Groundbreaking Salk study shows de-differentiation of mature neurons & astrocytes into glioma forming cells

A team from the Salk Institute in La Jolla led by Inder Verma has reported in a paper in Science some hugely important findings that I believe make their paper in the top 10 as a candidate for paper of the year.

The paper, entitled “Dedifferentiation of Neurons and Astrocytes by Oncogenes Can Induce Gliomas in Mice” makes striking findings.

The authors show that mature neurons and astrocytes (incredibly specialized, “terminally differentiated” cells) can be dedifferentiated. This is groundbreaking because these types of cells, especially the mature neurons, were thought to be really fixed in their fate and hence possessing much plasticity.

The dedifferentiated cells could also lead to formation of malignant brain tumors called gliomas. Interestingly, the glioma cells exhibited many hallmarks of stem cells.

An important open question is whether the neurons are dedifferentiated first as a discrete step back to a neural stem cell or even more primitive stem cell state, which then acts like a cancer stem cell or if partial dedifferentiation is enough to kick start the tumorigenic process. In this regard, the authors state:

We propose that the genetically altered differentiated cell acquires the capacity to dedifferentiate to a more progenitor (stem cell) state, and that tumor progression probably requires a permissive microenvironment composed of cell types and molecular signals that can sustain both differentiation of tumor cells as well as maintenance of tumor stem-like cells. 

Another key open question is whether these events that occur in mice also happen during brain tumor formation in people. I suspect that they do.

To trigger the dedifferentiation, the authors used a single lentiviral vector to produce virus injected into the mouse brain. They write:

…we have expanded the utility of our lentiviral system by generating a new construct that carries two shRNAs: one targeting neurofibromatosis type I gene (NF1: mutated in 18% of GBMs) and the other one targeting p53 (mutated in over 35% of GBMs)…

A picture of this vector from the paper is below.

Inder Verma

In a The Scientist article on the paper, Verma is quoted as follows:

“What we’re saying is, any cell in the brain that gets an oncogenic insult has the ability to dedifferentiate [and form tumors],” said Verma. This might seem a rather bleak outlook, but “by knowing the mechanism, we at least have a handle to start thinking about [treatments],” Verma said.

I agree that knowing the mechanism is key here and is actually encouraging. It is difficult to fight diseases that we do not understand. I believe this paper is a major advance for the cancer and stem cell fields.



  1. If I understand this correctly, then this would support the EMT model correct? I, however, have trouble reconciling the EMT model and the CSC model because the cells that become progenitors from the EMT model, based on what I have read but maybe wrong, do not fully exhibit all of the characteristics of CSCs. Based on this paper, do you feel that future drugs that are CSC targeting cannot be entirely effective if they also do not target progeny that exhibit stem-like characteristics?

    • Also, I am curious to hear your opinion of proposals of de-differentiating cancer cells from tumors in an effort to combat cancer. Do you view that as a feasible option? I personally feel that due to the reprogramming shortcomings that currently plague iPSCs, the resulting cells will still exhibit some characteristics of cancer and will not be fully genetically similar to healthy cells of the same cell type. As such, this idea is risky because now you have created a stem cell with cancer characteristics that can now potentially cause tumor formation.

    There would appear to be overwhelming evidence that OXYGEN Tension is the PRIMARY differentiating factor, ie that which determines at what stage of development or differentiation (going forward) that cells are at or what stage of DEDIFFERENTIATION (going backward) that cancer cells are at. What if, after transformation of breast duct epithelial cells, angiogenesis (the growth of blood vessels) is PROGRAMMED to lag behind tumor growth, so that as oxygen tension goes down in the tumor, it’s cells increasingly DEDIFFERENTIATE.

    Those furthest from the blood vessels and thus exposed to the LEAST oxygen tension. undergoing EMT (epithelial to mesenchymal transfer) to become cancer stem cells. Then as these cells with a mesenchymal phenotype crawl toward the blood vessels to metastasise they are exposed to increased oxygen tension and revert to an epithelial phenotype?

    This is a repeat of what happens to the normal mesenchymal cells that sit below the normal epithelial cells and presumably are in a more hypoxic environment and thus more ‘stemmie’, a precursor to the epithelial cell.. Come ovulation and the need for a lot more epithelial cells, then perhaps a vasodilating prostaglandin increases the oxygen tension bathing the mesenchymal cells just below the epithelia and they divide whilst differentiating INTO epithelia.

    The question is why all this order? HIF (hypoxia inducable factor) proteins affect 800 genes, 1/13 of all coding genes. That an AWFUL lot of genes for something that is supposed to be a stuff up, eh. Larmakism ??

Comments are closed.