IN-VITRO EVIDENCE FOR A MULTIFACTORIAL ETIOLOGY WITH LITTLE CONTRIBUTION FROM GLUCOSE PER SE

To determine the extent to which elevated glucose and 3-hydroxybutyrate (30HB) concentrations contribute to the embryotoxic properties of diabetic serum, we tested the effects of serum horn untreated or acutely insulin-treated diabetic rats on the development of mouse embryos during neurulation in vitro. Male Sprague-Dawley rats (n = 143) with streptozocin-induced diabetes for 1 week received infusions of insulin (n = 105) or saline (n = 38) for up to 120 min. The insulin-infused animals were exsanguinated when serum glucose concentrations fell to between 5.6 and 8.3 mM. Saline-infused animals were exsanguinated after a similar duration of infusion. Serum samples were tested for embryotoxic effects on 3-6 somite mouse embryos cultured in vitro for 24 h. Of embryos cultured in serum from untreated diabetic animals (glucose: 24 +/- 1 mM; 30HB: 2.0 +/- 0.3 mM), 36% (31 of 87) exhibited gross malformations, mostly of the neural tube. Only 16% (10 of 62) of embryos grown in serum fi om acutely insulin-treated animals (glucose: 7.4 +/- 0.2 mM; 30HB: 0.20 +/- 0.06 mM) were malformed. This rate that was less than half the rate caused by exposure to diabetic serum (P < 0.01), but a rate that remained much greater than the rate associated with culture in normal serum (0% in this study; <2% historically). In vitro addition of glucose to serum hom insulin-treated animals to re-establish hyperglycemia in the diabetic range (25 mM) resulted in a 17% (12 of 70) malformation rate, nearly identical to the 16% rate caused by normoglycemic serum from insulin-treated animals. Addition of D-30HB to approximate the highest concentration encountered in untreated diabetic serum (similar to 4 mM) increased the malformation rate to 29% (21 of 73 embryos; P = 0.08 vs. unmodified serum from insulin-treated animals). Addition of glucose (to similar to 25 mM) and 30HB (to similar to 4 mM) resulted in a 23% malformation rate, which was not significantly different from the rate caused by unmodified serum from the insulin-treated group. These findings demonstrate that factors other than glucose and 30HB account for much of the teratogenic potential of diabetic serum in this in vitro culture system. Moreover, the data suggest that glucose per se is not a major cause of the embryotoxicity of diabetic serum in this system, whereas 30HB may account for some of the toxic effects.