tail bud mesenchymal cells become radially arranged around the growing lumen and adopt characteristics of 0 1991 WILEY-LISS, INC.
neuroectoderm cells (Schoenwolf, '84; Shedden and Wiley, '87). In chick embryos, however, the tail bud cells initially differentiate into two populations; an outer population of neuroectodermal cells radially arranged around an inner group of cells which retain their mesenchymal appearance. Isolated cavities appear a t the boundary between the two populations. These cavities enlarge at the expense of the group of inner cells, coalesce and establish continuity with the lumen of the primary neural tube. Development of the caudal part of the avian neural tube from the tail bud is further complicated by the presence of an "overlap zone." Within this zone, the dorsal part of the neural tube is derived by primary neurulation from the tapering neural folds of the posterior neuropore, while the ventral part is derived by secondary neurulation from the tapering cranial end of the tail bud (Schoenwolf, '79; Schoenwolf and DeLongo, '80; Dryden, '80). Extirpation experiments and studies of chick-quail chimeras have been used to
Received August 20, 1990; accepted October 23, 1990. Address reprint request6 to Dr. Michael J. Wiley, Department of Anatomy, Medical Sciences Building, University of Toronto, Toronto, Ontario, Canada M5S 1A8.
218
C.M. GRIFFITH AND M.J. WILEY
demonstrate that the tail bud forms the lumbosacral segments of the chick neural tube (Schoenwolf, '77, '78; Schoenwolf et al., '85). However, the extent of the tail bud contribution to the mammalian neuraxis is less certain (Schoenwolf, '84; Shedden and Wiley, '87). Similarly, while there is evidence to suggest a contribution from the tail bud to the caudal part of the primitive gut and notochord, the extent of the contribution is not known (Svajger et al., '85; Gajovic et al., '89). The species differences in the normal development of the tail bud could be expected to lead to differences in response to teratogens which affect its development. Retinoic acid (RA) has been shown to induce a variety of notochordal abnormalities and dysplastic changes in the caudal part of the spinal cord in rodents exposed during the early period organogenesis. Such defects have been linked with the disruptive effects of the teratogen on the morphology of the early tail bud (Tibbles and Wiley, '88).Very little is known, however, of the effects of the teratogen on avian tail bud development. Jelinek and Kistler ('81) reported that administration of RA to chick embryos on day 2 produced a nonspecific syndrome of caudal regression. However, their study utilized only a small number of embryos (10) and no histological examination was reported. The objective of the present investigation was to provide a more detailed account of the effects of RA on the development of the secondary neuraxis and associated tail bud-derived structures in the chick. MATERIALS AND METHODS
Fertile White Leghorn eggs (Glen Fenelon Farms, Stroud, Ontario, Canada) were incubated in a humidified forced draft incubator at 38°C for 48 to 56 hours. Treatment Eggs were windowed and the embryos staged according to the criteria of Hamburger and Hamilton ('51). Embryos at stages 13-early 14 (the tail bud anlagen stage) were used in this study. This stage was chosen based on reports of tail bud abnormalities induced by RA when administered to rodent embryos at comparable stages (Shenefelt, '72; Shedden and Wiley, '87; Tibbles and Wiley, '88). Each embryo was given a subblastoderma1 injection of all-trans-retinoic acid (Sig-
ma Chem. Co., St. Louis, MO) suspended in 5 pl of 0.9% saline. Based on the report of RA teratogenicity by Jelinek and' Kistler (%l), doses of 0.5, 1.0, 2.0, 5.0, and 10.0 pg per embryo were tested. Twenty embryos were used in each group. In addition, control groups of 20 embryos each, that were either windowed only, or windowed and administered saline were prepared. The eggs were then resealed using sellotape and reincubated for another 48 hours. Stage response to retinoic acid exposure The 4 groups of embryos used in this study were a t different stages of tail bud development, namely (1) late primitive streak stage (HH stages 11-12), (2) tail bud anlagen stage (HH stages 13-early 14), (3) early tail bud stage (mid-late HH stage 14), and (4) late tail bud stage (HH stages 1516). Each embryo was given a subblastoderma1 injection of 1.0 pg RA suspended in 5 p1 saline. Eggs were then resealed and reincubated for approximately 48 hours. Forty embryos were used in each group. Stagematched saline-injected control groups were also prepared. Analysis of embryos The proportion of survivors recovered after the 48 hour reincubation period was noted and the embryos were fixed in Carnoy's fluid (Humason, '79). Fixed embryos were examined for gross malformations before processing for histological study. In order to examine embryos for internal malformations of the tail bud, the tissues at the level of, and caudal to the hind limb buds were dissected out and routinely processed for wax embedding (Humason, '79). Transverse serial sections of 6 pm thickness were made for each sample. The sections were then stained with haematoxylin and eosin (H and E; Humason, '79) for histological examination. All windowed and saline control embryos with external malformations and 10 "normal" embryos chosen at random were sectioned for microscopical examination. Statistical analysis The effect of increasing doses of RA on survival and teratogenicity was analyzed by linear regression and chi-square proportional analysis of a 2 x 2 contingency table, with a correction for continuity (Zar, '84). The chi-square analysis was also used to an-
EFFECTS OF RETINOIC ACID TABLE 1 . The effect of increasing dosages of R A on the survival and incidence of caudal axial malformations in stage 13-14 chick embryos after 48 hours exposure to the teratogen Survivors with caudal No. Survivors axial defects Dose (UP) treated (%) (%) Sham 20 95 0 Saline 20 90 0 0.5 1.o 2.0 5.0 10.0
20 20 20 20 20
65 75 20
31 47 75
10 0
-
-
alyze the survival rate and proportion of embryos with caudal defects caused by administration of RA at different stages of development. Statistical significance was set at Pc0.05. RESULTS
Dose-response There were no significant differences in the survival and malformation rates between windowed controls and those injected with saline (Table 1). The proportion of surviving embryos recovered 48 hours after RA exposure at HH stages 13-14 decreased rapidly with increasing doses of the teratogen (Table 1) and at a dose of 10.0 Fg per embryo, mortality was 100%.The proportion of survivors with caudal malformations increased up to a dose of 2 pg per embryo. However, further increases in dosage resulted in an apparent decline in abnormalities. This decline was likely due to the high mortality at higher dosages and the death of malformed embryos. Exposure of embryos to RA at stages 1314 produced a spectrum of caudal malformations ranging from shortened tails to complete caudal regression (Fig. 1).In addition to caudal axial defects, embryos with scoliosis and malpositioned hind limbs were also observed (Table 2). The dose which produced the optimal combination of low mortality and a high incidence of caudal malformations was 1.0 kg per embryo at stages 13-14. This dose was therefore used to test the pattern of response to the teratogen following treatment at other stages of tail bud development. Stage-response The effects of RA administration appeared to be stage dependent (Fig. 2). The
219
earlier stages (HH stages 11-12 and 13early 14) showed a significant decline in survival rates after exposure to 1.0 Fg RA, compared to saline-treated controls. However, embryos that were treated with RA during mid-late stage 14 and stages 15-16 showed no significant decrease in survival in comparison with saline controls. Significant increases in the proportion of survivors with caudal defects were found in RA-exposed embryos from all the stages studied. Caudal regression was the most common gross abnormality observed. Although the defect was found at all stages tested, partial regression was more prevalent in the younger embryos (stages 11-12, 13-early 14) while complete regression was more frequently observed in the more advanced embryos (stages mid-late 14, 15-16). Scoliosis and malpositioned limbs were also seen across all stages tested (Table 2). Transverse sections through the lumbosacral and more caudal regions of embryos showed that saline-treated controls had well-developed neural tubes, notochords, ganglia, and differentiating somites by 48 hours after treatment (Fig. 3). In comparison, RA treatment during the late primitive streak stages (stages 11-12) produced accessory neural tube lumina, accessory fragments of notochord, neural tube-notochord fusions, notochord-gut wall fusions, and ourenteric outgrowths in the lumbosacral region (Figs. 4, 5). In the ourenteric outgrowths, which are outgrowths of tissues into the gut that contain a combination of neural tube, notochord, and/or somites, the neural tube and notochord were continuous with the neural tube and notochord of the trunk. In more severe cases, i.e., when there was caudal regression at the gross level, the structural arrangement within the overlap zone was completely disrupted (Fig. 4F). Embryos treated at the tail bud anlagen stage (stages 13-early 14) showed similar lumbosacral defects, in proportions that were generally not significantly different from those of the pretail bud embryos (Fig. 5). The proportion of surviving embryos with neural tube-notochord fusions was, however, significantly higher than those of the pretail bud stages. Embryos exposed at mid-late stage 14 or stages 15-16 showed a significantly decreased response to the teratogen compared to the younger stages (Fig. 5). In addition to a decline in the severity of the defects, a
220
C.M. GRIFFITH AND M.J. WILEY
Fig. 1. Embryos fixed at 48 hours after exposure to 1.0 pg RA at stage 13-early 14, showing a range of caudal malformations at the gross level. (A) Control.
(B-D) Embryos showing increasing severity of caudal regression (from shortening of the tail t o the absence of any tail structures). Bar, 250 pm.
TABLE 2 . The effect of R A (1 pg per embryo) on the incidence of scoliosis, malpositioned hind limbs, and open posterior neuropore in chick embryos recovered 48 hours after treatment with the teratogen at different stages of development Survivors (%) with Malpositioned Open posterior Stages Treatment Survivors' Scoliosis hind limbs neuropore 11-12 Saline 38 5 0 0 29 3 10 10 RA 13-arly Saline 38 0 3 0 14 RA 26 12 27 8 Mid-late Saline 36 0 0 0 14 RA 29 3 10 3 15-16 Saline 38 0 0 0 35 0 3 0 RA 'Forty embryos were used in each treatment group.
caudal shift in the position of the defects was observed in embryos treated with RA during increasingly advanced stages in development. In embryos that were treated with RA at stages 15-16, the defects were usually observed near the tip of the tail. Vascular lesions (Fig. 4F) were also a common finding among the RA-treated embryos and the proportion of embryos affected also appeared to show a stage dependence. Thirty-one percent of embryos treated at stages 11-12 had vascular lesions as did 12% of embryos treated at stages 13+arly 14, 14% of mid-late stage 14 treated embryos, and 3% of stage 15-16 treated specimens.
DISCUSSION
Treatment of chick embryos by subblastodermal injection of RA resulted in defects of the caudal axis, which at the gross level were manifested as varying degrees of caudal regression. When examined histologically, the defects included total disruption of tail bud development, accessory neural tube lumina, accessory branches of the notochord, and abnormal relationships between the notochord and the neural tube or tail gut. While total disruption of tail structures was most frequently encountered, the presence of accessory neural tube lumina and the abnormal fusion between the neural
EFFECTS OF RETINOIC ACID
221
100 r
80
60
40 L
0
2
20
0 11-12
A
13-early
14
nxd-lafe
14
15-16
S t a g e of Development (n=4o in each group)
Fig. 3. Transverse section through the caudal region of a saline-treated control embryo. n, neural tube; c, notochord; g, ganglion; s, somites. Bar, 100 pm.
N (D
CII
' .?
40
m
C
C
-
11-12
O
Z
B
0sallne
13-early
C
n L
m m
m ID
m
20
14
C
mid-lafe
14
35-16
S t a g e of Development RA
Fig. 2. The effects of 1.0 p,g/embryo RA on (A) the survival and (B) the incidence of malformation in embryos exposed to the teratogen during different stages of caudal axial development. *Significant difference from saline controls (P
