I think the best way to explain, though, how we do this
is through the eyes of one of my patients.
When I opened my lab at Stanford,
Eduard, who's on the autism spectrum,
sent me this drawing
depicting how he thought we were studying brain disorders.
Now to paraphrase him, he said,
"What I think you're doing is you're climbing up a ladder,
poking holes in people's brains
and then use tiny telescopes to watch neural cells."
Of course, that's not what we do.
So I called him up, explained the process,
and then the next morning he sent me another drawing,
which I think ended up being a quite accurate representation
of the work that we and many others now are doing.
Again, to paraphrase him, he said,
"You're taking skin cells
from patients that have specific brain disorders,
then doing some mumbo jumbo to the cells
to push them back in time
and turn them into stem cells."
And then he knew that stem cells can be coaxed to become any cell type.
“So then you’re taking them and turning them into brain cells
that form brain circuits.”
That's right. We can build human brain circuits in a dish.
How is that possible?
Building on the hard work of biologists over the past 15 years or so,
we can today take any cell type from any individual
and then push it back in time to turn them into stem cells
and then guide those stem cells to become any other cell type.
We start by asking a patient to provide a small skin sample.
We then take those skin cells,
reprogram them by putting a series of genetic factors
and push them back in time
so that those skin cells become stem cells.
It's like cellular alchemy.
These stem cells have almost magical abilities
to turn into any other cell type.
So what do we do?
We take the stem cells, we dissociate them,
we then aggregate them
so that they form spheres or tiny balls of cells.
We then take those, move them into a special plate
where there is a kind of chemical soup.
And that chemical soup will allow them to grow
and transform and turn into a brain organoid.
By providing different cues, we can turn this brain organoid
to resemble specific regions of the central nervous system.
For instance, we have a recipe
that allows them to become a cerebral cortex,
the outer layer of the brain.
By using a slightly different combination of factors,
we can turn them into a spinal cord.
The secret to this process is careful guidance.
In the end, they look like this.
Tiny clusters of brain cells at the bottom of a dish.
And let me be clear.
They are not brains in a jar.
These are parts of the nervous system in a laboratory dish.
Each of them contains millions of cells,
and we can even listen as they fire electrical signals.
(Electrical signals firing)
Or we can watch them
as they sparkle with electrical activity.
Or we can image inside and watch the cells as they communicate with each other.
Isn't it remarkable to think that just a few months ago
these cells were skin cells in a patient,
and now they are neural cells at the bottom of a dish
that we can study at ease.
(Applause)
Thank you.
So with these models of brain growth,
we started wondering: Could we use them to start to understand disease?
So for instance, we wanted to know,
could we understand how low oxygen impacts the brains of premature babies?
So to do this, we took brain organoids and put them in a special incubator.
We then lowered the concentration of oxygen and watched them.
We discovered something quite interesting.
Only one specific cell type was affected by the low oxygen.
That cell type is responsible for the expansion of the human cortex.
We found exactly how that happens
and even found the drug that could prevent that process.
These clumps of three-dimensional tissue
can be grown in a dish for years.
In fact, we've maintained the longest cultures
that have been reported to date, going beyond 800 days.
At nine to 10 months, which is the equivalent of birth,
they slowly transitioned, and they started to resemble the postnatal brain.
We have discovered a brain clock
which keeps track of time in a dish and outside of the uterus.
Understanding the molecular mechanisms that underlie this brain clock
could be key to finding new strategies
to either accelerate or decelerate or rejuvenate human brain cells.
The work that I've shown you so far
is pioneering not just because of what it teaches us
about the human brain,
but also because of the frontiers of ethics.
Organoids and assembloids are not full replicas of the human brain.
They're not brains in a jar. They're not mini-brain.
They're not some stepping stone to a Frankenstein monster.
They have no blood flow,
they receive no meaningful inputs and outputs.
But at one point, they may become more complex.
At one point, they may receive sensory input.
So as the science advances,
we in the scientific community have been very careful
about discussing what are some of the ethical questions,
the societal implications and potential regulations.
Most of the work that I’ve shown you so far
has been in one specific brain region.
But to really understand circuits,
we actually need to build more complicated brain circuits.
And so to do this, six years ago,
we came up with a new approach to build human circuits
called an assembloid.
Assembloids are essentially blocks of tissue
that we build in a dish from multiple organoids put together.
When we put two brain organoids together,
we discovered something really fascinating.
First, they fused to each other.
But then they started to communicate,
and brain cells from one side
started to slowly migrate onto the other side
and form circuits,
much like they would in the actual brain.
In fact, we can even watch them live as they move from one side to the other.
I still remember how we were in the lab in absolute awe
when we saw for the first time
how human cells undergo this peculiar jumping behavior.
This is all fascinating, but what is it actually good for?
Dysfunction in the human brain causes brain disorders,
such as autism and schizophrenia and Alzheimer's disease,
devastating conditions that are poorly understood.
Nearly one in five individuals suffers from a psychiatric disease.
What is even more striking
is that the lowest success rate for finding new drugs
is in psychiatry, out of all the branches of medicine,
likely because until now we couldn't really access the human brain.
Using brain organoids and assembloids,
we can create avatars for a patient's brain development
and then use those to dissect the molecular mechanism of disease.
Let me give you one example.
As you have seen,
assembloids can be used
to model this healthy jumping behavior of neurons.
So what we did is we created assembloids from patients with Timothy syndrome,
which is a rare genetic disease associated with autism and epilepsy.
When we looked inside the assembloids, we noticed something remarkable.
The cells were moving much faster,
but every time they would jump, they would jump a shorter distance.
So in the end, they would be left behind.
Over the past six years in extensive studies,
we've actually dissected the molecular mechanism of this defect
and even found ways of restoring it.
And we're excited to be moving towards a potential therapeutic avenue
in the next year or so.
(Applause)
The promise of organoids and assembloids
is that they will slowly allow us to gain new insights
into the hidden biology of the human brain.
And by doing so,
they could revolutionize the way we think about human brain development, evolution,
function and disease.
So what's next?
Well, to really be able to gain insight into more complex brain disorders,
we need to build more complex circuits.
So in the last minute, let me show you
the most complicated circuit we have built to date.
The circuit that controls voluntary movement.
To do this, we've created three organoids.
One, shown here in purple, that resembles the cortex.
One, in yellow, that resembles the spinal cord,
and one, in red, that resembles human muscle.
We then put them together and watched them fuse
and noticed something really spectacular.
Neurons on the cortical side started extending axons,
find spinal motor neurons in the spinal side,
connect with them,
and then those farther project and connect to muscle.
When we put a light stimulus on the cortical site,
we noticed the muscle on the opposite side contract.
We have modeled for the first time a human cortical motor pathway.
(Applause)
And let me be clear.
These cells find each other.
Unlike in engineering, we don't have a master plan,
we don't provide a plan because the human brain builds itself.
And then in itself, it's a remarkable opportunity
to try to reverse engineer
what are some of the steps that underlie human brain development?
I know that this all sounds science fiction,
but we now do this routinely in the lab.
We have derived thousands and thousands of organoids and assembloids
from patients with various neuropsychiatric diseases,
including, for instance,
infecting them with viruses such as polio virus
to understand how diseases arise.
The statistician George Box famously said,
"All models are wrong, but some are useful."
I do the work that I do
because the promise and hope of brain assembloids and organoids
is that by allowing us to recreate circuits of the human brain,
we will gain new insights into human biology.
And this in itself will open a new era in the treatment of brain disorders.
Thank you.
