Researchers use AI to rescue embryos from brain defects by re-engineering cellular voltage patterns.

Improper neural patterning during brain development leads to serious disorders such as open neural tube defects, brain malformations, and susceptibility to autism or degenerative disorders. There are currently no clinically approved interventions for brain patterning disorders, therefore, developing repair strategies is a critical unmet need in medicine.  Now, a study from researchers at Tufts University shows electrical patterns in the developing embryo can be predicted, mapped, and manipulated in order to prevent defects caused by harmful substances such as nicotine. The team states their research suggests targeting bioelectric states may be a new treatment modality for regenerative repair in brain development and disease, utilizing artificial intelligence (AI) to find effective repair strategies.  The opensource study is published in the journal Nature Communications.

Previous studies have shown in fertility developing embryonic groups of cells creates patterns of membrane voltage preceding and controlling the expression of genes, with morphological changes occurring over the course of time.  The global medical community is now beginning to see how electrical patterns in the embryo are guiding large-scale patterns of tissues, organs, and limbs. It is highly desirable to decode this electrical communication between cells and use it to normalize development or support regeneration.  The current study uses deep-learning to predict the bioelectrical patterns occurring in normal and nicotine-exposed embryos, to identify reagents possessing the potential to restore normal patterning.

The current study develops a powerful computational simulation platform, called the BioElectric Tissue Simulation Engine (BETSE), to create a dynamic map of voltage signatures in the developing brain of a frog embryo.  Results show the BETSE model accurately replicates the distinct pattern of membrane voltage from normal embryonic brain development and also explains the erased electrical pattern observed due to nicotine exposure.

BETSE was then used to explore the effect of various reagents on the embryo’s voltage map. Data findings show one reagent, the hyperpolarization-activated cyclic nucleotide-gated channel (HCN2), when added to the cells selectively enhanced a large internal negative charge in areas of the brain diminished by nicotine.  Results show expressing HCN2 in live embryos rescues them from the effects of nicotine, restores a normal bio-electric pattern, brain morphology, markers of gene expression, and near regular learning capacity in the grown tadpole.

The team surmises they successfully used AI to map the biophysical mechanism of developmental brain damage to restore the bioelectrical patterns necessary for brain patterning after nicotine exposure.  For the future, the researchers state activation or expression of a specific channel is a general strategy for complex organ patterning in the context of birth defects, regenerative medicine, and bioengineering.

Source: Tufts University

Get Healthinnovations delivered to your inbox: