You may have heard of organoids (probably here on Massive Science) and how they are the next big thing in science, especially when it comes to studying the brain. Sure, cerebral organoids are pretty cool. They're formed when stem cells self-assemble into 3D structures that recapitulate brain development, making organoids a good model to study development as well as disease. But now, on top of the ethical debates brought on by these organoids, new research out today suggests that they may not be as reliable as the field thinks. 

A new study published in Nature thoroughly characterized lab-grown organoids (developed from human cortical cells) by looking at which genes they expressed and compared them with a developing human brain. They found that, compared to the creation of very specific cell types seen in normal developing human brain, organoids had broader cell types. What seems to be missing is the cell sub-type specification. Imagine a tree that grew a generic "fruit" rather than a specific thing like an “orange” or an "apple." 

The organoids also seemed to be more "stressed out" than their normal counterparts. Cellular stress markers that had previously been associated with blocking development of those specific cell types were over-activated in the organoids, but not in the human developing brain. Interestingly, transplanting the organoids into mouse brains blocked the increased cellular stress, and also increased cell type specification in the organoids.

Although this may seem as an unfortunate turn of events, it does not mean that all organoids should be disregarded, but rather that scientists should be more wary when drawing conclusions. As statistician George Box famously said, “All models are wrong, but some are useful.” Knowing these limitations, the field must think carefully about the questions they want to answer using the organoid as a model.