Retinoic acid-induced spina bifida: evidence for a pathogenetic mechanism.

Treatment of C57Bl/6J mice with three successive doses of all-trans retinoic acid (28 mg kg-1 body weight) on 8 day, 6 h (8d,6h), 8d,12h, and 8d,18h of gestation resulted in a high incidence (79%, 31/39 fetuses) of spina bifida with myeloschisis (spina bifida aperta) in near term fetuses. Twelve hours following the last maternal dose (9d,6h), the caudal aspects of treated embryos, were abnormal, with eversion of the neural plate at the posterior neuropore, as compared to its normal concavity in comparably staged control specimens. This eversion persisted in affected embryos through the time that the posterior neuropore should normally close. The distribution of cell death in control and experimental embryos was determined using vital staining with Nile blue sulphate and with routine histological techniques. Twelve hours following the maternal dosing regimen, experimental embryos showed evidence of excessive cell death, predominantly in the mesenchyme associated with the primitive streak and in the endoderm of the tail gut, both of which are readily identifiable sites of physiological cell death at this stage of development. In addition, the presumptive trunk neural crest cells located in the dorsal midline, cranial to the posterior neuropore, exhibited a marked amount of cell death in the experimental embryos. We propose that the major factor in the generation of spina bifida in this model is excessive cell death in the tail gut and mesenchyme ventral to the neuroepithelium of the posterior neuropore.(ABSTRACT TRUNCATED AT 250 WORDS)