Immunoteratological Studies of Diabetic Embryopathy Using Gene Expression Analysis
Diabetic embryopathy is a major complication of pregnant women with type I diabetes. Immune defects in the pathogenesis of diabetic embryopathy have been suggested. We hypothesized that activated immune system can counteract diabetic effect and result in prevention of diabetic embryopathy. Diabetes was induced in pregnant ICR mice by streptozocin injection. Three different techniques of maternal immune stimulation, complete Freund's adjuvant (CFA), granulocyte-macrophage colony-stimulating factor (GM-CSF), or interferon-gamma (IFN-g), were used to stimulate the maternal immune system. Approximately 50% of fetuses from hyperglycemic (>27 mM/L) dams were malformed, with neural tube defects predominating. Maternal immune stimulation during the time of normoglycemia, i.e. prior to onset of hyperglycemia, was necessary for reducing teratogenic effects associated with hyperglycemia. The immune-stimulated diabetic mice then produced significantly lower numbers of malformed fetuses: CFA 20.9%, GM-CSF 23.3%, IFN-g 13.9%. A gene microarray was then used to examine a selected panel of placental and splenic genes. We hypothesized that a shared profile of placental or splenic gene expression changes may correlate to the reduced birth defect outcome induced by the different immune stimulation procedures. Diabetes did not cause significant changes in placenta or spleen gene expression profile. In placenta, CFA and GM-CSF changed placental gene expression relative to control or diabetes, but differentially affected such genes relative to each other; further, IFN-g did not affect gene expression relative to control or diabetes. Thus no common pattern of improved placental cytokine, cell-cycle, apoptotic, transcription factor, or other gene expression was identified in the immune-stimulated mice. In spleen, all 3 immune activators produced a common altered gene expression profile. The overall gene expression profile after all immune stimulation procedures suggested increased splenocyte activity and cytokine production. The cytokine GM-CSF, in particular, was up-regulated in splenic leukocytes. This cytokine has previously been associated with reduced cleft palate in urethane-exposed mice after immune stimulation, and with reduced limb malformations in cyclophosphamide-treated mice after intra-uterine administration. In contrast, the TGF-beta3 gene was down-regulated in immune-stimulated diabetic mice. This gene was up-regulated in urethane-exposed mice, an effect that may be associated with reduced cleft palate. Thus unlike urethane, TGF-beta3 gene expression did not show a relationship with reduced diabetes-induced birth defects. Taken together, these data prove our hypotheses and suggest that mechanistically diverse forms of immune activation result in protection against diabetes-related teratogenesis, but only if given prior to onset of hyperglycemia. Such immune stimulation in mice may act through systemic immune organs, i.e. spleen, over-riding adverse effects of diabetes on development.