TERATOLOGY 43:71-81 (1991)
Parameters Determining lsotretinoin Teratogenicity in Rat Embryo Culture HELEN RITCHIE AND WILLIAM S. WEBSTER Department of Anatomy, University of Sydney, Sydney NSW 2006, Australia
ABSTRACT
At the in vitro threshold serum concentration of 500 ng/ml, isotretinoin induces defects of visceral arch development in 9.5-day rat embryos grown in culture for 48 h. Experiments were performed to determine the minimum period of exposure necessary to induce these arch defects and whether an increase in concentration of isotretinoin could compensate for reduced exposure time. The results showed that a minimum 6-h exposure to 500 ng/ml immediately prior to cranial neural crest migration was necessary to induce severe defects of the second visceral arch in a majority of embryos. Maximal increase in isotretinoin concentration t o 16,000 ng/ml did not compensate for shorter exposure periods. These results suggest that to cause malformations of the visceral arches, the embryo must be exposed to isotretinoin for a minimum period of time regardless of the concentration of isotretinoin above the threshold.
The isotretinoin (IT) or retinoic acid embryopathy is a restricted pattern of malformation seen in 22-40% of the offspring of women taking the drug Accutane during the first trimester of pregnancy (Lammer et al., '85; Rosa et al., '86). The embryopathy is characterized by facial dysmorphia, particularly reduced and abnormal external ears. There may also be a flat nasal bridge, mandibular underdevelopment and cleft palate. Many of the children also have conotruncal heart defects, interrupted aortic arch, and a variable deficiency of the thymus. Brain defects are another common feature and include hydrocephalus associated with hypoplasia of the cerebellum with or without agenesis of the cerebellar vermis. The association of facial defects, great vessel and conotruncal malformations and thymic deficiency in the human IT embryopathy indicates a common pathogenesis, since these structures all require a contribution from the cranial neural crest for their normal development (Johnston, '66; Noden, '75; Le Lievre and Le Douarin, '75; Bockman and Kirby; '84; Kirby et al., '83, '85). In vivo animal models of the IT embryopathy can duplicate many of the malformations seen in the human (Webster et al., '86; Willhite et al., '86), but the models are un-
0 1991 WILEY-LISS, INC.
satisfactory for two main reasons. First, depending on species, laboratory animals require 20-400 times the human daily IT dose to induce the malformations typical of the human embryopathy (Kamm, '82; Webster et al., '86; Willhite et al., '86). Second, in laboratory animals, a much wider range of malformations is induced, including limb (Kochhar et al., '84, '87), urogenital, and gastrointestinal defects (Willhite and Shealy, '84). By contrast, in a review of the outcome of human IT pregnancies, there were no substantiated reports of major defects affecting the limbs and only two minor defects affecting the gastrointestinal and urogenital systems (unspecified hepatic abnormality and hydroureter), despite evidence that in many cases IT was used during most of the first trimester (Rosa et al., '86). It is likely that these discrepencies are related and are due to differences in the pharmacokinetic metabolism of IT between humans and experimental animals. During normal chronic dosing with IT in the hu-
Received February 20, 1990; accepted August 7, 1990. Address reprint requests to H. E. Ritchie, Department of Anatomy, University of Sydney, Sydney NSW 2006, Australia.
72
H. RITCHIE AND W.S. WEBSTER
man, there are relatively constant trough blood levels of IT (132-186 nglml) and its main metabolite 4-oxo-IT (610-791 ng/ml) (Brazzell et al., '83). The accumulation of high levels of the metabolite is due to its long half-life of 24.5 h compared with 10.4 h for IT (Brazzell et al., '83). By contrast, the half-life of IT in the mouse is 19 min (Kalin et al., '82) and 0.8-0.9 h in the rat (Shelley et al., '82). The rapid metabolism of IT by rats and mice means that they have to be given relatively high doses of the drug for it to be present at a critical concentration for long enough to damage the embryos. A side effect of these high doses is the occurrence of very high peak blood levels of IT and its main metabolite, at least 15 times higher than levels in the human. For example, a teratogenic dose of IT in the mouse (100 mg/ kg) gave a blood level of 15,000 ng IT/ml of serum 1 h later (Kochhar et al., '87). Exposure of rodent embryos t o these high levels may damage other embryonic processes that are not affected by the lower concentration of retinoids that occur in the human. This may explain why limb defects are regularly induced by IT administration during limb development in pregnant mice (Kochhar et al., '84; Kochhar and Penner, '87) but rarely in the human despite exposure to IT during the period of limb development (Rosa et al., '86). This does not exclude the possibility that IT may induce limb defects in the human if the maternal blood retinoid levels are high enough. In this respect, it is of interest that there is a recent unpublished report of an abortus from a woman taking IT, which had severe abnormalities typical of the IT embryopathy as well as a minor limb reduction defect (unilateral absent thumb) (E.J. Lammer, '90, personal communication). One way to circumvent the problems associated with differing metabolic rates between species is to test drugs in vitro. Whole rodent embryo culture offers a way of exposing experimental embryos to the same drug levels that occur in the human. When IT or its main metabolite 4-0x0-IT was tested in vitro for 48 h, it was found that both were teratogenic at a concentration of 500 nglml with 250 ng/ml being a no-effect level (Webster et al., '86). Furthermore, both compounds induced the same malformation, namely underdevelopment of the visceral arches, particularly the second visceral arch. Since the second arch forms most of
the external ear, the in vitro system predicted the most common facial malformation seen in the IT embryopathy. The observation that IT and 4-oxo-IT induced the same malformations a t similar concentrations suggests that both compounds contribute to the human IT embryopathy, since both are present in human plasma during normal dosing at a combined trough concentration of 742-977 ng/ml (Brazzell et al., '83). For the purposes of this study 500 ng IT/ml is considered the threshold concentration for the induction of second arch defects in rat embryo culture. The aims of the present study using the in vitro model were first to establish the minimum period of exposure to IT, at the threshold concentration of 500 ng/ml, that will induce abnormalities of the second visceral arch and hence establish the most sensitive period of visceral arch development. Second, t o determine the relationship between teratogenesis and the concentration, duration and timing of exposure to IT. Some of this work has previously been presented in abstract form (Ritchie and Webster, '89). MATERIALS AND METHODS
For rat embryo culture, female SpragueDawley strain rats were housed with males overnight and examined for presence of sperm in the vaginal smear the next morning (day of positive smear is day 0). Embryos were explanted during the afternoon of day 9 of gestation and cultured by the technique of New ('73). Only those embryos at mid to late head-fold stage were used. Five embryos were cultured in 5 ml of media (100% heat-inactivated rat serum, 50 kg/ml streptomycin, and 50 p.g/ml penicillin G ) in each roller bottle. The cultures were initially equilibrated with a gas mixture of 5% 0,-5% C0,-90% N2. After 16 h, the gas mixture was changed to 20% 02-5% CO, 75% N,. Bottles were regassed with this mixture at 22 h and after 40 h with 40% 02-5% co2-55% N,. Experimental embryos were grown in media containing 500, 1,000, 2,000, 4,000, or 16,000 ng/ml IT, dissolved in sunflower oil (isotretinoin, Ro-43780, generously provided by Roche Products Pty Ltd, Australia). Three stock solutions of IT were prepared: 125 kg/ml, 500 kg/ml, and 2 mgiml; and 20-80 PI aliquots were added to the culture serum. The exact volume added may be slightly less because of the viscosity of the oil vehicle. The
ISOTRETINOIN TERATOGENICITY IN RAT EMBRYO CULTURE
solubility of IT in oil (Arachis) is 3 mg/ml (Roche Products manufacturer’s facts sheet). The stock solutions were either made fresh (highest concentration) or were aliquoted and stored in the dark at -20°C. IT was visibly dissolved in all solutions used. Preliminary experiments confirmed the results of Kitchin and Ebron (’84) that in rat embryo culture, concentrations of vegetable oil in serum up to 2.5% were not associated with a significant increase in malformations or a significant change in growth parameters such as crown-rump length and protein content. In the present experiments the maximum concentration of oil in serum was 1.6%. Oil was not subsequently administered to the control group. To determine the sensitive period of IT teratogenicity, embryos were exposed to IT for 48 h or for specific periods of 24, 12,6, 3, or 1.5 h, depending on previous experimental results. For exposure periods of less than 48 h, embryos were carefully removed at the end of the exposure period and washed in three changes of Hank‘s balanced salt solution (HBSS) before being replaced in fresh serum and regassed. Preliminary experiments showed that transferring embryos had no effect on the subsequent development of the embryo. To prevent photodecomposition of IT the culture bottles were completely enclosed in aluminum foil. All embryos were allowed to develop for 48 h and the embryos examined using a dissecting microscope. The parameters used to assess development were based on those by Brown and Fabro (’81) and included the presence of heart beat and yolk sac circulation, turning, arch development, crownrump length, yolk sac diameter, and somite number. Somite counts were only performed on embryos which had turned completely. The data on frequency of arch malformations, normal turning and vitality were examined using x2. Data on yolk sac diameter, crown-rump length, and somite number were examined using Student’s t-test. In each case, significance was taken at 0.01, to avoid type I errors. RESULTS
Morphological defects During the 48-h culture period, the predominant abnormality associated with IT exposure was reduction in the size of the visceral arches, particularly the second arch (11). Measurement of the arches was not at-
73
tempted as fusion between adjacent arches often made it difficult to determine their precise extent. For the purposes of tabulation and for further statistical analysis, abnormal embryos were assigned to one or more of five groups. These five groups were selected in an attempt to identify the principal teratogenic responses to IT at this stage of development. Some embryos had more than one malformation and hence appear in more than one group. The embryos in each group are expressed as a percentage of the total number of embryos in that treatment. Group 1 included all embryos with a ‘severe’ (approximately greater than 50%) reduction in the apparent size of I1 (Fig. 1BE). The first arch was usually normal in appearance except in the most affected embryos (Fig. 1B) when it was also reduced. In most of the embryos in this group, the otic vesicle appeared to be rostrally displaced with respect t o 11. However, its position appeared normal with respect t o the heart, Group 1 embryos were considered particularly significant, since in vivo studies had previously shown that these types of arch defects lead to marked abnormalities of the external ear (Webster et al., ’86). Group 2 included all embryos with a ‘moderate’ change in the size and/or shape of I1 (Fig. lF, Fig. 2B-E). No displacement of the otic vesicle was apparent in these embryos. The changes in I1 were manifested in a number of ways. Many embryos in this group had a I1 which appeared to be thinner and shorter (Fig. 2D) than in controls (Fig. 2A). In other embryos, adjacent first and second visceral arches were of almost normal size but appeared t o be completely fused (Fig. lF), and sometimes the boundaries between the first three arches were indistinct (Fig. 2E). Other embryos had a normal sized I1 but its distal end was turned caudally (Fig. 2B) compared with controls (Fig. 2A). A further small group of embryos had a 11 which appeared to consist of proximal and distal parts (Fig. 2 0 . The embryos shown in Figure 2B,E were only observed when cultured with IT for the last 24 h of culture while those in Figure 2C were only seen when embryos were cultured for 3 h with 4,000 ng/ml, which was the highest concentration used for this duration. Group 3 consisted of those embryos from groups 1 and 2, which in addition to their visceral arch defects had a markedly re-
74
H. RITCHIE AND W.S. WEBSTER
Fig. 1.
ISOTRETINOIN TERATOGENICITY IN RAT EMBRYO CULTURE
duced gap between the frontonasal process and the first visceral arch (Fig. 1B-El. Group 4 consisted of embryos with apparently normal visceral arches but the same markedly reduced gap between the frontonasal process and the first visceral arch (Fig. 2F) as seen in group 3. Group 5 included all other malformations. This small group consisted of embryos with open neural tubes or severe forebrain reduction. Embryos which had failed to turn normally were listed separately. Visceral arch malformations were not always identical on both sides of the embryo. In 9% of embryos with arch malformations there was a marked asymmetry. Of these embryos with asymmetry of 11, 67% were more affected on the right side. Incidence of defects To establish the critical period for second arch abnormalities embryos were exposed to a range of concentrations of IT for 48 h or for progressively shorter periods. Tables 1-5 show the results when data were grouped according to duration of exposure to IT (48, 24, 12, 6, 3, and 1.5 h, respectively).
48-h exposure When embryos were cultured for 48 h in IT at the lowest tested concentration (500 ng/ml), 77% of the embryos had severe malformation of I1 and 21% had moderate ab-
75
normalities (Table 1).All the ‘severe’ embryos had a reduced gap between the first arch and the frontonasal process. Only one embryo had completely normal visceral arches. Other parameters, such as yolk sac circulation, crown-rump length, and somite numbers, were also affected. At higher concentrations of IT, all embryos had ‘severe’ reduction of 11. The embryos were small and generally retarded. The absence of a yolk sac circulation was not the cause of abnormality of 11. Both normal and abnormal embryos were seen with a normal yolk sac circulation at the end of the culture period. Visceral arch defects were not seen in the control embryos.
24-h exposure The results (Table 1) showed that exposure to IT for the first 24 h of the culture period produced approximately the same incidence of abnormalities as seen with 48 h exposure to the same concentration. Exposure during the second 24 h was much less effective. Few of the embryos had a severe defect of 11, although some had moderate reduction of I1 in a concentration-dependent manner. Other parameters were not significantly different from controls. 12-hexposure
The first 24 h of the culture period were further analyzed by exposing embryos to IT for either the first or second 12 h of culture (Table 2). The results showed that exposure for the first 12 h was just as damaging as exposure for the first 24 h or the entire 48 h Fig. 1. All specimens are in vitro 11-day-oldrat em- at the same concentration. Exposure for the bryos. A Control embryo. x 33. B: Embryo with appar- second 12 h had much less effect on the visent absence of first and second visceral arches and malformation of the rostra1 part of the brain. This embryo ceral arches and other parameters were, in was exposed to 2,000 ng/ml IT for the first 24 h of cul- general, not significantly different from ture. x 37. C: Embryo with a wide first visceral arch (as controls. indicated by dotted line) and apparent absence of I1 or fusion of I and 11. The embryo has a reduced gap between the frontonasal process and the first visceral arch and the otic vesicle (ot) is displaced rostrally. This embryo was exposed to 500 ng/ml IT for 48 h of culture. x 50. D Embryo with apparent absence of 11, displacement of the otic vesicle, and reduced gap between the frontonasal process and the first visceral arch. This embryo was exposed to 2,000 ng/ml IT for 48 h of culture. x 31. E: Embryo with a reduced I1 fused with the first visceral arch, displacement of the otic vesicle, and reduced gap between the frontonasal process and the first visceral arch. This embryo was exposed to 1,000 ng/ml IT for 48 h of culture. x 52. F Embryo with fused I and I1 with little reduction in size of the individual arches. Otic vesicle is placed normally and the gap between the frontonasal process, and the first visceral arch is of normal size. This embryo was exposed to 4,000 ng/ml IT for the first 3 h of culture. x 32.
6-hexposure The first 12 h were further analyzed by exposing embryos to IT for either the first or second 6 h of culture (Table 3). The results showed that a more marked effect on the visceral arches was associated with exposure during the first 6 h. To determine whether there was a n earlier adjacent period of sensitivity, some embryos were cultured 6 h earlier than normal and grown for a total of 54 h (appears as -6-0 in Table 3). These embryos were not significantly smaller than the controls from the usual 48-h culture period. Embryos
76
H. RITCHIE AND W.S. WEBSTER
Fig. 2.
77
ISOTRETINOIN TERATOGENICITY IN RAT EMBRYO CULTURE TABLE 1 . Effects of 48-and 24-h culture with isotretinoin on embryonic growth and development
Treat-
Concn
(h)
serum) embryos viability'
merit (ngITim1
No.
%
Yolk sac diameter (mm t
SD)
Crownrump length (mm t
%
SD) turned
Somites (No 5
SD)
%
normal
Morphological defect (%)
embryos Group 1 Group 2 Group 3 Group 4 Group 5
0-48
0
143
99
4.1 ? 0.6 3.6
26.2 t 1.1
91
0
0
0
8
1
0-48 0-48 0-48
500 1,000 2,000
48 26 9
63' 15' 56'
3.8 i 0.5* 3.3 t 0.5' 3.1 t 0.5* 2.5 t 0.4' 3.7 k 0.7 2 . 6 t 0.7'
89 22.9 ? 4.2' 12' 22.3 i 0.6' 11' 20.0 t O.O*
2' 0'
77' 100' 100'
21' 0 0
73' 100' 100'
0 0 0
3 0 20
0-24 0-24 0-24
500 1,000 2,000
9 11 10
67' 9+ 10'
4.3 t 0.4 3.3 t 0.4 3.3? 0.5' 2.5 ? 0.4' 3.3 t 1.0' 2.8 t 0.6'
67' 36' 30'
25.7 t 1.4 20.3 -t 4.7' 19.3 2 2.3'
0'
67' 100' 100'
33'
0 0
0
67' 100' 100'
0 0 10
24-48 24-48 24-48
500 1,000 2,000
10 11 9
4.5 -C 0.3 3.5 t 0.2 3.4 t 0.9' 3.0 t 0.6' 4.2 2 0.7 3.9 t 0.3
100 100 100
25.9? 2.3 25.9 t 0.6 25.7 t 1.4
90 36'
0 36' 0
10 18' 100t
11 0 22'
90 82 100
?
0.4
97
0'
0' 0'
0'
0
0 0
0
20
11* 22*
0
'Viability based on embryos having a heart beat and yolk sac circulation a t the end of the culture period *P
Parameters Determining lsotretinoin Teratogenicity in Rat Embryo Culture HELEN RITCHIE AND WILLIAM S. WEBSTER Department of Anatomy, University of Sydney, Sydney NSW 2006, Australia
ABSTRACT
At the in vitro threshold serum concentration of 500 ng/ml, isotretinoin induces defects of visceral arch development in 9.5-day rat embryos grown in culture for 48 h. Experiments were performed to determine the minimum period of exposure necessary to induce these arch defects and whether an increase in concentration of isotretinoin could compensate for reduced exposure time. The results showed that a minimum 6-h exposure to 500 ng/ml immediately prior to cranial neural crest migration was necessary to induce severe defects of the second visceral arch in a majority of embryos. Maximal increase in isotretinoin concentration t o 16,000 ng/ml did not compensate for shorter exposure periods. These results suggest that to cause malformations of the visceral arches, the embryo must be exposed to isotretinoin for a minimum period of time regardless of the concentration of isotretinoin above the threshold.
The isotretinoin (IT) or retinoic acid embryopathy is a restricted pattern of malformation seen in 22-40% of the offspring of women taking the drug Accutane during the first trimester of pregnancy (Lammer et al., '85; Rosa et al., '86). The embryopathy is characterized by facial dysmorphia, particularly reduced and abnormal external ears. There may also be a flat nasal bridge, mandibular underdevelopment and cleft palate. Many of the children also have conotruncal heart defects, interrupted aortic arch, and a variable deficiency of the thymus. Brain defects are another common feature and include hydrocephalus associated with hypoplasia of the cerebellum with or without agenesis of the cerebellar vermis. The association of facial defects, great vessel and conotruncal malformations and thymic deficiency in the human IT embryopathy indicates a common pathogenesis, since these structures all require a contribution from the cranial neural crest for their normal development (Johnston, '66; Noden, '75; Le Lievre and Le Douarin, '75; Bockman and Kirby; '84; Kirby et al., '83, '85). In vivo animal models of the IT embryopathy can duplicate many of the malformations seen in the human (Webster et al., '86; Willhite et al., '86), but the models are un-
0 1991 WILEY-LISS, INC.
satisfactory for two main reasons. First, depending on species, laboratory animals require 20-400 times the human daily IT dose to induce the malformations typical of the human embryopathy (Kamm, '82; Webster et al., '86; Willhite et al., '86). Second, in laboratory animals, a much wider range of malformations is induced, including limb (Kochhar et al., '84, '87), urogenital, and gastrointestinal defects (Willhite and Shealy, '84). By contrast, in a review of the outcome of human IT pregnancies, there were no substantiated reports of major defects affecting the limbs and only two minor defects affecting the gastrointestinal and urogenital systems (unspecified hepatic abnormality and hydroureter), despite evidence that in many cases IT was used during most of the first trimester (Rosa et al., '86). It is likely that these discrepencies are related and are due to differences in the pharmacokinetic metabolism of IT between humans and experimental animals. During normal chronic dosing with IT in the hu-
Received February 20, 1990; accepted August 7, 1990. Address reprint requests to H. E. Ritchie, Department of Anatomy, University of Sydney, Sydney NSW 2006, Australia.
72
H. RITCHIE AND W.S. WEBSTER
man, there are relatively constant trough blood levels of IT (132-186 nglml) and its main metabolite 4-oxo-IT (610-791 ng/ml) (Brazzell et al., '83). The accumulation of high levels of the metabolite is due to its long half-life of 24.5 h compared with 10.4 h for IT (Brazzell et al., '83). By contrast, the half-life of IT in the mouse is 19 min (Kalin et al., '82) and 0.8-0.9 h in the rat (Shelley et al., '82). The rapid metabolism of IT by rats and mice means that they have to be given relatively high doses of the drug for it to be present at a critical concentration for long enough to damage the embryos. A side effect of these high doses is the occurrence of very high peak blood levels of IT and its main metabolite, at least 15 times higher than levels in the human. For example, a teratogenic dose of IT in the mouse (100 mg/ kg) gave a blood level of 15,000 ng IT/ml of serum 1 h later (Kochhar et al., '87). Exposure of rodent embryos t o these high levels may damage other embryonic processes that are not affected by the lower concentration of retinoids that occur in the human. This may explain why limb defects are regularly induced by IT administration during limb development in pregnant mice (Kochhar et al., '84; Kochhar and Penner, '87) but rarely in the human despite exposure to IT during the period of limb development (Rosa et al., '86). This does not exclude the possibility that IT may induce limb defects in the human if the maternal blood retinoid levels are high enough. In this respect, it is of interest that there is a recent unpublished report of an abortus from a woman taking IT, which had severe abnormalities typical of the IT embryopathy as well as a minor limb reduction defect (unilateral absent thumb) (E.J. Lammer, '90, personal communication). One way to circumvent the problems associated with differing metabolic rates between species is to test drugs in vitro. Whole rodent embryo culture offers a way of exposing experimental embryos to the same drug levels that occur in the human. When IT or its main metabolite 4-0x0-IT was tested in vitro for 48 h, it was found that both were teratogenic at a concentration of 500 nglml with 250 ng/ml being a no-effect level (Webster et al., '86). Furthermore, both compounds induced the same malformation, namely underdevelopment of the visceral arches, particularly the second visceral arch. Since the second arch forms most of
the external ear, the in vitro system predicted the most common facial malformation seen in the IT embryopathy. The observation that IT and 4-oxo-IT induced the same malformations a t similar concentrations suggests that both compounds contribute to the human IT embryopathy, since both are present in human plasma during normal dosing at a combined trough concentration of 742-977 ng/ml (Brazzell et al., '83). For the purposes of this study 500 ng IT/ml is considered the threshold concentration for the induction of second arch defects in rat embryo culture. The aims of the present study using the in vitro model were first to establish the minimum period of exposure to IT, at the threshold concentration of 500 ng/ml, that will induce abnormalities of the second visceral arch and hence establish the most sensitive period of visceral arch development. Second, t o determine the relationship between teratogenesis and the concentration, duration and timing of exposure to IT. Some of this work has previously been presented in abstract form (Ritchie and Webster, '89). MATERIALS AND METHODS
For rat embryo culture, female SpragueDawley strain rats were housed with males overnight and examined for presence of sperm in the vaginal smear the next morning (day of positive smear is day 0). Embryos were explanted during the afternoon of day 9 of gestation and cultured by the technique of New ('73). Only those embryos at mid to late head-fold stage were used. Five embryos were cultured in 5 ml of media (100% heat-inactivated rat serum, 50 kg/ml streptomycin, and 50 p.g/ml penicillin G ) in each roller bottle. The cultures were initially equilibrated with a gas mixture of 5% 0,-5% C0,-90% N2. After 16 h, the gas mixture was changed to 20% 02-5% CO, 75% N,. Bottles were regassed with this mixture at 22 h and after 40 h with 40% 02-5% co2-55% N,. Experimental embryos were grown in media containing 500, 1,000, 2,000, 4,000, or 16,000 ng/ml IT, dissolved in sunflower oil (isotretinoin, Ro-43780, generously provided by Roche Products Pty Ltd, Australia). Three stock solutions of IT were prepared: 125 kg/ml, 500 kg/ml, and 2 mgiml; and 20-80 PI aliquots were added to the culture serum. The exact volume added may be slightly less because of the viscosity of the oil vehicle. The
ISOTRETINOIN TERATOGENICITY IN RAT EMBRYO CULTURE
solubility of IT in oil (Arachis) is 3 mg/ml (Roche Products manufacturer’s facts sheet). The stock solutions were either made fresh (highest concentration) or were aliquoted and stored in the dark at -20°C. IT was visibly dissolved in all solutions used. Preliminary experiments confirmed the results of Kitchin and Ebron (’84) that in rat embryo culture, concentrations of vegetable oil in serum up to 2.5% were not associated with a significant increase in malformations or a significant change in growth parameters such as crown-rump length and protein content. In the present experiments the maximum concentration of oil in serum was 1.6%. Oil was not subsequently administered to the control group. To determine the sensitive period of IT teratogenicity, embryos were exposed to IT for 48 h or for specific periods of 24, 12,6, 3, or 1.5 h, depending on previous experimental results. For exposure periods of less than 48 h, embryos were carefully removed at the end of the exposure period and washed in three changes of Hank‘s balanced salt solution (HBSS) before being replaced in fresh serum and regassed. Preliminary experiments showed that transferring embryos had no effect on the subsequent development of the embryo. To prevent photodecomposition of IT the culture bottles were completely enclosed in aluminum foil. All embryos were allowed to develop for 48 h and the embryos examined using a dissecting microscope. The parameters used to assess development were based on those by Brown and Fabro (’81) and included the presence of heart beat and yolk sac circulation, turning, arch development, crownrump length, yolk sac diameter, and somite number. Somite counts were only performed on embryos which had turned completely. The data on frequency of arch malformations, normal turning and vitality were examined using x2. Data on yolk sac diameter, crown-rump length, and somite number were examined using Student’s t-test. In each case, significance was taken at 0.01, to avoid type I errors. RESULTS
Morphological defects During the 48-h culture period, the predominant abnormality associated with IT exposure was reduction in the size of the visceral arches, particularly the second arch (11). Measurement of the arches was not at-
73
tempted as fusion between adjacent arches often made it difficult to determine their precise extent. For the purposes of tabulation and for further statistical analysis, abnormal embryos were assigned to one or more of five groups. These five groups were selected in an attempt to identify the principal teratogenic responses to IT at this stage of development. Some embryos had more than one malformation and hence appear in more than one group. The embryos in each group are expressed as a percentage of the total number of embryos in that treatment. Group 1 included all embryos with a ‘severe’ (approximately greater than 50%) reduction in the apparent size of I1 (Fig. 1BE). The first arch was usually normal in appearance except in the most affected embryos (Fig. 1B) when it was also reduced. In most of the embryos in this group, the otic vesicle appeared to be rostrally displaced with respect t o 11. However, its position appeared normal with respect t o the heart, Group 1 embryos were considered particularly significant, since in vivo studies had previously shown that these types of arch defects lead to marked abnormalities of the external ear (Webster et al., ’86). Group 2 included all embryos with a ‘moderate’ change in the size and/or shape of I1 (Fig. lF, Fig. 2B-E). No displacement of the otic vesicle was apparent in these embryos. The changes in I1 were manifested in a number of ways. Many embryos in this group had a I1 which appeared to be thinner and shorter (Fig. 2D) than in controls (Fig. 2A). In other embryos, adjacent first and second visceral arches were of almost normal size but appeared t o be completely fused (Fig. lF), and sometimes the boundaries between the first three arches were indistinct (Fig. 2E). Other embryos had a normal sized I1 but its distal end was turned caudally (Fig. 2B) compared with controls (Fig. 2A). A further small group of embryos had a 11 which appeared to consist of proximal and distal parts (Fig. 2 0 . The embryos shown in Figure 2B,E were only observed when cultured with IT for the last 24 h of culture while those in Figure 2C were only seen when embryos were cultured for 3 h with 4,000 ng/ml, which was the highest concentration used for this duration. Group 3 consisted of those embryos from groups 1 and 2, which in addition to their visceral arch defects had a markedly re-
74
H. RITCHIE AND W.S. WEBSTER
Fig. 1.
ISOTRETINOIN TERATOGENICITY IN RAT EMBRYO CULTURE
duced gap between the frontonasal process and the first visceral arch (Fig. 1B-El. Group 4 consisted of embryos with apparently normal visceral arches but the same markedly reduced gap between the frontonasal process and the first visceral arch (Fig. 2F) as seen in group 3. Group 5 included all other malformations. This small group consisted of embryos with open neural tubes or severe forebrain reduction. Embryos which had failed to turn normally were listed separately. Visceral arch malformations were not always identical on both sides of the embryo. In 9% of embryos with arch malformations there was a marked asymmetry. Of these embryos with asymmetry of 11, 67% were more affected on the right side. Incidence of defects To establish the critical period for second arch abnormalities embryos were exposed to a range of concentrations of IT for 48 h or for progressively shorter periods. Tables 1-5 show the results when data were grouped according to duration of exposure to IT (48, 24, 12, 6, 3, and 1.5 h, respectively).
48-h exposure When embryos were cultured for 48 h in IT at the lowest tested concentration (500 ng/ml), 77% of the embryos had severe malformation of I1 and 21% had moderate ab-
75
normalities (Table 1).All the ‘severe’ embryos had a reduced gap between the first arch and the frontonasal process. Only one embryo had completely normal visceral arches. Other parameters, such as yolk sac circulation, crown-rump length, and somite numbers, were also affected. At higher concentrations of IT, all embryos had ‘severe’ reduction of 11. The embryos were small and generally retarded. The absence of a yolk sac circulation was not the cause of abnormality of 11. Both normal and abnormal embryos were seen with a normal yolk sac circulation at the end of the culture period. Visceral arch defects were not seen in the control embryos.
24-h exposure The results (Table 1) showed that exposure to IT for the first 24 h of the culture period produced approximately the same incidence of abnormalities as seen with 48 h exposure to the same concentration. Exposure during the second 24 h was much less effective. Few of the embryos had a severe defect of 11, although some had moderate reduction of I1 in a concentration-dependent manner. Other parameters were not significantly different from controls. 12-hexposure
The first 24 h of the culture period were further analyzed by exposing embryos to IT for either the first or second 12 h of culture (Table 2). The results showed that exposure for the first 12 h was just as damaging as exposure for the first 24 h or the entire 48 h Fig. 1. All specimens are in vitro 11-day-oldrat em- at the same concentration. Exposure for the bryos. A Control embryo. x 33. B: Embryo with appar- second 12 h had much less effect on the visent absence of first and second visceral arches and malformation of the rostra1 part of the brain. This embryo ceral arches and other parameters were, in was exposed to 2,000 ng/ml IT for the first 24 h of cul- general, not significantly different from ture. x 37. C: Embryo with a wide first visceral arch (as controls. indicated by dotted line) and apparent absence of I1 or fusion of I and 11. The embryo has a reduced gap between the frontonasal process and the first visceral arch and the otic vesicle (ot) is displaced rostrally. This embryo was exposed to 500 ng/ml IT for 48 h of culture. x 50. D Embryo with apparent absence of 11, displacement of the otic vesicle, and reduced gap between the frontonasal process and the first visceral arch. This embryo was exposed to 2,000 ng/ml IT for 48 h of culture. x 31. E: Embryo with a reduced I1 fused with the first visceral arch, displacement of the otic vesicle, and reduced gap between the frontonasal process and the first visceral arch. This embryo was exposed to 1,000 ng/ml IT for 48 h of culture. x 52. F Embryo with fused I and I1 with little reduction in size of the individual arches. Otic vesicle is placed normally and the gap between the frontonasal process, and the first visceral arch is of normal size. This embryo was exposed to 4,000 ng/ml IT for the first 3 h of culture. x 32.
6-hexposure The first 12 h were further analyzed by exposing embryos to IT for either the first or second 6 h of culture (Table 3). The results showed that a more marked effect on the visceral arches was associated with exposure during the first 6 h. To determine whether there was a n earlier adjacent period of sensitivity, some embryos were cultured 6 h earlier than normal and grown for a total of 54 h (appears as -6-0 in Table 3). These embryos were not significantly smaller than the controls from the usual 48-h culture period. Embryos
76
H. RITCHIE AND W.S. WEBSTER
Fig. 2.
77
ISOTRETINOIN TERATOGENICITY IN RAT EMBRYO CULTURE TABLE 1 . Effects of 48-and 24-h culture with isotretinoin on embryonic growth and development
Treat-
Concn
(h)
serum) embryos viability'
merit (ngITim1
No.
%
Yolk sac diameter (mm t
SD)
Crownrump length (mm t
%
SD) turned
Somites (No 5
SD)
%
normal
Morphological defect (%)
embryos Group 1 Group 2 Group 3 Group 4 Group 5
0-48
0
143
99
4.1 ? 0.6 3.6
26.2 t 1.1
91
0
0
0
8
1
0-48 0-48 0-48
500 1,000 2,000
48 26 9
63' 15' 56'
3.8 i 0.5* 3.3 t 0.5' 3.1 t 0.5* 2.5 t 0.4' 3.7 k 0.7 2 . 6 t 0.7'
89 22.9 ? 4.2' 12' 22.3 i 0.6' 11' 20.0 t O.O*
2' 0'
77' 100' 100'
21' 0 0
73' 100' 100'
0 0 0
3 0 20
0-24 0-24 0-24
500 1,000 2,000
9 11 10
67' 9+ 10'
4.3 t 0.4 3.3 t 0.4 3.3? 0.5' 2.5 ? 0.4' 3.3 t 1.0' 2.8 t 0.6'
67' 36' 30'
25.7 t 1.4 20.3 -t 4.7' 19.3 2 2.3'
0'
67' 100' 100'
33'
0 0
0
67' 100' 100'
0 0 10
24-48 24-48 24-48
500 1,000 2,000
10 11 9
4.5 -C 0.3 3.5 t 0.2 3.4 t 0.9' 3.0 t 0.6' 4.2 2 0.7 3.9 t 0.3
100 100 100
25.9? 2.3 25.9 t 0.6 25.7 t 1.4
90 36'
0 36' 0
10 18' 100t
11 0 22'
90 82 100
?
0.4
97
0'
0' 0'
0'
0
0 0
0
20
11* 22*
0
'Viability based on embryos having a heart beat and yolk sac circulation a t the end of the culture period *P
