Pages

Search This Blog

A brain part mouse, part human

10/1/17

Researchers placed human cells into the brains of mice. The human brain cells began to replicate and take over much of the mouse brain, and the effects on mouse behavior were bizarre.  True.  Not science fiction.

Researchers at the University of Rochester in New York figured out a way to put human brain cells into mouse brains, taking steps to ensure that the mouse immune system wouldn't reject the human cells.   The mice they created are 'chimeric', their brains are part mouse, part human. (Chimera were mythical hybrid creatures with body parts from more than one animal.)

In order to appreciate what the researchers did, and what they found, it helps to know about the kind of human brain cells they introduced into the mouse brain.  Like mice, we have two basic types of brain cells: neurons and glia.  The cells they implanted into the mouse brain were human glial cells.

Most people have heard a lot about neurons, but not much about glia.  Glia are sometimes described as the brain's 'support staff'. Some glial cells clean up the debris from neurons that have been injured (the microglia), others provide nutrients to nearby neurons (they're called astrocytes, the type that were implanted into the mice), other glial cells play other roles that help neurons survive and function.

So it was surprising that mice implanted with human glial cells became smarter, or at least faster learners, than other mice. Glia aren't normally thought of as playing a key role in learning.  If you've seen the movie Good Will Hunting, you may remember Matt Damon's character working as a janitor in the halls of Harvard University.  His mathematical mind far surpassed that of the Harvard faculty.   The genius in the Harvard math building was the guy sweeping the floors.  The fact that human glial cells, our brain's support staff, would power-up the learning abilities of the mouse brain was surprising, very much in that way.

Here are some of the details:

The experimenters aimed to implant the mouse brain with a particular kind of human glial cell called astrocytes ('astro' because it's shaped like a star).
An astrocyte is a star-shaped glial cell. The image above shows an
astrocyte with its many thin processes radiating out from a round center.

Instead of implanting fully formed human astrocytes, they implanted an immature form of the astrocyte that exists in the brain during late stages of embryonic development. They're called glial progenitor cells or GPCs. The GPCs turn into astrocytes at some point during brain development. Experimenters placed human GPCs into the brains of newborn mice (Goldman, Nedergaard, & Windrem, 2015), taking steps to ensure that the mouse immune system wouldn’t reject the human cells.  This allowed the human astrocytes to develop and take their places within the mouse brain as it developed.

The brains of newborn mice were
implanted with human GPCs, immature
cells that eventually turn into astrocytes.
The immature human cells traveled throughout the mouse brain and increased in number. They soon developed into human astrocytes, which competed with and replaced most of the mouse astrocytes. Within about 10 months, most of the astrocytes in the mouse brain were human in origin! In the  figure below, the human astrocytes are stained red. As you can see, they spread to virtually all parts of the mouse brain. Researchers sometimes refer to this procedure as “glial humanization,” or more specifically, “astrocytic humanization” of the mouse brain.

The front of the mouse brain is to the left, the
back is to the right. The red dots are human  astrocytes.
They multiplied throughout the mouse brain.
The neuroscientists presented the mice with tones a few seconds before a mild shock was delivered to the floor of the cage. Once mice learned the tone-shock association, they would ‘freeze’, immobilized in their place, as soon as they heard the tone.  The more experience they had with tones followed by shock, the more time they spent freezing when the tone came on.  The black line below shows the behavior of a group of mice that was not transplanted (not grafted) with foreign cells. Over the course of 4 days of training, they increased the time they spent freezing when the tone came on. The purple line shows the behavior of another group of mice that received a graft containing astrocytes from other mice (allografted control mice). These mice gradually learned the tone-shock association as well. However, the mice with human astrocytes (the red line) were by far the fastest learners. After one day of training, they showed better learning than the other groups of mice showed after four days’ training.


The long, thin processes that radiate out from the center of the astroctyes often wrap around synapses between neurons. Evidence suggests that astrocytes enhance synaptic plasticity, the ability of neuronal connections to strengthen, and thereby enhance learning (Han et al., 2013). What is the precise role that these astrocytes, the supposed 'support staff' for the neurons, play in learning?  Do some people with learning disorders have a lack of, or dysfunction in, astrocytes? If so, can we do something to enhance their astrocyte activity? And what is going through the minds of these unusually smart mice with human astrocytes as an experimenter walks into the room?  I imagine an experimenter noticing a look of intelligence in the eyes of the chimeric mouse. Then I imagine the experimenter quickly looking away.

Most of the time I spend thinking about the brain, I think about the basal ganglia

Most of the time I spend thinking about the brain, I think about the basal ganglia.

The basal ganglia is mostly involved in choosing which action to do at the moment  -- e.g., right now should I write a blog post or make coffee or take a walk or call my friend on the phone?  Research on the basal ganglia suggests that it essentially 'assigns' a value to each action.

How does the basal ganglia know how much value to assign to a particular action?  It remembers how rewarding the action was last time I carried it out, and the time before that.  It basically keeps a running average of the reward I received for each particular action in the past.

But wait.  Doesn't the value of a particular action (how much I want to do it) depend upon the situation I find myself in, my mood, where I am, who I'm with, and so on?   That's exactly where the basal ganglia comes in.  It doesn't simply assign a value to an action per se.  Instead, the value it assigns varies according to the situation I'm in.  The basal ganglia remembers the situation I was in (at least some aspects of it) when I carried out the action in the past. And it keeps track of how rewarding the action was in the context of situation 1, situation 2, and so on. In order to tell me if a particular action is likely to be rewarding right now, it takes into account my current situation, looks for actions I've carried out under similar situations in the past, and tells me which ones led to the greatest reward (positive feelings).  The answer the basal ganglia gives me doesn't come in words. It comes in terms of how much motivation I feel when I consider choosing that particular action.

How does the basal ganglia keep track of values for each of the different situations I find myself in?  How does it know what the current situation is?  Areas of the the basal ganglia are mostly comprised of input neurons and output neurons.  The 'situation' is defined by the inputs. An area of my basal ganglia, right now, is receiving input from visual areas of the cortex telling it that I'm at a table in my in-laws' garden in Spain.  It also receives input about the temperature of the sun on my skin (a gentle warmth), my mood (relaxed) and signals from my body (my legs are tired from walking a lot this afternoon). The cerebral cortex is in charge of sensing these components of my current 'situation', and it sends input to inform the basal ganglia about it.

The outputs of the basal ganglia correspond to the particular actions I choose (e.g., writing this blog post).

The value of an action under a particular situation is represented as the strength of the connections between the inputs and outputs.

Dopamine -- which is released in large quantities when I perceive my experience to be better-than-expected - strengthens the input-output connections of the basal ganglia -- i.e., increases the value assigned to a given situation/action pair.

The details get technical.  Probably more technical than you're in the mood to hear about, or I'm in the mood to write about.  Besides, it would require me to sketch a diagram.

Like many things we do, I'm entering into behavior that doesn't resemble past actions I've taken.  My basal ganglia doesn't give me much of an indication of how much reward to expect from writing a blog post with details about how the basal ganglia works.  That means I have to rely on other parts of my brain that imagine outcomes of new behaviors.  The hippocampus, best known for its role in memory,  is also involved in imagining scenarios that never occurred.  As I consult with my hippocampus and related areas, my basal ganglia steps in and tells me to bring this post to a close.   There's fresh coffee on, and in my current situation (neuroscientists like to use the term 'state'), in my current state, serving myself coffee is almost guaranteed to be a high-value action.

Learn a little bit each day


I've discovered a fun hobby.  I take a subject that I know almost nothing about. Maybe a subject that I missed in high school because I was daydreaming in class.   Back then I could decide quickly that a subject like Chemistry wasn't my kind of thing.   That was lucky, because there's a special joy in learning about a subject fresh as an adult.  And the less attentive you were in class, the fresher and more exciting it is to learn about now.

So, with my morning coffee I open up this Chemistry Book, and I'll read just a page or two; sometimes half a page.  But if you do that - or anything else - every morning, a page a day adds up to getting through a textbook in a year.  Even in three months, you have a feel for the topic.


When I decided to learn something about Chemistry, I Googled "BEST INTRODUCTORY CHEMISTRY TEXTBOOK" and soon I had a list of the 5 or so textbooks that were often described as "very good'.  And I chose one that was, sure enough, very good.