Practice makes perfect, but how? Two groups of neuroscientists using MRI brain imaging announced last month that they were able to see changes inside the brains of people after mastering a new skill.  The big surprise is that the part of the brain that changed has no neurons or synapses in it!  The cerebral remodeling during learning was seen in the mysterious and still largely unexplored “white matter” region of the brain.

“Grey matter” is synonymous with smarts, but in fact only half of the human brain is grey matter.  White matter, the “other brain tissue”, is rarely mentioned.   Neurons in the cerebral cortex are packed into in the top layers of the brain, where they are connected together through synapses.  Learning takes place in the grey matter by linking neurons together into new circuits by strengthening synapses or forming new ones.

But beneath the topsoil of the brain lies a dense network of fibers packed into a spaghetti-like snarl that is so complicated it is difficult to study or comprehend.   These fibers are the wire-like axons projecting out from neurons in grey matter that transmit electrical impulses.  Like buried telephone lines, these tightly bundled cables transmit information over long distances to communicate between distant regions of the cerebral cortex that are specialized to carry out different aspects of a complex cognitive function.  

To understand the importance of white matter, consider what is happening under the baseball cap of a left fielder leaping over the wall to snatch a baseball in mid air.  Visual processing in the back of his brain perceives and tracks the flying object and at the same time it monitors all the other objects on the field as the athlete runs to catch the ball.  Then the motor control centers in the parietal region of his brain engage to launch his body on a running trajectory to intercept the projectile.  Finally, precisely timed fine motor control extends his arm into space with millimeter precision to clench fingers at the right instant to pluck the speeding ball out of the sky.  All the while the player simultaneously perceives the fluid situation on the field as runners advance and strategies unfold so that he can make critical split-second decisions—“Do I hold the ball or hurl it to home plate?”  This higher level decision making is calculated in the frontal lobes, just behind the eye brows.  All this vital communication sweeps across the entire brain from the back of the skull to the front to activate different regions of cerebral cortex specialized in executing individual aspects of the skill. 

That’s the job of white matter—long distant speedy communication.  The tissue is white because many axons are coated with tightly wrapped layers of electrical insulation called myelin.  This insulation, made by non-neuronal cells (called oligodendrocytes), speeds the transmission of electrical impulses 100 times faster than transmission rates through bare axons.  The complex skill of catching a baseball is a far cry from Pavlov and his slobbering dog learning to associate the sound of a bell with food.  Skill learning is likely to involve different mechanisms.  The kind of complex learning involved in mastering new skills such as catching a fly ball, takes time to learn and repetition over the course of days,weeks or years.  This type of learning is what these neuroscientists dared to tackle.

In the first study,  Jan Scholz and colleagues at the University of Oxford, England, used MRI brain imaging to obtain a detailed scan of the brain of 48 right-handed adults.  Then they taught half of them to juggle.  Anyone who has tried to master the three-ball-toss knows how difficult juggling is and how much practice it takes to learn it.  But as in learning to ride a bike, once the complicated skill is mastered, suddenly everything “clicks” and the process becomes mysteriously automatic.  Learning to read is like that too, which is what the second research group investigated, but first let’s have a look at the fascinating study peering into the brain of jugglers. 

Six weeks after training, the jugglers had their brains re-scanned, as did the other half of the group who were not taught to juggle.  The untrained individuals comprise the vital experimental control group, which allows researchers to check whether any brain changes they find in the jugglers might have happened by chance.  What the researchers found is that the structure of white matter in the region beneath the cortical area known to handle visuo-motor processing became more highly organized after learning to juggle (the right posterior intraparietal sulcus, IPS).  A third MRI taken a month later without any further training showed that the changes in the white matter in this area of the juggler’s brain were still evident.

The study’s lead author, a juggler herself, was not too surprised to see changes in this part of the brain, “It is an area of visuo-motor integration, which is an essential aspect of learning to juggle,” she told me.  “The IPS is quite important for the co-ordination of quick and precise arm and grasping movements with the visual tracking of the juggling balls.”  But the big surprise was to find that the white matter regions in this part of the brain had changed at all.  Not unexpectedly, changes were seen in the grey matter above these white matter fibers, but the grey matter changes seemed to be independent of the white matter changes.  “The white matter changes seemed to be primarily training or activity related.  In contrast, once triggered, the grey matter changes seem to continue even after 4 weeks of training abstinence, suggesting a more sluggish underlying mechanism.”  Changes in dendrites or vascular supply could have caused the grey matter changes, but what about the white matter? 

The brain imaging cannot tell us exactly what has changed at a cellular level in white matter after learning to juggle.  The technology, called diffusion tensor imaging, is sensitive to how uniformly water diffuses between the fibers in white matter.   The larger the fiber diameters and the more densely packed axons are, or the more thickly wrapped with myelin insulation, the better water flows along the fibers than in all directions.  Just as paint will flow up the bristles of a paint brush, but stains diffuse symmetrically through the fibers of a carpet because they are less organized, the microstructure of these white matter tracts carrying signals coordinating vision and hand motion became more organized after learning to juggle. 

Since none of the jugglers were willing to donate their brains to science, we can only wonder what has changed on a cellular level in their white matter tracts as they learned the new skill.  Research on experimental animals shows that experience can increase myelin formation, and recently research has shown that impulse activity in axons is communicated to the myelin forming oligodendrocytes, stimulating them to form more myelin.  This is what the researchers would be most excited to learn, because changes in myelin during learning would affect the speed of information transmission through neural circuits, and optimizing the speed and synchrony of nervous signals transmitted between the distant cortical regions could in theory explain part of the process that enables us to learn new complex skills. 

Another difficult but very important skill for everyone to master is learning to read.  Brain imaging has detected differences in certain white matter tracts in the brains of people with dyslexia, and differences in white matter have been observed in children with different reading abilities, but this does not necessarily mean that learning to read changes white matter.  These differences could be individual differences or related to a large number of other changes taking place in the brain of children as they mature.  To test this hypothesis, one would need a population of adults who were never taught to read, and then after giving them reading lessons, scan their brain for any changes.  But where would you find such a population of illiterate adults?

Dr. Manuel Carreiras and colleagues at the Bosque Center of Cognition Brain and Language in Spain, told me he stumbled upon the perfect set of experimental subjects by chance.  “I was looking for illiterates and they were very difficult to find in Spain.  One of my doctoral students was from Colombia, so I asked her.”  She related the troubled history of guerrilla warriors in her country who were now being re-integrating into mainstream society and learning to read for the first time as adults.  “So I asked her whether it would be possible to get them in Bogota, where we could find an MRI machine.” 

The brain scan gave the answer as clear as a picture.  The splenium of the corpus callosum was bigger in the guerrilla warriors who had completed the reading lessons.  The corpus callosum is the large bundle of fibers that connects the left and right sides of our brain together.  This was just the spot that previous research had found was sometimes underdeveloped in people with dyslexia. 

A remarkable contribution of this study is that this is a snapshot of brain structure, not a DTI measure of water diffusion.  Carrieras and colleagues also did DTI and functional brain imaging, which backed up their findings with structural MRI.  Since the fibers in the splenium of the corpus callosum are laid down during embryonic development, the increased bulk of white matter in this pathway must have developed during the process of learning to read as an adult.  Again, we can’t say exactly what has changed on a cellular level in rewiring the brain during reading, but Dr. Carreiras also suspects increased myelin could be involved. 

This raises some interesting new leads for helping children with dyslexia.  Most children with dyslexia struggle with reading, but eventually most do learn to read, and some become quite proficient in reading and writing later in life.  Their biggest problem, many argue, is a rigid school system that cannot adapt to the fact that there are great individual differences in the way everyone’s brain is wired, and this affects the way and the rate of learning different kinds of skills, such as reading and mathematics.  Because this study shows that this white matter region vital for reading changed in the process of learning to read, this casts a new light on the large body of literature that had documented differences in dyslexic brains.  “This new study therefore suggests that some of the differences seen in dyslexia may be a consequence of reading difficulties rather than a cause,” Carreiras told me.  This new insight also offers a hopeful outlook for people with dyslexia because it suggests the possibility that dyslexics could (or perhaps they do) modify these pathways through experience as they eventually learn to cope with the reading difficulty.  “This is precisely one of the questions we are addressing,” he shared with me in explaining his plans for a large project on dyslexia in Spain. 

Discovering changes in white matter turns traditional concepts of cellular learning on its head, because these are modifications of the output of neurons rather than changing the synaptic input.  Historically myelin was of no interest to neuroscientists working to understand how the brain learns.   Myelin was thought to be static, a structural element that was laid down on axons during development, but the insulation never changed unless it was damaged or diseased, as in multiple sclerosis.   These old assumptions are now being re-examined.

It's now clear that we will never fully understand the mechanism of learning if all attention is focused only on what happens at tiny synapses and we fail to consider the efficiency of information flow through the global system of networks in the brain.  By analogy, neuroscientists have broadened their scope of investigation from the transistor to the internet.  Following this cerebral information highway is leading us on a fascinating road into the future.