Nabiha Saklayen Could you recover from illness ... using your own stem cells TED
You’re sitting in the doctor’s office
waiting for test results.
She comes in and says,
“You have Parkinson’s disease.”
Your heart sinks, and you think about
everything that will go wrong:
you’ll be unable to walk, unable to feed
yourself, your hands trembling, drooling,
unable to swallow.
But before you say anything, she says,
“Not to worry, we’ll put
in an order for your cells today.”
You come back a week later,
and a surgeon transplants
brand new neurons into your brain.
You just received an on-demand
functional cure for Parkinson’s,
made from your cells.
It sounds like science fiction,
but in the future,
we will all have the option of having
our stem cells banked ahead of time
so that any time you need new neurons,
new muscle cells, new skin cells,
they’d be generated from this bank.
And because they’re
100 percent your cells,
your immune system is extremely unlikely
to reject or attack those cells.
In fact, the body has no idea
that these cells were
actually made in a cell factory.
All of this is possible
because of a breakthrough
at the intersection of biology,
laser physics and machine learning.
We’ll start with biology.
The human body is an absolute miracle.
Trillions of cells are working
in synchronicity
to pump blood, secrete dopamine
and let me see and speak to you right now.
But as we age, our cells age, too.
That’s why our skin starts to sag,
our cartilage wears away,
and your five-mile run
might turn into a 20-minute walk.
Yes, we’re all getting older.
Our bodies are ticking time bombs.
But stem cells could offer a solution,
because one stem cell can become
almost any cell in your body.
My grandma passed away
due to diabetes in 2012.
If the technology
were available at the time,
we could have used her stem cells
to generate new pancreatic cells,
and it could have cured her.
Now, unfortunately, stem cells
are notoriously difficult to engineer.
One fundamental problem
relates to how they’re made,
which involves taking
a patient’s blood cells
and adding chemicals to those blood cells
to turn them into stem cells.
Now, during this chemical process,
you never end up with
a perfect set of stem cells.
In fact, you get a very messy plate
of cells going in different directions –
towards the eye, brain, liver –
and every random cell must be removed.
Until recently, the main way
to remove cells was by hand.
I remember the first time I visited
the Harvard Stem Cell Institute.
I watched a highly skilled scientist
sitting at a bench looking at stem cells,
evaluating them one at a time
and removing the unwanted cells by hand.
It’s a slow, tedious
and artisanal process,
which is why generating
a personalized stem cell bank today
costs about one million dollars.
Now, using a donor’s stem cells
is much cheaper,
but your immune system will likely
attack or reject those cells
unless you take immunosuppressants,
which, unfortunately, is not an option
for a lot of people,
especially the elderly.
To avoid this problem,
some scientists are banking stem cells
from individuals with the most
common genetic backgrounds.
Here in the US,
let’s say we made a cell bank
with 100 of the most common cell lines.
It could work for about
75 percent of Caucasians,
50 percent of African Americans.
But it gets harder.
My cofounder is Filipina-Mexican,
and it’s unclear if she would
be ever covered by a bank.
And regardless,
if you could choose between using
a stranger’s cells versus your own,
wouldn’t you choose your own?
Personalized stem cells
are our opportunity
to make medicines that truly work
for me, for you and everyone.
And in order to make this process of stem
cell production affordable and scalable,
we have to automate it.
Different people are taking
different approaches to doing that,
and I decided to use physics.
Since childhood, I’ve been
a die-hard physics fan,
gazing at the stars,
daydreaming about space travel.
Thanks, Mom, for not thinking I was weird!
My family moved around a lot,
from Saudi Arabia to Germany
to Sri Lanka to Bangladesh,
and each time, I had to learn
new languages and cultures.
Eventually, I fell in love with physics
because it was a universal language
that I didn’t have to relearn every time.
When I started my PhD,
I joined a laser physics lab,
because lasers are the coolest.
But I also decided to dabble in biology.
I started using lasers
to engineer human cells,
and when I talked to biologists
about it, they were amazed.
Here’s why: scientists are always looking
for ways to make biology more precise.
Sometimes cell culture
can feel a lot like cooking:
take some chemicals, put it in a pot,
stir it, heat it, see what happens,
try it all over again.
In contrast, lasers are so precise,
you can target one cell in millions
at precise intervals –
every second, every minute,
every hour – you name it.
I realized that instead of doing
this tedious process of stem cell culture
by hand,
we could use lasers to remove
the unwanted cells.
And to automate the entire process,
we decided to use machine learning
to identify those unwanted cells
and zap them.
Algorithms today are great at finding
useful information and images,
making this a perfect use case
for machine learning.
Here’s how it works:
Take some blood cells,
put it in a cassette.
Add chemicals to those blood cells to turn
them into stem cells like always.
Now, instead of having a human
look for those unwanted cells
and remove them by hand,
the machine identifies the unwanted cells
and zaps them with a laser.
As you can see, this entire process
happens by machine.
The computer decides when
and how often to print the cells
and uses a fully automated system
to run the process.
After repeated pruning,
you end up with a perfect culture
of your stem cells,
ready to be banked and used at any time.
In the future, we’re going to have
stem cell farms
with stacks and stacks of hundreds
and then eventually millions of cassettes,
each cassette a personalized
bank for one human.
Nurses will take a sample
of your cord blood right at birth
and ship it off for cultivation,
so that for the rest of your life,
your stem cells are on file, banked,
ready to go, should any
medical need arise.
Let’s say you develop heart disease.
Your doctor can order up new heart cells.
Hair loss. They can order up new hair.
The most immediate application
of this technology is for implants.
Dr. Kapil Bharti’s research
at the National Eye Institute
has informed a breakthrough clinical trial
for a stem cell derived
therapy for blindness.
As the process becomes cheaper,
scientists can run larger and larger
clinical trials at scale
to develop new treatments
that don’t exist today,
because what costs
one million dollars today
will soon be less than 50,000,
and then even cheaper with time.
Now, it gets even more
interesting than that.
And perhaps you have longevity in mind.
That is certainly a possibility.
In the future, we might use
these exact same stem cell banks
to generate entire new organs,
new tissues, new skin …
New bone, teeth, anyone?
This technology also has the potential
to revolutionize
personalized pharmaceuticals.
Today, taking medicine is,
to some degree, trial and error.
You don’t really know
if the drug is going to work for you
until you put it in your body.
But what if we had a miniature
human replica of you with your cells –
eye cells, brain cells, heart cells,
muscle cells, blood cells –
on a chip?
A miniature human replica of you.
We could take the drugs, test
them on the cells in the lab first
to see how it works.
If it works, fantastic.
Go ahead and take the drug.
If it doesn’t, pharmacists can
order up custom drugs just for you.
This has been the hope and dream
of scientists for decades.
With this technology,
we can finally realize
the true potential of stem cells:
on-demand functional cures
made from your cells.
Cures that your body won’t reject.
Cures that truly work for everyone.
The future of regenerative medicine
is 100 percent personalized,
and it’s a lot closer than you think.
Thank you.
(Applause)