February 18, 2004
New Type Of Stem Cells Found In Adult Human Brain

Human brain astrocyte cells are capable of dividing and turning into all three types of mature brain cells.

February 18, 2004— Researchers have found an unexpected source of stem cells in the adult human brain. They have demonstrated for the first time that human astrocytes — brain cells thought to play more of a secondary role by providing a supportive, nurturing environment for the neuron — can actually function as stem cells. The astrocytes can form new stem cells and are able to generate all three types of mature brain cells.

But these astrocytes are different: They form a novel ribbon-like structure in the brain's lateral ventricle. Stem cells from comparable areas in the rodent brain follow a distinct path from their place of origin to the olfactory bulb (a brain region that processes smells), where they create new neurons.

The work, led by former HHMI medical student fellow Nader Sanai and Arturo Alvarez-Buylla, Heather and Melanie Muss Professor of Neurological Surgery at the University of California, San Francisco, opens the possibility that such stem cells could be harnessed and one day used to regenerate damaged areas in the central nervous system. The scientists reported their findings February 19, 2004, in the journal Nature.

“We've found a structure in the human brain that represents a significant departure from other species,” Sanai said. “The differences we see imply that this region in the human brain doesn't necessarily do the same things as its primate and rodent counterparts. This is a cell population that has the potential to regenerate parts of the brain, though it's not clear what regions those may be. Neurons generated in this area may migrate to other areas of the brain and potentially regenerate those areas.”

What is not clear to me is whether this exact same experiment has been tried in the primate species (chimpanzees and bonobos) that are closest to humans in evolution. Would a repeat of this experiment on chimp astrocytes from the same part of the brain yield the same result? Does anyone know whether this has been tried? I'm not quite ready yet to accept this a feature of neurobiology that is totally unique to humans. Does any reader have enough expertise in the relevant areas of research to answer this question?

What seems surprising about this result is that only now in the year 2004 has anyone even checked to see if astrocytes can become nerve cells.

They studied brain tissue from the lateral ventricles - two cerebrospinal fluid-filled cavities in the center of the brain - available from either surgery patients or from pathology samples after autopsy. The researchers first stained the tissue to locate astrocytes, and immediately saw the ribbon of astrocytes lining the ventricle walls. They subsequently determined that cells within the ribbon were dividing, implying that they were part of a region of proliferative stem cells.

Next, the scientists decided to look for the stem cells. They took representative sections of tissue from the lining of the lateral ventricles, and found that these specimens could generate neurospheres in a dish. Neurospheres contain all of the precursors for the major central nervous system cell types the stem cell produces: neurons, astrocytes, and oligodendrocytes. They result from a stem cell being put in a culture dish with various growth factors.

To make sure, they subsequently isolated individual human astrocytes and put each in a dish with growth factors, showing they could form neurospheres as well.

This was the first time anyone had shown that a single human astrocyte could function as a stem cell. Alvarez-Buylla, Sanai, and their co-workers then found that single astrocytes from the lateral ventricle could generate neurons without added growth factors — direct evidence that a single astrocyte could generate a neuron.

The findings are provocative because astrocytes have traditionally been considered simple helper cells, Sanai said.

“This speaks to the plasticity of the human brain,” he said. “Certain cell types may have hidden potential.” These subtypes of astrocytes appear no different from any other astrocytes, implying that “it's possible that other astrocytes in other regions of the body have the same potential.”

The hippocampus has been previously known as a site in the brain for adult neural stem cells. The existence of stem cells in the brain was first discovered in canaries and this discovery upset the received wisdom for many decades that the adult brain never gained new nerve cells. In fact, it has been previously reported that astrocytes provide growth factors that help hippocampal stem cells convert more rapidly into neurons. Now that it is known that astrocytes can convert into neurons this opens up the potential to stimulate astrocytes to divide and create neurons to do repair for various neurological disorders and even to replace nerve cells lost to aging. This result also means that future development of the ability to replenish astrocytes with rejuvenated replacement cells could turn out to be a useful rejuvenation therapy for the brain.

This particular discovery is also part of a larger pattern of discovery in which new sources of adult stem cells are being found in different parts of the body. It seems likely that many more sources of adult stem cells are still waiting to be discovered.

Share |      Randall Parker, 2004 February 18 11:50 AM  Biotech Organ Replacement

Reason said at February 18, 2004 6:24 PM:

Pretty interesting when combined with this item reported at Betterhumans:


It seems fairly clear that the understand of basic cellular processes and mechanics is entering the lego phase - researchers know just about enough to start putting bricks together to get cool new items. Developments in all fields that build directly atop of this sort of knowledge have been astoundingly rapid for the past two or three years.

Founder, Longevity Meme

Randall Parker said at February 18, 2004 6:48 PM:


I see a lot more work that needs to be done to make this useful. One problem that I kinda alluded to is that there are a lot of different stem cell reservoirs in the body of a large assortment of different partially differentiated types. We need the ability to make each of those types. That represents a lot of work, especially in order to be sure that one has produced exactly the type one wants.

One of the mysteries is just how many different partially differented stem cell types are there? We need to know a lot more about all the different differentiation states. How much of any given state is due to genetic methylation patterns? How much is it due balances of compounds just floating around each cell floating around and keeping the cell in a given control state?

As for the danger of cancer and telomerase: We need the ability to turn up telomerase to rebuild the telomeres but then to turn it back down again before injecting the cells into the body. Aubrey de Grey goes so far as to advocate the removal of telomerase from cells that are going to be used to replenish aged stem cell reservoirs. An open question is just what portion of all cancers come from stem cells gone amok?

In order to prevent stem cells delivered as part of stem cell therapy from going cancerous we also need ways to take whatever cells we are going to use to replenish stem cell reservoirs (or to grow replacement organs) and select ones that have intact and unmutated genes for regulating cell growth. So we need, for instance, a way to check the p53 gene to make sure it is still functioning properly. Ditto for probably dozens of other genes involved in cell growth regulation. One way to do that would be to use gene arrays to test gene expression pattern from cells taken from a batch of rejuvenated stem cells. Look for expression patterns that suggest the potential presence of dangerous mutations.

Then there is yet another problem: Tissue Engineering. A lot of people spend a lot of time debating about the ethics of human embryonic stem cell research. But one reason I spend far less time on the topic is that we need to develop tissue engineering techniques using other species for the same familiar reasons that we use rats and mice and other animals in medical research. Well, there are reports of success in growing extra body parts of rodents. But the work is still pretty preliminary. Scientists still have a lot of problems to solve before new kidneys, livers, pancreases, and other organs can routinely be grown for other animals let alone for humans.

Randall Parker said at February 18, 2004 7:17 PM:

Another point about just how far we are away from where we need to be in stem cell manipulation: When it comes to neurodegenerative disorders it isn't enough just to be able to produce the desired type of stem cell. We need a way to tell it what to do once it is in the body.

Keep in mind that the brain already has at least two kinds of stem cells. So why aren't those stem cells creating the neurons needed to replace lost neurons in Parkinson's? Are they exhausted by too many cell divisions? My guess is that more is involved. The replacements do not know exactly where to go and what configuration to take.

Reason said at February 18, 2004 10:28 PM:

Oh, yes, there's a lot of work to go. I've envisaged the next fifty years in medicine as looking very much like the war on cancer a thousand times over in parallel. There is definitely a lot of work ahead - but the fact that you or I can sit down and make sensible statements (over several paragraphs) about what that work looks like speaks volumes. That's a big difference from biomedical studies in the 70s.

I wrote a little about the fast and slow of things at:


Founder, Longevity Meme

Stephen Gordon said at February 19, 2004 5:51 PM:


I see no conflict between Reason's statement, "researchers know just about enough to start putting bricks together to get cool new items" and your position that there's much left to do.

Surely useful applications will come from stem cell research long before we have anything near full understanding. The Wright brothers didn't make a jumbo jet, but they did get off the ground. :-)

Randall Parker said at February 19, 2004 6:11 PM:


I'm certainly an advocate of taking an engineering approach well before the science is all figured out. We can start using things long before we understand how they work. Look at aspirin for a classic example of that. Plusm there are already therapeutic successes reported with human use of stem cells.

However, having said all that I do believe that most of the future therapeutic uses of stem cells require far more knowledge before we can succeed in making those uses work. I also think that most of those future uses need to be worked out in mouse models first.

As for the whole embyronic stem cell controversy: I think time spent on it is time wasted. By the time most applications for stem cells are worked out in mouse models we will have the ability to turn adult fully differentiated cells into much less differentiated and even into entirely undifferentiated cells. That kind of progress is already happening. Check out reversine for example.

BTW, the discover of reversine checked through 50,000 compounds to find it. To repeat a theme I like to hit on: the ability to parallelize and automate biological researoh is going to accelerate many types of research by orders of magnitude.

stephen.c.bensley said at October 3, 2005 9:21 AM:

Hey,,,,,,,,,,,,,fascinating,wonder if this new discovery could be linked to a similar study in the development of other life,,,or growth,,,,,,,,,,,flowers or buds or other growing things to replenish our own,,,,,,,,,,,,or stimulate our own,,,,,,i.e. young fresh,,,,,,,,,,,,,cathy ranthus roseaus vine, and encourage some sort of hope......

stephen.c.bensley said at October 3, 2005 9:22 AM:

Or even from our own liver???????????????????

Post a comment
Name (not anon or anonymous):
Email Address:
Remember info?

Go Read More Posts On FuturePundit
Site Traffic Info
The contents of this site are copyright ©