Retinoid-induced limb defects 1: inhibition of cell proliferation in distal mesenchyme of limb buds in rats

Reproductive Toxicology, Vol. 13, No. 2, pp. 103–111, 1999 © 1999 Elsevier Science Inc. Printed in the USA. All rights reserved 0890-6238/99/$–see front matter

PII S0890-6238(98)00069-0

RETINOID-INDUCED LIMB DEFECTS 1: INHIBITION OF CELL PROLIFERATION IN DISTAL MESENCHYME OF LIMB BUDS IN RATS HISASHI TSUIKI and KURAJIRO KISHI Developmental Research Laboratories, Shionogi & Co., Ltd., Osaka, Japan Abstract — The present study was undertaken to investigate the effects of all-trans-retinoic acid (RA) on cell death and limb bud growth in forelimb buds and also to examine whether these events are involved in limb bone defects induced by RA in rats. RA was given at doses of 50 and 100 mg/kg to pregnant rats on Day 12 of pregnancy. Although RA did not show teratogenecity in the 50 mg/kg group, micromelia was observed in the 100 mg/kg group in all live fetuses on Day 21 of gestation. Micromelia was characterized by high incidences of proximodistal reduction of forearm bones without reduction of the humerus. The incidence of cell death in prechondrogenic areas, which differentiate into humerus and forearm bone, significantly increased 24 h after RA treatment in not only the 100 mg/kg, but also the 50 mg/kg, group. There was no difference in the incidence of cell death in the prechondrogenic area between the two groups. These observations indicate that the bone-specific defects were not the result of cell death alone in the prechondrogenic area. We examined the effects of RA on early forelimb bud growth, which is indispensable for the morphogenesis of the forelimb. Proximodistal length and protein content were decreased significantly in the forelimb bud 24 h after RA treatment at a dose of 100 mg/kg, but not 50 mg/kg. The immunohistochemical detection of bromodeoxyuridine (BrdU) incorporated into cells showed that at a dose of 100 mg/kg, cell proliferation was reduced in the distal mesenchyme, but not in the forearm-bone prechondrocytes of the forelimb bud. As the distal margin provides the cells differentiating into the prechondrocytes of future bones in the limb bud, these observations suggested that RA-induced inhibition of cell proliferation in the distal margin resulted in a decrease of forearm-bone prechondrocytes localized at more distal sites. We conclude that RA may inhibit the chondrogenesis of forearm bones by reducing cell proliferation in the distal margin of the forelimb bud, not by increasing cell death, and that this results in reduction defects in forearm bones. © 1999 Elsevier Science Inc. Key Words: retinoic acid; cell death; reduction defect; forelimb bud; proximodistal growth; protein content; cell proliferation.

reports are available for direct demonstration that an increase in cell death caused by RA is responsible for the limb malformations. Kurishita found that 5-azacytidine (5-AC), known as a teratogen that causes limb reduction defects, increases the cell death in the rat limb bud (6). His report also pointed out that caffeine suppresses the limb defects, although post-treatment with caffeine enhances the cell death caused by 5-AC in the limb bud. These findings suggest that induction of cell death is not essential for teratogenesis. In mice, RA is reported to produce 100% limb defects at a dose of 100 mg/kg on Day 11 of pregnancy (plug day 5 Day 0) (7). Under these experimental conditions, Alles and Sulik have suggested that retinoid-induced excessive cell death in the prechondrogenic area may inhibit chondrogenesis followed by long bone reduction (8). They also showed that excessive cell death in the core mesenchyme of the mouse limb, which is related to programmed cell death, primarily interferes with initial aggregation of the prechondrogenic mesenchyme, causing abnormal bone development. However, Abbott et al. showed that no increase in cell death occurs in mouse limb bud treated with a teratogenic dose of RA under the same conditions

INTRODUCTION Administration of excess retinoids to pregnant mammals causes malformations in their offspring. The type of malformation depends on the dose and developmental stage at which the retinoid is administered (1). Rodents given retinoids in the midgestational period mainly develop limb defects or cleft palate (2). These limb defects often coincide with severe micromelia or phocomelia (3). Many hypotheses have been presented to interpret the limb reduction defects caused by retinoids, which have also been shown to interfere in a number of developmental events in the limb. One of the major mechanisms involved in teratogenesis is cell death (4). It has been also found that an increase in cell death occurs in limb buds treated with retinoic acid (RA) (5), suggesting that excessive cell death induces reduction defects in limb. However, no Address correspondence to Hisashi Tsuiki, Developmental Research Laboratories, Shionogi & Co., Ltd. 3-1-1, Futaba-cho, Toyonaka, Osaka 561-0825, Japan. Received 10 August 1998; Revision received 9 November 1998; Accepted 15 November 1998. 103

104

Reproductive Toxicology

(same strain of mouse and same dose) (9) as Alles and Sulik used in their experiments except for the treatment time; Abbotts’ group treated the dams on Day 10 to Day 10.5, counted in the same fashion as Alles and Sulik. The discrepancy of results from the two groups suggests that there may be different mechanism(s) responsible for limb defects that depends on the time of administration. In addition, it also has been shown that retinoids inhibit chondrogenesis in limb bud cells in vitro but do not exhibit a cytotoxic effect (10). Therefore, although the experimental conditions are different, these reports suggest that cell death alone cannot explain the mechanism of the limb defects caused by RA. Further studies are needed to elucidate the actual role of cell death in limb malformations caused by RA. RA regulates gene expression through its interaction with intracellular binding proteins or receptors, and it plays an important role in embryogenesis via these nuclear interactions (11). In the developing limb bud, endogenous RA is present (12), and various binding proteins and receptors are expressed restrictively in space and time during limb development (13). Moreover, the destruction of these receptors causes abnormal limb formation (14). These facts indicate that endogenous RA regulates pattern formation in the limb through complex nuclear interactions, and that exogenous RA seems to perturb these interactions, causing abnormal pattern formation. At the macroscopic level, proximodistal pattern formation depends on the proximodistal length and/or local change in cell proliferation rates, which may be indispensable to cell differentiation in the limb bud (15). Therefore, proximodistal growth and cell proliferation may be good candidates for examining the effects of excessive RA on limb pattern formation. Teratogenic doses of RA inhibits proximodistal growth in the chick wing bud (16). However, there are no reports that clarified the relationship between proximodistal growth and limb malformation caused by RA in mammals. Only a few laboratory groups showed that RA affects cell proliferation in the limb bud (16,17), suggesting the possibility that excessive RA perturbs normal pattern formation via inhibition of proximodistal growth or cell proliferation rather than induction of cell death, which then causes the limb defects. However, the hypothesis needs to be proven clearly, and the mechanism of RA-inhibition of limb bud growth, in which nuclear interaction must be involved, remains to be clarified. It has been shown that excessive RA treatment (120 mg/kg) on Days 12 to 13 of gestation (plug day 5 Day 0) caused almost 100% limb defects in rat fetuses (18). In our previous report, retinoid treatment on Day 12 of gestation also produced 100% limb defects in rat fetuses (19). In addition, we also found that in vitro TGF-b2 is related to limb malformation caused by retinoids rather

Volume 13, Number 2, 1999

than induction of cell death. However, it remained unclear in our reports whether cell death alone is responsible for the limb reduction defects caused by retinoids. Thus, we examined the relationship between amount of cell death and limb bud growth in the limb defects caused by RA to elucidate the teratogenic mechanism of retinoids in rats. MATERIALS AND METHODS Animals Mature male and female Sprague-Dawley rats were purchased from Clea Japan, Inc. (Shiga, Japan). Female rats were mated with male rats overnight. Females with a seminal plug on the following morning were considered pregnant at Day 0 and housed individually or in pairs in cages throughout the experiments. Experimental schedule We performed the administration of RA according to the schedule used in previous studies (19). Pregnant rats on Day 12 of gestation were gavage-fed 0, 50, or 100 mg/kg body weight RA (Sigma, St. Louis, MO). RA was suspended in corn oil (Sigma) and the application volume was 10 mL/kg body weight. Eighteen dams were killed by cervical dislocation 24 h after administration. Eight embryos per dam were used to examine cell death in the forelimb bud, and four embryos per dam to measure proximodistal length and protein content in the forelimb bud. Six embryos obtained from each dam in the control and 100 mg/kg RA groups were used to examine cell proliferation in the forelimb bud. Sixteen dams were killed on Day 21 of gestation, and approximately half of the fetuses were harvested for alizarin staining. Alizarin staining The fetuses were killed by excessive i.p. administration of pentobarbital sodium (Nembutal Sodium Solution, Abbott Laboratories, North Chicago, IL) followed by decortication for fixing with 70% ethanol for 10 d. After treatment with 1% KOH for 12 h, the fetuses were stained with 0.01% alizarin red S (Nacalai Tesque Inc., Kyoto, Japan) in 0.1% KOH (w/v) for 8 h. Forelimb or hindlimb skeletal elements were observed after washing with a solution of glycerin:70% ethanol: benzylalcohol (2:2:1). Examination of cell death Forelimb buds were dissected from the embryos 24 h after exposure to RA, then fixed in 4% paraformaldehyde, dehydrated, and embedded in paraffin. Next 3 mm serial transverse sections were cut, followed by

Retinoid inhibition of limb bud development ● H. TSUIKI and K. KISHI

105

Table 1. Fetal development in pregnant rats given RA by peroral application on Day 12 of gestation

Treatment Control mean SD RA mean SD

Dose (mg/kg)

No. of dams

—

6

50 100

Dead fetuses

No of corpora lutea

No. of implants

Implant ratio (%)

Live fetuses

Resorbed

Macacerated

Dead

Fetal viability (%)

18.3 1.9

17.8 2.2

98 2.9

17.3 1.9

0.5 0.8

0 0

0 0

98 4.2

17.6 1.5

17.2 1.8

98 3.3

17.0 1.6

0.2 0.4

0 0

0 0

99 2.2

16.8 1.1

16.0 1.2

95 4.7

16.0 1.2

0 0

0 0

0 0

5 5

mean SD

100 0

Fetuses were obtained from pregnant rats on Day 21 of gestation, as described in Materials and Methods. Implant ratio (%) 5 (No. of implants/No. of corpora lutea) 3 100 Fetal viability (%) 5 (No. of live fetuses/No. of implants) 3 100

dewaxing and staining with hematoxylin and eosin (HE). Eight sections per embryo were observed. The frequency of pyknotic cells, which were considered to be dead cells, occupying the prechondrogenic area was calculated. The incidence of cell death was obtained from eight sections per embryo, averaged and classified into four groups (2: ,1%, 1: ,10%, 11: ,20%, 111: ^20%). Protein contents The forelimb buds obtained from the embryos 24 h after exposure to RA were evaluated for proximodistal length, which corresponds to the distance from the mid position of the proximal end to the apical tip in the limb bud, by micro measure, and then were sonicated in phosphate buffered saline. An equal volume of 10% perchloric acid was added to the homogenate followed by incubation at 70°C. The protein pellet was isolated by centrifugation at 600 g, rinsed once with 5% PCA, and dissolved in 0.1 M NaOH. The protein was assayed by Lowrys’ method (20). Cell proliferation The whole embryo culture system (21) was used for estimating cell proliferation in the limb bud treated with or without RA as reported previously (17). The conceptuses obtained from the dams in control and 100 mg/kg RA administration after 24 h were transferred to sterile saline. The embryos were dissected free of decidua and

Reichert’s membrane, and the yolk sac and aminion were opened. Treated embryos were incubated in heat-inactivated male rat serum supplemented with 2 mg/mL glucose and 400 mM of BrdU (Sigma) for 1 h at 37°C in 95% O2 and 5% CO2. After culture, the forelimb buds were dissected from embryos and fixed in 4% paraformaldehyde. Next the forelimb buds were dehydrated and embedded in paraffin. Transverse 3-mm sections of forelimb buds were cut and used for immunohistochemical detection of BrdU. After dewaxing, the sections were treated with 0.3% hydrogen peroxide in methanol for 30 min to block endogenous peroxidase, permeabilized with 2 M HCl and 0.1% trypsin (Sigma) in PBS for 30 min and 10 min, respectively, blocked with 1% normal goat serum for 30 min, and incubated overnight at 4°C with mouse antiBrdU antibody (Becton Dickinson Immunocytometry System, San Jose, CA). The sections were then incubated with biotin-conjugated goat anti-mouse IgG antibody (Vector Laboratories, Burlingame, CA) for 1 h, and visualization was performed by VECTASTAIN Elite ABC kit and DAB Substrate kit (Vector Laboratories). The sections were counterstained with methyl green, and the BrdU-positive cells and the negative cells were counted respectively. The percentage of positive cells compared with the total counted cells was calculated. Values were obtained from six slides per embryo and averaged for two regions (distal mesenchyme, distal prechondrocytes).

Table 2. Teratogenic effects of RA given on Day 12 of pregnancy evaluated in Day 21 rat fetuses No of surviving fetuses malformed (%) Treatment Control RA

Dose (mg/kg)

No. of live fetuses (No. of dams)

Micromelia

Oligo-, brachyor syn/dactyly

Cleft palate

— 50 100

105 (6) 85 (5) 80 (5)

0 (0) 0 (0) 80** (100)

0 (0) 0 (0) 80** (100)

0 (0) 0 (0) 25* (31)

*,**: Statistically significant compared to control at P , 0.05 and P , 0.01, respectively.

106

Reproductive Toxicology

Volume 13, Number 2, 1999

Table 3. Congenital limb defects induced by RA given on Day 12 of gestation in Day 21 rat fetuses

Treatment Control RA

Proximal-distal reduction (%) of

Dose (mg/kg)

No. of fetuses examined

Humerus

Ulna

Radius

Femur

Tibia

Fibula

— 50 100

51 25 41

0 (0) 0 (0) 0a (0)

0 (0) 0 (0) 38 (93)

0 (0) 0 (0) 39 (95)

0 (0) 0 (0) 14 (34)

0 (0) 0 (0) 36 (88)

0 (0) 0 (0) 37 (90)

a

Four embryos had humerus lacking the deltoid eminence.

Statistical analyses The numbers of implants, corpora lutea, and live fetuses were analyzed by one-way analysis of variance and subsequently by Duncan’s multiple range test (22). The implantation ratio or fetal viability per litter was analyzed by the Kruskal-Wallis test and Dunn test (23). The mean values for proximodistal length and protein content per litter were analyzed by the Bartlett test (22) and Dunnet test (24,25). The numbers of fetuses with external anomalies were analyzed by chi-square (2 3 2 contingency table) and then by Fisher’s exact probability test (22). The incidence of cell death in the prechondrogenic area was analyzed by cumulative chi-square test (26). The cell proliferation rate per litter was analyzed by F test and then by Student’s t test (22). Differences of P , 0.05 were considered significant. RESULTS RA-induced abnormality No significant differences between control and RAtreated groups were found in the number of implants, corpora lutea, or live fetuses, and implantation ratio or fetal viability (Table 1). The types and numbers of external anomalies in Day 21 live fetuses caused by RA are shown in Table 2. No fetuses with external anomalies were found in the control and 50 mg/kg groups. However, maternal treatment with 100 mg/kg of RA caused limb defects, especially proximodistal reduction defects such as micromelia or branchydactyly in all live fetuses. At the dose of 100 mg/kg, RA also induced cleft palate in the fetuses at low, but significant (P , 0.05), rates compared with the fetuses of the control group. No skeletal anomalies in the limbs were found in fetuses of the control and 50 mg/kg groups (Table 3). However, RA at 100 mg/kg produced an extremely high incidence of short and/or curved long bones in the fetuses (Table 3 and Figure 1). Alizarin red S staining demonstrated that of the fetuses exposed to 100 mg/kg of RA, 93% showed reduction of the ulna and 95% showed a defect of the radius. In the 100 mg/kg group, the frequencies of fetuses with reduced tibia or fibula were also fairly high (tibia: 88%, fibula: 90%). There was no shortening of the humerus, although some humeri lacked the deltoid eminence, and lower frequencies of reduced

femur (34%) were found in the fetuses of the 100 mg/kg group. Cell death The incidence of cell death in the mesenchymal core region, which corresponded to the future forearm bone or humerus prechondrocytes, was examined for control and RA-treated embryos. When examined at Day 12 to 14, Day 13 embryos of the control group showed the maximal incidence of cell death in the prechondrogenic area. Therefore, the embryos 24 h after exposure to RA on Day 12 of gestation were used. Prechondrocytes of the forearm bone and humerus were aggregating cell populations (Figure 2). The maximal incidence of cell death in the prechondrogenic area of forelimb bud exposed to RA did not exceed 30%. RA at both 50 mg/kg and 100 mg/kg significantly (P , 0.01) enhanced the incidence of cell death in the prechondrogenic area of the humerus as well as the forearm bones compared with the control embryos, but the frequency of cell death showed a tendency to increase in the prechondrocytes of the humerus in comparison with those of the forearm bones (Table 4 and Figure 2). Although RA at 100 mg/kg could cause limb reduction defects in fetuses but not at 50 mg/kg, no significant differences for the incidence of cell death between the two groups treated with 50 mg/kg and

Fig. 1. Forelimb bone defects in Day 21 fetuses treated with RA. A. Control; B. 50 mg/kg RA; C. 100 mg/kg RA. The rats were exposed to RA on Day 12 of gestation. At a dose of 50 mg/kg, RA did not cause limb bone defects. Although RA at 100 mg/kg caused shortening of the radius, ulna, and digital bones, no proximodistal reduction was found in the humerus.

Retinoid inhibition of limb bud development ● H. TSUIKI and K. KISHI

107

Fig. 2. Histological analysis of cell death in the prechondrogenic area of the forelimb bud 24 h after exposure to RA. A,D,E: Control; B,F,G: 50 mg/kg RA; C,H,I: 100 mg/kg The figures show H.E-stained humerus prechondrogenic area (A,B,C) or ulna prechondrogenic area (D,E,F,G,H,I). E,G, and I are magnified views of areas indicated by the thick arrows in D,F, and H, respectively. The peripheral proximal mesenchyme in the limb bud consisted of presumptive myogenic cells (arrowheads in D,F,H). Thin arrows indicate dead cells showing pknosis. In the humerus and ulna prechondrogenic area, RA at 50 mg/kg, as well as 100 mg/kg, increased the number of dead cells compared to the control (A vs B,C in humerus; E vs G,I in ulna). Magnification: A,B,C,E,G,I 3208; D,F,H 352. Magnification bar 5 100 mm.

108

Reproductive Toxicology

Volume 13, Number 2, 1999

Table 4. Effects of RA on cell death in forelimb prechondral region 24 h after administration of RA Cell death in prechondral Treatment Control RA

Dose (mg/kg)

No. of embryos examined

2

Humerus 1 11

50 100

24 24 24

0 0 0

13 2 2

11 10 10

111

2

Ulna or radius 1 11

0 12** 12**

0 0 0

18 6 5

6 16 17

111 0 2** 2**

Eight sections per embryo were observed. The incidence of cell death observed in the prechondrogenic area was classified into the following four groups: 2: ,1% cell death area in prechondral area, 1: ,10% cell death area in prechondral area, 11: ,20% cell death area in prechondral area, 111: ^20% cell death area in prechondral area. **: Significant difference from control at P , 0.01.

100 mg/kg of RA were found in either the future humerus or forearm bones.

100 mg/kg RA was significantly (P , 0.01) decreased to 122.0 6 2.4 mg/limb bud compared with the control.

Proximodistal length of the forelimb bud Figure 3 shows the proximodistal length in the forelimb bud 24 h after RA treatment. The axial length from the midposition of proximal end to the tip of distal end was obtained for the proximodistal length. No significant differences between the control and 50 mg/kg groups were found in the proximodistal length (mean values 6 SE; control: 1.80 6 0.02 mm, 50 mg/kg of RA: 1.80 6 0.02 mm). However RA at 100 mg/kg reduced the proximodistal length (1.65 6 0.02 mm) slightly, but significantly (P , 0.01), compared with the control (Figures 3 and 4).

Cell proliferation in forelimb buds Distal mesenchyme (cells in the distal marginal zone: area from the tip, underlying apical ectodermal ridge (AER) to the front of distal prechondrocytes) or distal prechondrocytes in the core mesenchyme in the forelimb bud were examined (Figure 5). The cell proliferation rates, represented as percentages of BrdU-positive cells in these zones, were 9.2 6 0.5% (mean 6 SE) and 9.7 6 0.6% for the core zones of the control and 100 mg/kg groups, respectively, with the difference not being significant (Figure 6). On the other hand, at a dose of 100 mg/kg, RA significantly (P , 0.05) reduced the cell proliferation rates in the distal marginal zone to 46.2 6 3.2% compared with 59.5 6 3.2% of those in the control (Figure 6). These reductions in cell proliferation were attributable to the decrease of BrdU-positive cells in the loose-packed cell population located between the AER and the forearm-bone prechondrogenic area (Figure 7).

Protein content of forelimb buds Figure 3 also shows the protein content in the forelimb buds 24 h after RA treatment. The mean values for the control and 50 mg/kg groups were 147.3 6 4.8 mg/limb bud (mean 6 SE) and 140.9 6 1.6 mg/limb bud, respectively, with the difference not being significant. However, protein content in the forelimb bud exposed to

DISCUSSION Our main findings can be summarized as follows:

Fig. 3. Effects of RA on forelimb bud development. The forelimb buds were obtained from embryos 24 h after exposure to RA. Numbers in parentheses represent the number of embryos examined. **Significant difference from control at P , 0.01.

Fig. 4. External morphogenesis in forelimb buds 24 h after RA treatment. Forelimb buds dissected from control embryo (left) and embryo exposed to 100 mg/kg of RA (right). The proximodistal (P-D) length in the forelimb bud exposed to 100 mg/kg of RA is inferior to the control’s. D: Distal, P: Proximal.

Retinoid inhibition of limb bud development ● H. TSUIKI and K. KISHI

Fig. 5. Area measured for cell-proliferation rate in forelimb bud. Cell proliferation in the distal mesenchyme and distal prechondrogenic area of the forelimb bud were examined.

1. No fetal malformations were seen in the litters of rat dams given a single dose of 50 mg/kg of RA on Gestation Day 12. A dosage of 100 mg/kg caused forelimb defects (reduction defects in ulna and radius). Nevertheless, the humerus was not affected. 2. At 24 h following treatment on Gestation Day 12: ● Both 50 and 100 mg/kg of RA produced a similar enhancement in the incidence of cell death in the prechondrogenic areas of humerus, ulna, and radius. ● One hundred, but not 50, mg/kg of RA significantly reduced the proximodistal length of the forelimb bud. ● Protein content in the forelimb buds exposed to 100 (but not 50) mg/kg of RA was significantly decreased. ● The dose of 100 mg/kg significantly reduced cell proliferation in the distal mesenchyme of the forelimb buds; cell proliferation in the core mesenchyme (future radius and ulna) was not affected. Thus, although both 50 and 100 mg/kg of RA produced a very high incidence of cell death in the humerus and forearm-bone prechondrocytes of the fore-

Fig. 6. Effects of RA on cell proliferation of the mesenchyme in the forelimb bud. Forelimb buds were obtained from embryos cultured for 1 h in BrdU. Numbers in parentheses represent the number of embryos examined. Cell proliferation rate 5 (No. of cells incorporated BrdU/No. of counted cells) 3 100. *Significant difference from control at P , 0.05.

109

limb, forelimb defects were found only in the litters of dams treated with 100 mg/kg of RA. Also, humerus defects were not seen at any dosage. This suggests that RA-induced cell death alone is not sufficient to explain the observed terata in term fetuses. It would seem that RA-induced inhibition of the proximodistal growth of the forelimb buds, seen at 100, but not 50, mg/kg of RA, is related to the reduction defects of the forearm bones. In the mouse, a teratogenic dose of RA inhibited proximodistal growth of limb buds (27). Similarly, local application of RA to the chick wing bud failed to produce cell death but did reduce bud growth and ultimately caused shortened long bones (16). We and others (28) found that teratogenic doses of RA reduced the protein content in forelimb buds. Thus, the decrease of protein content in the limb bud may be related to the limb defects caused by RA, because neither protein reduction nor limb defects were observed at the 50 mg/kg dose of RA. The reduction in the protein content was not attributable to excessive cell death because there were no differences in the incidence of cell death between groups treated with 50 (nonteratogenic) and 100 (teratogenic) mg/kg of RA. The decline in the protein content could indicate a decrease in the cell population and/or reduction in the proximodistal length in the forelimb bud. The proximodistal growth of the limb bud is known to depend on the distal mesenchyme proliferation in and around the progress zone underlying the AER (15). Therefore, a decrease in the protein content and proximodistal length might suggest a decrease in the amount of distal mesenchyme. Indeed, we did find decreased cell proliferation in this region in the 100 mg/kg group. Others have reported a similar decrease in the mesenchymal cells of the forelimb bud in the RA-treated mouse (9) and in limb bud cells in culture (29). However, Kochhar reported that RA reduces cell proliferation in core mesenchyme as well as distal mesenchyme (17). In his experiment, the cell proliferation evaluated by 3H-thymidine incorporation in the limb bud of cultured embryos was performed at an earlier development-stage than we used. Therefore, it is possible that the discrepancy could be attributable to the different stage used in the experiments. Moreover, the long time culture (4 h) may result in a higher labeling index in his report than in ours. According to the developmental processes in the limb bud, the mesenchymal cells committed to osteogenesis migrate from the progress zone to a more proximal region and then begin differentiation to prechondrocytes (15). The prechondrogenic zone incorporates the migrating mesenchyme and enlarges, followed by limb bud growth. Thus, the increase of prechondrocytes depends on the amount of mesenchyme migrating from the distal region and its proliferation rate. Because 100 mg/kg of

110

Reproductive Toxicology

Volume 13, Number 2, 1999

Fig. 7. Immunohistochemical localization of Day-13 limb bud mesenchyme incorporated BrdU. (A) Control. (B) 100 mg/kg RA; In the control limb bud, the dark cells recognized as BrdU-incorporated were widely distributed in the loose-packed mesenchyme at the distal region. In the RA-treated limb bud, the stained cells were more restricted, and none could be seen in the central-distal region (arrowhead). Note that cells showing less proliferation in RA-treated forelimb bud were adjacent to the prechondrogenic area of the ulna. Arrows: Apical ectodermal ridge U: Ulna prechondrogenic area. Magnification: (A),(B) 3130. Magnification bar 5 100 mm.

RA suppressed cell proliferation in the distal mesenchyme, but not in the core mesenchyme (prechondrocytes), we can assume that RA treatment decreases the number of cells migrating out of the progress zone into the prechondrogenic zone, which ultimately leads to bone reduction defects. Finally, because the humerus prechondrocytes are localized in a more proximal region than the forearm-bone prechondrocytes, distal mesenchymal cells may not significantly contribute to the humerus; therefore, a reduction in the amount of distal mesenchyme may have little or no effect on the growth of the humerus. This could explain why the radius and ulna, but not the humerus, are affected by RA treatment. The mechanism of RA inhibition of cell proliferation in the distal mesenchyme is unclear. The AER is important for controlling cell proliferation in the progress zone (15). It is possible that excessive RA disturbs AER action, reducing cell proliferation. However, local application of RA to the apex of the limb bud reduced the proximodistal growth, which depends on cell proliferation in the limb bud. This growth inhibition was not attributable to the effects of RA on the AER (16). In this report, it was shown that a tissue recombination made between RA-treated mesenchyme and untreated epithelia containing the AER caused truncated limbs, but the recombination between RA-treated epithelia and untreated mesenchyme did not. We found no pathogenic changes in the AER (data not shown). Therefore, excessive RA may affect the limb bud mesenchyme directly and inhibits cell proliferation. CRABP is an RA-binding protein expressed in the limb bud at a much higher level distally than proximally,

reaching a peak level in the subapical mesenchyme of the progress zone (13,30). This expression pattern suggests that CRABP indicates potential fitness in the scheme of proximodistal pattern control in the limb. Because CRABP seems to act as a sink to restrict the free RA available to nuclear receptors and remains at a required constitutive level (31), the presence of abundant CRABP in and around the progress zone suggests that these cells may need little or no RA for proliferation. Therefore, it is likely that the free RA escaping from CRABP may disturb the nuclear interaction and inhibit proliferation of the cells in and around the progress zone. However, the RA–nuclear interaction that brings about a reduction of cell proliferation remains unclear. Hox genes, which regulate axial pattern formation in embryogenesis or various growth factors the expression of which is regulated by RA, are involved in limb development (15,32– 34), and one or more of these factors may be related to RA inhibition of cell proliferation. From this point of view it is interesting that cycloheximide, an inhibitor of protein synthesis, has been shown to significantly protect the long bones against the teratogenic effects of RA in the rat fetus (18). Some protein factors may be related to retinoid inhibition of proliferation of the distal mesenchyme in the forelimb buds. We have explored this further in the accompanying paper. In conclusion, we have demonstrated that excessive cell death is not essential for limb reduction defects caused by RA, but a teratogenic dose of RA decreases distal marginal mesenchyme in the forelimb buds of affected rat fetuses. Because these cells migrate proximally and contribute to the future long bones, a reduction in their proliferation leads to reductions in the protein

Retinoid inhibition of limb bud development ● H. TSUIKI and K. KISHI

content and proximodistal length of the limb bud, and ultimately to the reduction defects of long bones. Acknowledgment — We thank Dr. Yonetaka Fukiishi of Development Product Managers Unit, Shionogi & Co., Ltd., for his helpful suggestions and discussions.

REFERENCES 1. Shenefelt RE. Morphogenesis of malformations in hamsters caused by retinoic acid: Relation to dose and stage at treatment. Teratology. 1972;5:103–18. 2. Kochhar DM, Penner JD, Tellone CI. Comparative teratologenic activities of two retinoids: effects on palate and limb development. Teratogenesis Carcinog Mutagen. 1984;4:377– 87. 3. Kochhar DM. Limb development in mouse embryos: I. Analysis of teratogenic effects of retinoic acid. Teratology. 1973;7:289 –98. 4. Sandor S, Mic AF. Cell death in normo- and teratogenesis II. Cell death in teratogenesis. Morphol Embryol. 1994;3– 4:83–91. 5. Kochhar DM, Agnish ND. Chemical surgery as an approach to study morphogenic events in embryonic mouse limb. Dev Biol. 1977;61:388 –94. 6. Kurishita A. Histological study of cell death in digital malformations induced by 5-Azacytidine: suppressive effect of caffeine. Teratology. 1989;39:163–72. 7. Kochhar DM, Penner JD, Tellone CI. Comparative teratogenic activities of two retinoids: effects on palate and limb development. Teratog Carcinog Mutagen. 1984;4:377– 87. 8. Alles AJ, Sulik KK. Retinoic-acid-induced limb-reduction defects: Perturbation of zones of programmed cell death as a pathogenetic mechanism. Teratology. 1989;40:163–71. 9. Abbott BD, Hill LG, Birnbaum LS. Processes involved in retinoic acid production of small embryonic palatal shelves and limb defects. Teratology. 1990;41:299 –310. 10. Renault J-Y, Melcion C, Cordier A. Limb bud cell culture for in vitro teratogen screening: Validation of an improved assessment method using 51 compounds. Teratog Carcinog Mutagen. 1989;9: 83–96. 11. Morris-Kay G. Retinoic acid receptors in normal growth and development. Cancer Surv. 1992;14:181–93. 12. Satre MA, Kochhar DM. Elevations in the endogenous levels of the putative morphogen retinoic acid in embryonic mouse limbbuds associated with limb dysmorphogenesis. Dev Biol. 1989;133: 529 –36. 13. Ruberte E, Nakshatri H, Kastner P, Chambon P. Retinoic acid receptors and binding proteins in mouse limb development. In: Morris-Kay G, ed. Retinoid in normal development and teratogenesis. Oxford: Oxford University Press; 1992:99 –111. 14. Kastner P, Mark M, Chambon P. Nonsteroid nuclear receptors: what are genetic studies telling us about their role in real life? Cell. 1995;83:859 – 69. 15. Tabin CJ. Retinoids, homeoboxes, and growth factors: Toward molecular models for limb development. Cell. 1991;66:199 –217. 16. Tickle C, Crawley A, Farrar J. Retinoic acid application to chick wing buds leads to a dose-dependent reorganization of the apical ectodermal ridge that is mediated by the mesenchyme. Development. 1989; 106:691–705.

111

17. Kochhar DM. Cellular basis of congenital limb deformity induced in mice by vitamin A. In: Bergsma D, Lenz W, eds. Morphogenesis and malformation of the limb. New York: Alan R. Liss; 1977:111– 54. 18. Kistler A. Teratogenesis of retinoic acid in rats: susceptible stages and suppression of retinoic acid-induced limb malformations by cycloheximide. Teratology. 1981;23:25–31. 19. Tsuiki H, Fukiishi Y, Kishi K. Relation of TGF-beta 2 to inhibition of limb bud chondrogenesis by retinoid in rats. Teratology. 1996;54:191–7. 20. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin phenol reagent. J Biol Chem. 1951;192: 265–75. 21. Eto K, Takakubo F. Improved development of rat embryos in culture during the period of craniofacial morphogenesis. J Craniofac Genet Dev Biol. 1985;5:351–5. 22. Gad SC, Weil CS. Statistics for Toxicologists. In: Hayes AW, ed. Principles and methods of toxicology. New York: Raven Press; 1982:273–320. 23. Dunn OJ. Multiple comparisons using rank sums. Technometrics. 1964;6:241–57. 24. Dunnet CW. A multiple comparison procedure for comparing several treatments with a control. J Am Statist Assoc. 1955;50: 1096 –121. 25. Dunnet CW. New tables for multiple comparison with a control. Biometrics. 1964;20:482–92. 26. Yoshimura K, Ohashi Y. Statistical analysis for Toxicology data. Tokyo, Chijin Shokan;1992:45–7. 27. Mendelsohn C, Ruberte E, Lemeur M, Morriss-Kay G, Chambon P. Developmental analysis of the retinoic acid-inducible RAR-beta 2 promoter in transgenic animal. Development. 1991;113:723–34. 28. Kochhar DM. Skeletal morphogenesis: comparative effects of a mutant gene and a teratogene. In: Lash JW, Saxen L, eds. Developmental Mechanisms: Normal and Abnormal. New York, Alan R. Liss; 1985:267– 81. 29. Gardiner DM, Gaudier C, Bryant SV. Mouse limb bud cells respond to retinoic acid in vitro with reduced growth. Exp Zool. 1992;263:406 –13. 30. Maden M, Ong DE, Summerbell D, Chytil F. The role of retinoid-binding proteins in the generation of pattern in the developing limb, the regenerating limb and the nervous system. Development. 1989;107(Suppl.):109 –19. 31. Ong DE, Newcomer ME, Chytil F. Cellular retinoid-binding proteins. In: Sporn MB, Roberts AB, Goodman DS, eds. The retinoids: biology, chemistry, and medicine, 2nd edition. New York: Raven Press; 1994:283–317. 32. Favier B, Rijli FM, Fromental-Ramain C, Fraulob V, Chambon P, Dolle´ P. Functional cooperation between the non-paralogous genes Hoxa-10 and Hoxd-11 in the developing forelimb and axial skeleton. Development. 1996;122:449 – 60. 33. Rijli FM, Chambon P. Genetic interactions Hox genes in limb development: learning from compound mutants. Curr Opin Genet Dev. 1997;7:481–7. 34. Niederreither K, Ward SJ, Dolle´ P, Chambon P. Morphological and molecular characterization of retinoic acid-induced limb duplications in mice. Dev Biol. 1996;176:185–98.