EXPERIMENTAL MAMMALIAN TERATOLOGY 241 



that it is impossible to foretell which effect is related to the production of a malformation. 



A useful approach to this question involves the use of combinations of teratogens. 

 The assumption is made that two teratogens which in combination give the same 

 frequency of a malformation as that produced by the one with the higher effect given 

 singly (a nonadditive effect) act on the same metabolic pathway to produce the mal- 

 formation, whereas when the combined teratogens produce a frequency of defects 

 which is the sum of the frequencies produced by each one singly, different pathways 

 are involved. This is well demonstrated by the work of Runner and co-workers on 

 the embryonic effects of maternal fasting in mice. Malformations of the vertebrae and 

 ribs were prevented by giving small quantities of glucose or casein during the fasting 

 period, and (to a lesser extent) certain amino acids and acetoacetate. This suggested 

 that protection was provided by supplying substrate for the citric acid cycle. Maternal 

 treatment with insulin, iodoacetate, or a PGA antagonist produced a similar array of 

 malformations, and could reasonably be postulated to interfere with the citric-acid 

 cycle. 1088 Further evidence came from the results of fasting combined with one of a 

 variety of other teratogens. 1090 



When pregnant females were fasted and exposed to hypoxia, the frequency of 

 defects in the offspring was approximately the sum of the frequencies produced by each 

 teratogen separately. This is reasonable on the assumption that the two teratogens 

 decrease two substrates, involving separate metabolic pathways. In Runner's hypo- 

 thetical scheme, hypoxia affects both the ectodermal and mesodermal components of 

 the inductive system leading to differentiation of the axial skeletal system. Fasting in 

 addition reduces the glucose substrate necessary for the ectodermal component, so the 

 combined treatment should (and does) give a higher frequency of defective offspring 

 than either one alone (fasting 24 per cent, hypoxia 47 per cent, combined 75 per cent) . 

 On the other hand, iodoacetate combined with fasting gives almost the same frequency 

 of defects (66 per cent) as iodoacetate alone (62 per cent), as would be expected on the 

 basis that if glycolysis is blocked by the iodoacetate, reducing the glucose substrate 

 by fasting will make no difference, since it involves the same (blocked) pathway. 



Synergistic effects (for fasting and cortisone on cleft palate frequency) 678, 878 are 

 more difficult to interpret, especially if the malformation concerned falls into the class 

 of quasicontinuous variations, 501 as postulated for cleft palate. 405 Here, two agents 

 which altered the threshold in an additive manner could produce an apparently 

 synergistic effect as measured by the frequency of induced malformations, by moving 

 the threshold in from the flat tail of the curve to the more steeply sloping portion. 



5. What are the relations of mother and fetus with respect to teratogens? — The fact that a 

 teratogen must pass through the mother to act on the fetus in most cases is a problem 

 that is not always given sufficient attention. In fact, for many teratogens, it is not clear 

 whether the effects on embryonic development result from direct action of the teratogen 

 on the embryo or from secondary metabolic effects of the teratogen on the mother. 

 Giroud et a/. 435 have reported that maternal riboflavin deficiency in rats is accompanied 

 by a decreased concentration of riboflavin on the maternal liver and an even greater 



