Changing DNA replication errors into key gene regulatory mechanisms.

Even though until 50 years ago scientists believed that all the non-coding regions of the DNA -meaning those that are not translated into proteins- were basically useless and just a relic of evolution, it is now largely known they play a key role in regulating gene expression.       

During DNA replication, errors may occur, in some cases leading to the synthesis of short non-coding DNA regions repeated few to several times (DNA repeats), which are normally extremely unstable.

Being so unstable, could one ever think of them as reliable sequences to modulate critical gene expression programs?

Probably not.

However, it has been recently shown that in evolution these accidentally arising repeats –whose role in cell physiology was still not completely understood- must have somehow conferred an evolutionary advantage, as they have been positively selected, fixed in the genome, and assumed specific roles, becoming part of the regulatory networks of particular gene programs.

The first example of this newly discovered mechanism is the DNA repeats bound by a protein named ZEB1. ZEB1 binding -together with other proteins- to these sequences ultimately results into the inhibition of the gene program modulating a cell process called epithelial-to-mesenchymal-transition (EMT), namely the loss of epithelial features, such as the ability to make junction with neighboring cells or their polarity (intrinsic feature of a cell due to differences in structure and function in different areas of a cell), and gain of mesenchymal features, making them acquire the ability to migrate away from the original site in the tissue (which is, for instance, a primary step of the process ending into metastatic dissemination, and a key process in development). Experiments performed in pancreatic cancer cell lines showed that when these ZEB1-specific DNA repeats are deleted, cells normally showing mesenchymal properties assume instead epithelial features, whereas their migration ability (associated to mesenchymal features) is impaired.

Having said that, two main questions may arise:

Question n1. Is this newly identified DNA-repeats-based regulatory mechanism restricted to the control of EMT in cancer cells? Evidence suggests the existence of an analogous regulatory mechanism in neurogenesis. Indeed, ZEB1 has a role in modulating cell polarity and adhesion, which are two key processes for neural migration and differentiation, and defective neuronal migration may cause brain malformations, including  microencephaly (abnormal small brain); is this control exerted through ZEB1 binding to DNA repeats in neurogenesis as well? It is likely.

Question n2. Why these DNA replication errors have been positively selected? Considering ZEB1, the binding to a specific DNA region controls crucial cell processes; in case of mutations altering this region, ZEB1 binding might get lost, likely prejudicing the related cell process. The emergence of DNA repeats containing multiple ZEB1 binding sites would provide additional regulatory factors to ensure the required gene expression.

Therefore, a region of the genome arising from an error during replication, selected by evolution, becomes a (probably) common regulative mechanism modulating key cellular processes involved in EMT (thus in malignant transformation) and in neurodevelopment. Can we see this as another impressive example of the incredibly great outcomes arising from mistakes when these are retained?

Reference: Co-optation of Tandem DNA Repeats for the Maintenance of Mesenchymal Identity. Balestrieri C, Alfarano G, Milan M, Tosi V, Prosperini E, Nicoli P, Palamidessi A, Scita G, Diaferia GR, Natoli G. Cell. 2018