Latest / Talking Science / CRISPR – Editing the Code of Life

Transcript

CRISPR – Editing the Code of Life · Talking Science

0:00
Usually when we hear about a massive paradigm
0:03
shifting leap in human technology, right, especially
0:07
something that feels like it was ripped straight
0:09
out of a science fiction novel. Oh, absolutely.
0:11
There's, you know, this expectation of a certain
0:13
kind of dramatic origin story. Like you picture
0:16
a pristine billion dollar futuristic lab. Yeah.
0:20
Lots of glowing holographic screens. Exactly.
0:23
Or perhaps a dramatic thunderstorm raging outside
0:26
the window while some brilliant eccentric. scientist
0:29
in a white coat has a sudden earth -shattering
0:32
epiphany. All right, right. Taking off her glasses
0:34
and whispering something dramatic. Yes, whispering,
0:36
we've cracked the code of life. But the reality
0:39
of humanity's most powerful biological tool,
0:42
I mean, the tool that literally allows us to
0:44
rewrite the fundamental code of living things,
0:46
it doesn't start in a futuristic lab at all.
0:49
Not even close. It starts with a problem that
0:51
is incredibly wonderfully mundane. It starts
0:53
with yogurt. It really is the ultimate deflation
0:57
of the mad scientist trope, isn't it? It really
0:59
is. Because for decades, biologists and futurists
1:02
imagined how we might one day re -engineer life.
1:06
And the assumption was always that it would require
1:08
us to invent some entirely new and possibly complex
1:13
synthetic machinery from scratch. Like we'd have
1:16
to build it atom by atom. Exactly. We thought
1:19
we'd have to build the tools ourselves. And all
1:21
the while, the actual solution was just sitting
1:25
there. literally floating in our breakfast. That
1:27
is wild. That had been perfected by single -celled
1:30
organisms over billions of years of evolution.
1:33
And that contrast, honestly, is the heart of
1:35
what we're doing today. Our mission on this deep
1:38
dive is to completely demystify CRISPR. Yeah.
1:43
For you listening, we're going to strip away
1:44
the science fiction Hollywood tropes, cut through
1:47
the intimidating academic jargon, and just break
1:49
this down so that absolutely anyone can understand
1:51
it. Because you really don't need a PhD to get
1:53
why this is so huge. Exactly. We're going to
1:56
take our time, keep things conversational, and
1:58
walk through exactly how this microscopic system
2:00
functions, why it has completely upended modern
2:04
medicine, and where it's ultimately taking the
2:06
human species. It's a huge journey. Because the
2:09
story of CRISPR isn't just a niche topic for
2:11
geneticists, you know. It is fundamentally a
2:14
story about our future. To really grasp the magnitude
2:18
of those implications, though, we have to start
2:20
with the biology itself. But, you know, at a
2:22
very accessible level. We're going to unpack
2:25
how a simple bacterial defense mechanism. A tool
2:29
that bacteria use just to survive their mundane
2:32
day -to -day lives in the dirt and the water
2:35
was noticed by observant researchers. And then
2:37
reverse engineered, right? Yeah, reverse engineered
2:39
piece by piece and then repurposed into a universal
2:43
programmable tool. A tool that lets us edit the
2:46
master instruction manual of literally any living
2:49
organism on Earth. I just, I love that phrase,
2:51
master instruction manual. It's the best way
2:53
to think about it. But to understand how we can
2:55
rewrite that manual today, we actually need to
2:57
travel back in time to the late 1980s. Yeah,
3:00
long before anyone was debating the ethics of
3:02
editing human genes. Right. There was this strange
3:05
little mystery hidden deep inside the genetic
3:08
code of bacteria. Back in 1987, a researcher
3:12
in Japan named Yoshizumi Shino was studying E.
3:16
coli bacteria. And we have to remember, sequencing
3:18
DNA in the 1980s was not like it is today. Oh,
3:22
definitely not. You didn't just put a sample
3:24
into a machine and get a digital readout on a
3:26
screen. Right. It was agonizingly slow, manual.
3:30
chemical work. Yeah, it was incredibly tedious.
3:33
Researchers were painstakingly reading through
3:35
the bacterial genome letter by chemical letter.
3:38
And just to define our terms clearly early on
3:40
here, if you think of an organism as a complex
3:42
machine, its genome is the complete set of instructions
3:45
for how to build and operate that machine. It's
3:48
a manual. Exactly. Every cell carries a copy
3:50
of this manual written in the four chemical letters
3:52
of DNA, A, C, T, and G. So Ishino was reading
3:55
a tiny fraction of this e -caller manual, right?
3:58
trying to understand a specific gene. And then
4:00
he noticed an anomaly just past the end of the
4:02
gene he was actually interested in. Yeah, imagine
4:04
reading a highly technical textbook. And suddenly,
4:07
right at the end of a chapter, you encounter
4:09
a short, nonsensical paragraph that just repeats
4:12
over and over again. This weird, jimmerish. But
4:15
it's not just a continuous block of repetition.
4:17
There is a distinct rhythm to it. Right. Ashino
4:20
saw these short segments of DNA that repeated
4:22
identically. But between each repeating segment,
4:25
there sat a unique spacer sequence of DNA. So
4:29
the pattern was like repeat, unique spacer, repeat,
4:32
unique spacer. Repeat, unique spacer. And those
4:35
unique spacer sequences didn't seem to match
4:37
any known biological function in the bacterium
4:39
at all. And at the time, it was treated as a
4:42
complete curiosity. Like a typo. Yeah. Ishino
4:45
essentially published a paper saying, we found
4:47
this weird repeating pattern. We have no idea
4:49
what it does. Moving on. Moving on. And for a
4:51
long time, the scientific community. largely
4:53
ignored it it fell into the broad rather dismissive
4:56
category of you know junk DNA but then throughout
5:00
the 90s a Spanish researcher named Francisco
5:03
Mojica started noticing the exact same repeating
5:06
pattern in completely different types of microbes
5:08
right organisms living in extreme environments
5:11
like the highly salty marshes of Santa Pola in
5:15
Spain So Mojica realized this wasn't just a weird
5:18
quirk of E. coli. This pattern was ancient and
5:21
it was everywhere. And Mojica is actually the
5:23
one who gave this mystery its famous name. Which
5:26
honestly is one of my favorite little trivia
5:28
facts about this whole field. It's so funny.
5:30
Because scientists had zero clue what this pattern
5:33
actually did. They couldn't name it based on
5:36
its function. So they named it by literally just
5:39
describing its physical shape in the DNA. They
5:41
called it CRISPR. Clustered? Regularly interspaced
5:45
short palindromic repeats. A massive mouthful.
5:48
Is a massive mouthful of a biological acronym.
5:51
But it perfectly describes the architectural
5:53
shape of the genetic puzzle pieces. Right. The
5:56
repeats are clustered together. They are regularly
5:58
interspaced by those unique spacer sequences.
6:01
They're short. And they are palindromic. Yeah,
6:03
meaning the genetic letters read the same forwards
6:06
and backwards, which causes the DNA in those
6:08
spots to fold into little hairpin shapes. But
6:11
even with a catchy new acronym, the actual function
6:14
of CRISPR remained a total black box. The acronym
6:18
described the shape of a lock, but nobody had
6:21
the key. Exactly. The massive aha moment. The
6:25
realization that elevated this from an obscure
6:27
genetic quirk to a biological revelation came
6:31
when computing power finally caught up with biology.
6:34
In the mid -2000s, right. Yeah. As global DNA
6:37
databases grew massive and search algorithms
6:40
improved, scientists finally thought to ask a
6:42
very simple question. Which was, let's stop looking
6:45
at the repeating bookends. What do those unique
6:48
spacer sequences in the middle actually say?
6:50
Right. Where do they come from? Yeah. And when
6:52
they ran those spacer sequences. through the
6:54
databases, they didn't match bacterial DNA at
6:57
all. What did they match? They matched viruses.
6:59
Wow. And not just any random viruses. The sequences
7:02
in those spacers matched the DNA of bacteriophages,
7:05
which are specific types of viruses that are
7:07
the natural predators of bacteria. Bacteriophages,
7:10
or phages for short. And these are terrifying
7:12
little biological machines. Oh, they look alien.
7:15
If you've ever seen a picture of one under an
7:17
electron microscope, they literally look like
7:20
the Apollo lunar landers. They really do. They
7:23
land on the surface of a bacterium, inject their
7:25
own lethal viral DNA into the cell, hijack the
7:29
bacterium's machinery to multiply, and then cause
7:32
the bacterium to just explode. Releasing thousands
7:34
of new viruses, it is an endless microscopic...
7:38
Billions of years old war. It really is. So what
7:40
Mojica and other researchers realized was that
7:43
this wasn't a genetic typo or a filing error.
7:46
They had stumbled into an incredibly sophisticated
7:48
adaptive bacterial immune system. And more importantly,
7:52
it was an immune system with a long -term memory.
7:54
Yes. To visualize how elegant this is, imagine
7:58
the timeline of a viral attack. A lunar lander
8:01
phage attaches to the bacterium and injects its
8:03
lethal DNA. And usually the bacterium dies. Usually,
8:06
yes. But occasionally the bacterium manages to
8:09
fight off the infection and survive. When it
8:11
does, it performs an extraordinary microscopic
8:13
surgical procedure. What does it do? The bacterium
8:16
snips off a tiny fragment of the defeated attacking
8:19
virus's DNA, and it physically integrates that
8:22
fragment into its own genome. It files that viral
8:25
DNA fragment directly into one of those spacer
8:28
slots in the CRISPR region. So when a virus attacks
8:31
and the bacterium survives, it snips off a piece
8:34
of the virus and files it away in its own DNA.
8:36
It's essentially taking a molecular mugshot and
8:39
putting it on the bacterium's most wanted wall.
8:41
That is the perfect analogy. It is permanently
8:44
updating its own instruction manual with a record
8:47
of the things that have tried to kill it. And
8:49
that most wanted wall isn't just an archive,
8:51
is it? No, it is a live, active defense grid.
8:54
If that exact same species of virus ever returns
8:57
and attempts to infect the bacterium again, the
9:00
bacterium is ready. It remembers. It accesses
9:02
the CRISPR archive. It transcribes that stored
9:05
viral sequence, the mice shot, into a tiny mobile
9:07
molecule called a guide RNA. Okay, we should
9:10
define RNA quickly for those who might be rusty
9:12
on high school biology. If DNA is the heavy master
9:16
reference book that never leaves the library,
9:18
what is RNA? Think of RNA as a temporary, highly
9:21
portable photocopy of a single page of that master
9:23
book. The cell makes an RNA copy of the mugshot
9:26
to send out into the field. Like handing out
9:28
a flyer to a bounty hunter. Exactly. This guide
9:31
RNA acts like a bloodhound. It attaches itself
9:35
to a specialized, weaponized protein, an enzyme
9:38
called a Cas protein. Cas stands for CRISPR -associated
9:42
protein. And the most famous of these is Cas9,
9:44
right? Yep, Cas9. So you have this Cas9. cutting
9:47
protein, which is basically a biological pair
9:49
of scissors, and it's holding this guide RNA
9:52
bloodhound. And the Cas9 protein prowls around
9:54
the inside of the bacterial cell. It scans any
9:57
foreign DNA it encounters, continuously unzipping
10:00
the DNA slightly to check the sequence. Like
10:03
a security guard checking IDs. Exactly. It is
10:06
looking for a perfect letter -for -letter match
10:08
to the viral mugshot it is carrying. The moment
10:10
it finds a 100 % match, the Cas9 protein clamps
10:13
down and physically severs the viral DNA. Chopping
10:17
it in half. Wow. And a severed DNA strand is
10:19
lethal to the virus. The invader is destroyed
10:22
before it can even begin to replicate. It's a
10:24
completely brilliant targeted defense mechanism.
10:28
Now, early on, we mentioned that this story begins
10:31
with yogurt. Yes, the dairy connection. Because
10:34
while geneticists and universities were geeking
10:36
out over the evolutionary elegance of this immune
10:39
system, the dairy industry was looking at it
10:41
with dollar signs in their eyes. They saw an
10:44
immediate. incredibly lucrative, practical application.
10:47
Right, because making industrial dairy products,
10:50
turning thousands of gallons of milk into cheese
10:53
and yogurt, relies entirely on massive vats of
10:56
specific bacterial cultures, like Streptococcus
10:59
thermophilus. Those bacteria are the microscopic
11:02
factory workers. And just like any dense population,
11:06
they are highly susceptible to pandemics. I never
11:08
thought about bacterial pandemics in cheese factories.
11:11
Oh, it's a huge problem. A single phage virus
11:14
can get into a dairy plant, sweep through an
11:16
industrial vat, and wipe out the entire bacterial
11:19
workforce overnight. The milk doesn't ferment,
11:21
the product is ruined, and it costs the industry
11:23
millions upon millions of dollars. So if you
11:26
are a scientist working for a major food conglomerate
11:29
like Denisco in the mid -2000s, which is exactly
11:33
where a researcher named Philippe Horvath was
11:36
working. And you suddenly read that bacteria
11:38
have a natural built -in system to remember and
11:42
destroy the specific viruses that kill them.
11:44
That is a massive industrial revelation. Horvath
11:48
and his team realized they could essentially
11:50
vaccinate their cheese cultures. They could deliberately
11:54
expose their industrial bacteria to the most
11:57
common factory viruses, let the surviving bacteria
12:00
incorporate the viral mugshots into their CRISPR
12:02
arrays, and then selectively breed those hyper
12:05
-resistant strains. And they patented the process,
12:08
of course. Oh, absolutely. It was a total game
12:10
changer for protecting the global dairy supply
12:11
chain. If you've eaten a piece of commercial
12:14
cheese or a cup of yogurt in the last decade,
12:16
there's a very high probability you have consumed
12:18
bacteria that were protected by the CRISPR immune
12:21
system. That is wild. It grounds the science
12:24
in such a tangible, everyday reality. But protecting
12:27
dairy cultures, as massive as it was for the
12:29
food industry, was still just taking advantage
12:31
of a natural bacterial quark. The true scientific
12:34
revolution, the moment that altered the trajectory
12:37
of human medicine, required a leap of imagination.
12:40
It required scientists to realize that this defensive
12:43
system could be hijacked and turned into a programmable
12:46
tool. The monumental insight, the concept that
12:50
completely blew the doors off the field of genetics,
12:52
was a realization about how that Cas9 cutting
12:57
protein actually behaves. What was the realization?
13:00
When scientists really started analyzing Cas9,
13:02
they realized... it is essentially a blind, obedient
13:05
worker. It doesn't know and fundamentally doesn't
13:08
care where the guide RNA sequence it is holding
13:10
came from. Oh, I see. It has no biological loyalty
13:14
to the bacterium's most wanted wall or to defeating
13:16
viruses. So the protein that does the actual
13:19
cutting is just mindlessly following whatever
13:21
chemical instructions it happens to be holding
13:23
at that exact moment. Its only function in the
13:25
universe is to take the guide RNA it has handed,
13:27
scan any DNA it bumps into, and if it finds a
13:30
perfect match, make it cut. That is its entire
13:32
biological programming. Which opens up a huge
13:35
possibility. The brilliant Nobel Prize winning
13:37
leap in logic was realizing, what if we stop
13:42
waiting for the bacteria to supply the mugshot?
13:45
What if we synthesize our own fake mugshot in
13:47
a lab, hand it to Cas9, and send it into a completely
13:51
different kind of cell? And that brings us to
13:54
the breakthrough of 2012, driven by Emmanuel
13:57
Charpentier and Jennifer Doudna. They were the
13:59
ones who proved that we aren't limited to the
14:01
viral sequences bacteria collected. Exactly.
14:04
We could theoretically give this Cas9 worker
14:06
any genetic sequence we want to target. Charpentier
14:09
and Doudna met at a scientific conference in
14:12
Puerto Rico in 2011 and started a transatlantic
14:14
collaboration. And through intense laboratory
14:17
work, they unraveled the exact biochemical mechanics
14:20
of how the guide RNA and Cas9 bind together.
14:22
Because in nature, the guide mechanism actually
14:24
relies on two separate pieces of RNA that have
14:26
to link up, right? Yeah. Doudna and Charpentier's
14:29
team figured out how to engineer those two pieces
14:31
into a single, streamlined guide RNA. They essentially
14:34
created a simplified synthetic skeleton key.
14:37
They demonstrated that you could write a custom
14:38
sequence of genetic letters in a test tube, fuse
14:41
it to this simplified guide, hand it to the Cas9
14:44
scissors, and direct it to cut a specific gene
14:47
in a test tube. They published this in 2012 and
14:50
it immediately set the scientific world on fire.
14:53
But I want to pause here and contextualize this
14:55
because gene editing, broadly speaking, existed
14:58
before 2012. We had genetically modified organisms
15:01
in the 90s and 2000s. We had tools with very
15:04
complex names like zinc finger nucleuses and
15:07
talons. Yes, we did. Why was the Doudna -Sharpentier
15:11
paper such an earth -shattering big deal if we
15:13
already knew how to cut DNA? The older methods
15:16
were incredibly powerful, but they were profoundly,
15:19
agonizingly impractical for widespread use. Before
15:23
CRISPR, the tools we had, like the zinc fingers
15:25
you mentioned, did not use a simple piece of
15:28
RNA to find their target. How did they work?
15:30
The targeting mechanism was built directly into
15:32
the complex three -dimensional folding structure
15:34
of the protein itself. So to find a new target,
15:37
you couldn't just swap out a cheap piece of RNA
15:40
paper. You had to physically redesign and synthesize
15:43
a completely new bespoke protein machine from
15:47
scratch every single time. Exactly. To use a
15:50
mechanical analogy, imagine a massive mansion
15:52
with a million locked doors. The older gene editing
15:56
techniques were like having to hand forge a complex
15:59
custom -shaped metal key in a blacksmith's fire
16:01
for every single individual lock you wanted to
16:04
open. That sounds exhausting. It required deep
16:07
expertise in protein engineering. It took months
16:09
of trial and error by highly specialized postdoctoral
16:12
researchers. And it cost thousands of dollars
16:14
just to attempt to target one single gene. It
16:17
was an elite artisanal craft. So if older methods
16:20
were like hand forging a custom key for every
16:23
single... lock in a mansion, CRISPR is like having
16:26
a universal lockpick where you just type in the
16:27
code. Yes. CRISPR completely obliterated that
16:30
bottleneck. Because with CRISPR, you aren't forging
16:33
a new metal key. You are using the exact same
16:35
Cas9 protein every single time. You're using
16:37
the same universal lockpick. And you simply type
16:39
in the genetic code of the door you want to open.
16:41
You just order a short custom sequence of RNA
16:44
from a biological supply company for a few dozen
16:47
dollars. It arrives in the mail a few days later.
16:50
You mix it with Cas9 and you are ready to edit.
16:53
That is a massive shift. It took a process that
16:55
required a massive specialized laboratory and
16:59
turned it into something that a biology undergrad
17:01
could do in an afternoon. It completely democratized
17:05
the science, which, as a side note, is exactly
17:07
why the subsequent patent battle over who owned
17:10
this technology became so incredibly fierce and
17:13
dramatic. Oh, the patent war was intense. Shortly
17:16
after Doudna and Charpentier's paper, a researcher
17:20
named Feng Zhang at the Broad Institute in MIT
17:22
showed that the system didn't just work in test
17:25
tubes or bacteria. It worked in human cells.
17:28
That sparked a billion -dollar legal war over
17:31
the intellectual property. The stakes were astronomical,
17:34
precisely because the tool was so universally
17:36
accessible. The accessibility was the accelerant.
17:39
Within months of those initial papers, thousands
17:42
of laboratories around the world that previously
17:44
couldn't afford to do gene editing were suddenly
17:47
running CRISPR experiments. They were using it
17:49
in plants, in mice, in human cell lines. The
17:54
sheer volume of biological research exploded
17:57
overnight. But having a universal lockpick is
18:00
one thing. What actually happens once you kick
18:02
the door open? Or in CRISPR's case, what happens
18:05
after you make the cut? That's the key question.
18:07
Because the term gene editing is fundamentally
18:09
a bit misleading. If Cas9 is just a pair of programmable
18:12
molecular scissors, scissors don't write. Scissors
18:15
just sever things. That is the most pervasive
18:17
misconception about CRISPR in the public consciousness.
18:20
And it is vital to clear it up. Let's do it.
18:22
CRISPR itself only performs one single action.
18:25
It creates a targeted double -strand break in
18:28
the DNA helix. It just cuts the rope in half.
18:30
The actual editing, the alteration of the genetic
18:33
code, comes entirely from the cell's panicked
18:36
reaction to being cut. Because a severed DNA
18:38
strand isn't just an inconvenience for a cell.
18:40
It is a catastrophic emergency, isn't it? It
18:43
is an existential threat. If the master instruction
18:46
manual is ripped in half, the cell's machinery
18:48
will stall, and the cell will trigger a self
18:50
-destruct sequence and die. Oh, wow. So the moment
18:53
CRISPR makes that cut, The cell's emergency repair
18:56
proteins rush to the scene, and cells generally
18:59
rely on one of two primary repair pathways to
19:03
fix a broken DNA strand. Okay, what's the first
19:06
one? The first pathway, which is by far the most
19:08
common, is called non -homologous N -joining,
19:11
or NHEJ. But we can just call it the quick and
19:15
sloppy method. Quick and sloppy does not sound
19:17
like the kind of methodology you want applied
19:19
to the master instruction manual of a human being.
19:21
No, but it is an act of sheer biological desperation.
19:25
The cell's repair proteins essentially grab the
19:27
two loose, flailing ends of the broken DNA and
19:30
frantically try to glue them back together as
19:33
fast as possible to prevent the cell from dying.
19:35
Just taping it up. But because it is a messy,
19:38
hasty repair in an emergency environment. The
19:41
cell frequently makes mistakes. It will accidentally
19:43
delete a few genetic letters at the site of the
19:45
cut, or it will haphazardly insert a few random
19:48
letters while trying to seal the gap. So it basically
19:50
scrambles the spelling of the sentence right
19:52
where the cut happened. It introduces a localized
19:55
mutation. And because genes are read in a very
19:58
precise three -letter rhythm, adding or deleting
20:01
even a single letter shifts the entire reading
20:04
frame. The gene suddenly turns into unreadable
20:07
gibberish. So what happens to the gene? The cell
20:11
can no longer read the instructions to build
20:13
the protein the gene codes for, so the gene essentially
20:15
shuts down. It is knocked out. I have to challenge
20:18
the utility of this. If our ultimate goal is
20:21
to cure devastating genetic diseases, why would
20:24
our primary tool rely on a sloppy repair mechanism
20:26
that permanently breaks a gene? isn't the whole
20:29
point to fix things. It seems counterintuitive,
20:32
but sometimes breaking a gene is exactly what
20:34
the doctor ordered. Really? Yes. Many genetic
20:37
diseases are not caused by a missing protein.
20:39
They are caused by a mutated gene that is actively
20:42
producing a toxic, malformed protein that is
20:45
poisoning the body. Give me an example. Think
20:47
of a recipe book that has a typo in it. Instead
20:50
of saying, add one cup of sugar, it says, add
20:52
one cup of arsenic. Okay, bad typo. If you can't
20:55
easily erase the word arsenic, and rewrite sugar,
20:58
your next best option is to take a black marker
21:01
and just aggressively cross out that entire sentence
21:03
so the chef ignores it entirely. Okay, I see.
21:06
By scrambling the gene, you stop the production
21:09
of the toxic protein. Exactly. If your clinical
21:12
goal is simply to disable a problematic gene,
21:14
the quick and sloppy repair pathway is incredibly
21:17
effective. And more importantly, from a clinical
21:20
perspective, it is by far the easiest kind of
21:22
edit to achieve reliably. The cell wants to use
21:25
this pathway. But what about the diseases where
21:27
the gene isn't producing a poison but is simply
21:30
broken and failing to produce something vital?
21:32
Where scribbling out the recipe won't work because
21:35
the chef actually needs that step? That brings
21:38
us to the second cellular repair pathway. Yes,
21:41
the precise method. How does that work? This
21:43
is called homology -directed repair. In this
21:46
scenario, we don't just want to break the gene,
21:48
we want to rewrite it. To accomplish this, scientists
21:51
do two things. First, they send in the CRISPR
21:54
scissors to make the cut at the exact site of
21:56
the mutation. Second, they flood the inside of
21:59
the cell with thousands of copies of a custom
22:02
designed template piece of DNA. Right. This template
22:05
contains the exact correct genetic spelling that
22:09
we want to install. So you are basically handing
22:11
the panic repair crew a blueprint and saying,
22:14
don't just glue it together blindly. Look at
22:16
this picture and rebuild it exactly like this.
22:19
That is the goal. If the environment is right,
22:21
the cell's repair machinery will recognize the
22:24
template, grab it, and use it as a guide to bridge
22:27
the gap in the broken DNA. Meaning it copies
22:29
the good sequence. It effectively copies the
22:32
template, writing your chosen, corrected sequence
22:35
perfectly into the genome. This is true gene
22:38
editing in the way the public imagines it. You're
22:40
literally swapping out the defective code for
22:42
the healthy code. That sounds like the holy grail.
22:44
But whenever a biological process sounds that
22:47
clean and perfect, there's usually a massive
22:49
catch. The catch is efficiency and cellular state.
22:53
Cells heavily favor the quick and sloppy method.
22:56
It's their default setting. Coaxing them into
22:59
using the precise template method takes a lot
23:01
of engineering optimization. And does it work
23:03
reliably? A lot of the time, the cell simply
23:06
ignores the template and uses the sloppy method
23:08
anyway, resulting in a mixture of broken genes
23:11
and fixed genes. Furthermore, this precise pathway
23:14
really only operates efficiently in cells that
23:17
are actively dividing and multiplying. So it
23:20
doesn't work in all cells. It works very poorly
23:22
in cells that are largely static and non -dividing,
23:25
like the neurons in your brain, the cells in
23:27
your retina, or the muscle cells in your heart.
23:29
So while precision editing is the holy grail,
23:32
it is still vastly more difficult to achieve
23:35
in an adult human patient than simply knocking
23:38
a gene out. And this leads right into a technical
23:41
reality that we have to confront. We talked about
23:43
how the guide RNA is essentially a 20 letter
23:46
molecular mugshot. Yes. Cas9 unzips the DNA looking
23:50
for those exact 20 letters. But the human genome
23:52
is roughly three. billion letters long. That's
23:55
like trying to find a specific 20 -letter phrase
23:58
in a whole library of books. What happens if
24:01
that same phrase appears in the wrong book and
24:03
the scissors cut there? How can we possibly guarantee
24:05
that Cas9 won't find a matching sequence on the
24:08
wrong chromosome and cut there? We can't guarantee
24:10
it perfectly. And you have just identified the
24:14
single greatest technical hurdle the field of
24:16
CRISPR has faced over the last decade. off -target
24:19
effects. It's a huge risk. You are mathematically
24:22
correct. In a sequence of three billion letters,
24:25
sequences that are identical or just similar
24:28
enough to fool the Cas9 protein can exist elsewhere
24:30
in the genome. And if the Cas9 scissors scan
24:33
that area, bind to it, and make an unintended
24:36
double -scran break, the consequences can be
24:39
catastrophic. Because if you accidentally cut
24:41
a random gene and the cell uses that sloppy repair
24:44
method... Yeah. You might accidentally knock
24:47
out a tumor suppressor gene. You could theoretically
24:50
cure a genetic disease, but trigger an aggressive
24:52
cancer in the process. That is the nightmare
24:55
scenario for regulators and scientists alike.
24:57
That is why an immense amount of computational
25:00
power is now used to design the guide RNAs. To
25:03
make sure they're unique. We map the entire genome
25:06
and intentionally select 20 letter targets that
25:09
are completely unique, ensuring there are no
25:11
close matches anywhere else. We're also continuously
25:14
engineering the CAS proteins themselves, creating
25:17
high -fidelity variants of the scissors that
25:19
are incredibly strict, refusing to cut unless
25:22
the match is 100 % physically perfect without
25:25
a single letter out of place. But I understand
25:27
the field is actually trying to move away from
25:30
the scissors entirely to solve this. They are
25:32
trying to find ways to edit without making that
25:34
dangerous double -strand cut at all. Yes. The
25:38
next evolution of the tool, pioneered by researchers
25:40
like David Liu at the Broad Institute, involves...
25:43
base editing, and prime editing. What's the difference?
25:46
If CRISPR Cas9 is a pair of scissors that severs
25:48
the rope, base editors are more like a pencil
25:50
with an eraser or a precision chemical whiteout.
25:53
They still use the guide RNA to navigate to the
25:56
specific 20 -letter address in the genome, but
25:58
they use a disabled version of the Cas9 protein
26:01
that cannot cut the DNA. So it just finds the
26:03
spot and stops. Instead, attached to this disabled
26:06
Cas9 is a different enzyme. a demonase, that
26:10
chemically alters a single DNA letter. It can
26:13
literally rearrange the atoms of an A to turn
26:16
it into a G or a C into a T directly on the DNA
26:19
strand without ever breaking the backbone of
26:22
the double helix. That is phenomenal. So you
26:24
completely bypass the cell's panicked, sloppy
26:27
repair response because the cell doesn't even
26:30
realize it has been damaged. You just sneak in,
26:33
chemically swap a single letter, and sneak out.
26:35
It is profoundly elegant. And it dramatically
26:38
lowers the risk of off -target mutations and
26:40
unintended chromosomal rearrangements. Because
26:43
a huge percentage of human genetic diseases are
26:45
caused by point mutations. A single incorrect
26:48
letter out of 3 billion base editing is increasingly
26:51
viewed as the direction the field is heading
26:53
for maximum safety. Okay, so the underlying mechanisms
26:56
are becoming clearer and the tools are becoming
26:57
remarkably refined. But a listener hearing all
27:00
this molecular theory is inevitably asking, has
27:03
this actually crossed over into reality? Right,
27:05
has it cured anyone? Beyond the Petri dish, beyond
27:08
the mice in the laboratory, has a human being
27:11
actually been cured of a disease using CRISPR?
27:14
And the answer is a resounding yes. But the reality
27:17
of the treatment is intense, harrowing, and deeply
27:21
complicated. It is the reality of pushing the
27:23
absolute boundary of medical science. When we
27:26
look at what has been achieved clinically in
27:28
just over a decade since the Doudna -Charpentier
27:30
paper, it is staggering. But we must ground that
27:34
scientific triumph in the physical and economic
27:37
reality of the patient experience. Let's focus
27:39
on the clearest success to date, the first approved
27:42
CRISPR therapy for humans, which targets blood
27:45
disorders like sickle cell disease and beta thalassemia.
27:48
Late 2023, regulators in the U .S. and U .K.
27:51
approved the first CRISPR -based therapy for
27:53
this. And to understand the cure, we have to
27:55
understand the sheer brutality of sickle cell
27:58
disease itself. It's caused by faults in hemoglobin,
28:01
right? Sickle cell disease is a devastating inherited
28:03
blood disorder. It is caused by a single -point
28:06
mutation, one single incorrect genetic letter
28:09
in the gene that produces hemoglobin. Hemoglobin
28:13
is the vital protein packed inside our red blood
28:15
cells. It is the molecular machine that grabs
28:18
oxygen in our lungs and delivers it to every
28:20
tissue in our body. In a healthy person, red
28:23
blood cells are round, squishy, and flexible,
28:25
allowing them to squeeze through microscopic
28:27
capillaries. But in sickle cell, that one tiny
28:30
genetic error changes the shape of the hemoglobin
28:33
molecule. The mutation substitutes a single amino
28:36
acid, which causes the hemoglobin molecules inside
28:39
the cell to polymerize, or link together into
28:42
long, rigid chains when they release oxygen.
28:44
And that deforms the cell. These rigid chains
28:46
physically deform the red blood cell, warping
28:49
it from a squishy disk into a hard... And the
29:02
result of that blockage is what patients call
29:04
a pain crisis. I've read accounts from patients
29:08
like Victoria Gray, who was the first American
29:10
treated with the CRISPR therapy. She described
29:12
the pain as feeling like lightning strikes in
29:14
her bones or being crushed by a truck. It is
29:17
a horrific burden. It requires heavy intravenous
29:20
opioids to manage. It causes cumulative catastrophic
29:23
damage to the heart, lungs, and kidneys, drastically
29:27
shortening the patient's lifespan. So when regulators
29:30
approved a CRISPR -based therapy called Casgevi
29:32
in late 2023 to treat this, it was rightfully
29:36
heralded as a historic milestone. Yeah. But the
29:39
therapeutic approach they used is incredibly
29:40
ingenious because they didn't actually do what
29:42
you might expect. Right. They didn't use that
29:44
precise template method to go in and fix the
29:47
broken spelling in the adult hemoglobin gene.
29:49
Because we established earlier that the precise
29:52
repair method is highly inefficient. Trying to
29:55
perfectly correct the broken gene in billions
29:57
of a patient's blood stem cells just isn't feasible
30:00
yet. So what did they do instead? They engineered
30:02
a brilliant biological workaround using the sloppy
30:05
repair method. The therapy relies on a fascinating
30:09
quirk of human biology. Humans actually possess
30:12
two entirely different sets of genetic instructions
30:15
for making hemoglobin. Wait, two? We have a fetal
30:18
version, which we use while we are developing
30:20
in the womb, and an adult version, which we use
30:22
for the rest of our lives. Why do we have two
30:25
different versions? Because a fetus needs a type
30:27
of hemoglobin that is incredibly aggressive at
30:30
stealing oxygen from the mother's bloodstream
30:32
through the placenta. Fetal hemoglobin binds
30:35
oxygen much more tightly. That makes sense. But
30:37
shortly after birth, we don't need that aggressive
30:40
fetal version anymore. So biology flips a genetic
30:43
switch specifically, a regulatory enhancer called
30:46
BCL11A, which effectively silences the fetal
30:50
hemoglobin gene and turns on the adult hemoglobin
30:53
gene. So in a patient with sickle cell disease,
30:55
the mutation only exists in the adult gene. Their
30:58
fetal hemoglobin gene, which is currently turned
31:00
off and gathering dust in the genome, is perfectly
31:03
healthy. Exactly. The fetal gene has no mutation
31:06
whatsoever. It is a perfectly functional backup
31:09
oxygen carrier just sitting there in the DNA.
31:12
So the scientists realize, what if we don't try
31:16
to fix the broken adult gene? What if we just
31:19
use the CRISPR scissors to break the genetic
31:21
switch that is keeping the healthy fetal gene
31:24
turned off? That is an incredibly clever lateral
31:26
thinking solution. They program the guide RNA
31:29
to target the BCL11A genetic switch. Cas9 makes
31:33
the cut. The cell panics, uses the sloppy repair
31:36
method, scrambles the sequence, and permanently
31:38
breaks the switch. And then what happens? With
31:40
the suppressor destroyed, the cell essentially
31:42
reverts to its biological state in the womb.
31:45
It starts churning out massive quantities of
31:47
healthy, functional fetal hemoglobin. And because
31:50
fetal hemoglobin doesn't polymerize, it prevents
31:53
the red blood cells from sickling, even with
31:55
the faulty adult hemoglobin still present. It
31:57
is a genuine cure. Patients who suffered recurrent
32:00
pain crises and needed constant transfusions
32:02
are going years without them. It is nothing short
32:05
of miraculous. Scientifically, it is a triumph.
32:07
And we should briefly touch on other applications,
32:09
too. Yeah, tell me about those. There is progress
32:11
in inherited blindness, where surgeons are injecting
32:14
the CRISPR machinery directly under the retina.
32:17
In oncology, they are extracting patients' immune
32:20
cells using CRISPR to turn them into hyperaggressive
32:23
cancer hunters and infusing them back in. There
32:26
is even a rare liver condition treated via a
32:29
single infusion of lipid nanoparticles carrying
32:31
CRISPR. It's moving so fast. And agricultural
32:34
uses, too, disease and drought resistance in
32:37
crops without introducing foreign genes like
32:39
older GM crops. This sounds like a miracle, but
32:42
I'm guessing this isn't as simple as taking a
32:44
pill. No, not at all. Because a listener might
32:46
assume that treating sickle cell in 2026 involves
32:48
going to a clinic, getting an IV drip for a few
32:51
hours and going home. It is worlds away from
32:54
a simple IV drip. The harsh reality of this therapy
32:57
is grueling. It is highly invasive and deeply
33:00
punishing on the patient's body. Why is it so
33:02
intense? Remember, to cure a blood disorder permanently,
33:05
you cannot just edit the circulating red blood
33:08
cells. because they die and are replaced every
33:10
few months. You have to edit the source. You
33:13
have to edit the blood -producing factory, which
33:16
resides in the stem cells living deep inside
33:18
the bone marrow. And we don't currently have
33:20
a safe way to send CRISPR directly into the bone
33:23
marrow inside a living person, right? Exactly.
33:26
So the process must be done outside the body.
33:29
First, doctors must remove a patient's blood
33:32
stem cells. Those cells are then edited in a
33:35
highly specialized manufacturing lab over several
33:38
weeks. But the terrifying part is what happens
33:41
right before those edited cells are returned
33:43
to the patient. The preparation for receiving
33:45
the cells is the most dangerous part. Yeah. The
33:47
patient is admitted to the hospital and given
33:49
harsh chemotherapy. Why do they need chemotherapy
33:51
if they don't have cancer? Because you have to
33:53
clear out their bone marrow. If you just infuse
33:56
the new cells, they won't have anywhere to graft
33:58
and grow. The chemotherapy is designed to intentionally
34:01
wipe out their entire immune system and blood
34:03
-producing capacity. Oh my god. They lose their
34:06
hair, suffer severe mouth sores, and are kept
34:08
in strict isolation because a common cold could
34:11
kill them instantly. Only then are the edited
34:14
cells returned. It is a systemic reboot of the
34:17
human body. The physical toll is unimaginable.
34:21
But that physical toll leads directly to the
34:24
access dilemma. Yes. These therapies cost millions
34:27
per patient. And this price tag highlights a
34:30
deeply tragic irony. Sickle cell disease is most
34:34
common in sub -Saharan Africa and India. These
34:37
are precisely the regions of the world where
34:39
a million dollar chemotherapy dependent hospital
34:42
infrastructure is absolutely fundamentally unavailable
34:46
to the masses. We've literally invented the cure
34:48
and we're administering it to a few dozen people
34:51
in wealthy nations while hundreds of thousands
34:53
of children in Africa will continue to suffer
34:55
because the delivery mechanism is too complex.
34:57
It is a devastating reality. And if treating
34:59
a living patient, subjecting them to millions
35:01
of dollars of grueling, highly toxic chemotherapy
35:04
is this. and expensive, it raises an inevitable
35:07
question. It does. Why wait? Why wait until a
35:10
person is born to fix their genes and put them
35:13
through that? Why not edit them at the very beginning?
35:15
And that brings us to the ethical heart of the
35:17
CRISPR debate. Because intervening at the embryonic
35:19
stage crosses a boundary that terrified the scientific
35:22
community from the moment CRISPR was invented.
35:25
And to understand why... we have to clearly define
35:28
the fundamental distinction that underpins this
35:31
entire ethical conversation somatic editing versus
35:34
germline editing exactly the entire ethical debate
35:37
hinges on one distinction If you don't understand
35:40
the difference, you can't grasp why certain CRISPR
35:43
applications win Nobel Prizes while others result
35:46
in prison sentences. So let's lay it out. The
35:49
sickle cell treatment we just discussed, extracting
35:51
bone marrow, editing it, and putting it back,
35:54
that is somatic editing. What does that mean?
35:56
The word soma comes from the Greek word for body.
35:59
Somatic editing means editing the cells of a
36:01
living patient, their blood, liver, immune cells.
36:04
And the ethics there are broadly accepted, right?
36:07
Yes. The changes affect only that person. If
36:11
I edit your liver, those edited genes die with
36:14
you. They cannot be passed down to your children.
36:16
So it's weighed like any other medical intervention,
36:19
risks, benefits, informed consent. The risk is
36:22
confined. But the other category, germline editing,
36:25
fundamentally changes the stakes. What is germline
36:28
editing? Germline editing means editing sperm,
36:31
eggs, or embryos. And why is that categorically
36:35
different? Because of human development. If you
36:37
edit a single cell embryo, that exact genetic
36:40
edit is perfectly copied into every single cell
36:43
of the resulting human being. And crucially,
36:45
it will be in their reproductive cells. Meaning
36:48
the edit is permanent and it is inheritable.
36:50
Exactly. The changes appear in every cell of
36:53
the resulting person and are inherited by all
36:55
their descendants indefinitely. You are permanently
36:58
introducing an artificial genetic modification
37:00
into the human gene pool. So the ethical burden
37:04
is infinitely heavier. If a germline edit goes
37:07
wrong, you haven't just harmed one person. You've
37:09
permanently encoded a man -made genetic error
37:12
into the human lineage. And the future generations
37:14
can't consent. Precisely. Because of that, the
37:18
scientific community had a strong consensus.
37:20
Germline editing of human embryos was a hard
37:23
red line. But in 2018, that red line was crossed.
37:26
The He Jiankui scandal. Walk us through what
37:28
happened. In 2018, a Chinese scientist named
37:31
He Jiankui announced the birth of twin girls,
37:34
Lulu and Nana, with edited embryos. He claimed
37:37
he had altered their germline. What was he trying
37:39
to edit? Was it a deadly disease? No. He intended
37:42
to give them a genetic advantage resistance to
37:44
HIV. He targeted a gene called CCR5, trying to
37:48
recreate a natural mutation that makes people
37:50
resistant to the virus. And what was the global
37:52
reaction? Was he celebrated? It was near universal
37:55
condemnation. He was utterly excoriated. Why
37:58
was it so universally condemned? Let's break
37:59
down his failures. First, it was medically unnecessary.
38:03
He justified it because the father was HIV positive.
38:06
But the mother was negative, and safe ways to
38:08
prevent HIV transmission already exist. You don't
38:11
need experimental germline editing for that.
38:13
Second. Consent was inadequate. The parents didn't
38:16
fully comprehend the novelty in generational
38:18
risks. And the editing itself. It was imprecise.
38:21
He produced unintended changes, novel mutations
38:24
in the CCR5 gene. Nobody knows what these new
38:28
mutations will do. He was imprisoned, and the
38:31
scientific community paused. heritable editing.
38:33
OK, that case was botched and unnecessary. But
38:36
let's look forward. What happens when the technology
38:39
is genuinely safe and precise? The uncomfortable
38:41
gray area. Yes. If two parents carry a devastating
38:45
genetic disease and we could spare their child
38:48
and every future generation from suffering. Doesn't
38:51
it become uncomfortable to say no? It does. It
38:54
becomes incredibly difficult. Now, it's worth
38:56
noting that embryo screening currently handles
38:58
most of these cases safely. You can select embryos
39:00
that don't carry the disease without editing
39:02
them. True, but screening has limits. What if
39:04
they don't have healthy embryos? Or what happens
39:07
when we move past severe disease? This is the
39:09
gray area. If it's safe enough to prevent a fatal
39:11
disease, where do we draw the line between disease
39:13
prevention and enhancement? And that is the slippery
39:16
slope. Once you open the door to germline editing,
39:19
how do you prevent it from sliding into enhancement?
39:22
Parents demanding edits for height, muscle, or
39:25
cognition. If I can legally ensure my child won't
39:29
get Alzheimer's, by what logic can a doctor tell
39:31
me I can't also ensure they are six feet tall
39:34
with perfect vision? As a parent, Isn't it my
39:37
instinct to give my child an advantage? It is.
39:40
And the brutal truth is that nobody has a coherent
39:43
ethical framework for this. Biology doesn't draw
39:46
a clear line. A genetic tweak can be both a cure
39:49
and an enhancement. And cultural viewpoints differ
39:52
widely. What's seen as playing God in the West
39:55
might be seen as a parental duty elsewhere. It
39:58
threatens to create literal biological inequality.
40:01
But the concept of making irreversible, inheritable
40:04
changes doesn't just apply to human embryos,
40:06
does it? No, it applies to nature, too. If we
40:09
can alter the human gene pool, what happens when
40:11
we decide to alter an entire ecosystem? That
40:13
brings us to gene drives. Gene drives? What are
40:15
they? Gene drives are CRISPR constructs engineered
40:18
to cheat the normal rules of inheritance. How
40:21
do they cheat? Normally, there's a 50 -50 chance
40:23
a gene is passed on. But a gene drive... encodes
40:27
the CRISPR machinery right into the animal's
40:29
DNA. When the animal mates, the drive wakes up,
40:32
cuts the normal gene from the other parent, and
40:35
copies itself. It changes that 50 -50 coin toss
40:38
into a 100 % guarantee. So every single offspring
40:42
inherits the modification. And they pass it to
40:45
100 % of their offspring. They spread a genetic
40:47
modification through a wild population incredibly
40:50
fast, potentially reaching every individual in
40:53
a few generations. It's like releasing a biological
40:55
software update into the wild that automatically
40:58
installs itself on every computer it touches.
41:01
That is a terrifyingly accurate analogy. But
41:03
why would we do that? What is the ultimate trolley
41:06
problem here? The target is malaria. Hundreds
41:08
of thousands die annually, mostly children. The
41:11
obvious application of a gene drive is to push
41:13
a modification like sterility through mosquito
41:16
populations to eliminate the disease. You release
41:19
a few engineered mosquitoes, the sterility spreads,
41:21
the population crashes, and you save hundreds
41:23
of thousands of lives. It sounds so tempting.
41:25
It does. But going back to my software update
41:28
analogy, what if there's a bug in the code? That
41:32
is the terrifying ecological risk. Releasing
41:34
a self -propagating genetic change into an ecosystem
41:37
is not reversible. Ecosystems are incredibly
41:40
complex, and we cannot easily predict the knock
41:43
-on effects of rapidly altering or crashing a
41:46
wild species. What if a new, worse disease vector
41:49
moves in? What if the drive mutates and crosses
41:52
into pollinator species? The science is answering
41:55
the technical questions of how to rewrite life,
41:57
but the question of what we actually do with
41:59
that power is no longer up to biologists. It's
42:03
up to all of us. It has to be. Let's do a quick
42:05
recap. We journeyed from a weird repeating pattern
42:07
in bacterial DNA that yogurt makers cared about,
42:10
to programmable molecular scissors, to a functional
42:13
cure for previously incurable blood diseases,
42:16
and finally to the ethical precipice of altering
42:18
human descendants and wild ecosystems. And remember,
42:21
CRISPR is cheap. It's accessible. Its technical
42:23
limitations, precision, cost, delivery are steadily
42:26
shrinking. Its use will only broaden. Which leaves
42:29
you, the listener, with a final lingering thought
42:32
to mull over. We've spent our whole history adapting
42:35
ourselves and our societies to the limitations
42:37
of our biology and our environment. But now that
42:40
we hold the pen to rewrite the code of living
42:42
things directly, will we have the wisdom to know
42:45
when to stop editing? That is the big question.
42:48
If the tool is cheap and accessible enough, who
42:50
gets to decide what a perfect human or a perfect
42:53
ecosystem even looks like? It's a heavy thought.
42:55
It really is. Thank you for joining us on this
42:57
deep dive into CRISPR. Keep questioning, keep
43:00
exploring, and we will see you next time.