TED2019
David R. Liu: Can we cure genetic diseases by rewriting DNA?
Filmed:
Readability: 6.4
1,976,595 views
In a story of scientific discovery, chemical biologist David R. Liu shares a breakthrough: his lab's development of base editors that can rewrite DNA. This crucial step in genome editing takes the promise of CRISPR to the next level: if CRISPR proteins are molecular scissors, programmed to cut specific DNA sequences, then base editors are pencils, capable of directly rewriting one DNA letter into another. Learn more about how these molecular machines work -- and their potential to treat or even cure genetic diseases.
David R. Liu - Chemical biologist
David R. Liu leads a research group that combines chemistry and evolutionary techniques to create revolutionary new medicines. Full bio
David R. Liu leads a research group that combines chemistry and evolutionary techniques to create revolutionary new medicines. Full bio
Double-click the English transcript below to play the video.
00:13
The most important gift
your mother and father ever gave you
your mother and father ever gave you
0
1286
4031
00:17
was the two sets
of three billion letters of DNA
of three billion letters of DNA
1
5341
2720
00:20
that make up your genome.
2
8085
1564
00:22
But like anything
with three billion components,
with three billion components,
3
10014
2477
00:24
that gift is fragile.
4
12515
1400
00:26
Sunlight, smoking, unhealthy eating,
5
14815
3540
00:30
even spontaneous mistakes
made by your cells,
made by your cells,
6
18379
2992
00:33
all cause changes to your genome.
7
21395
1923
00:36
The most common kind of change in DNA
8
24942
3278
00:40
is the simple swap of one letter,
or base, such as C,
or base, such as C,
9
28244
4229
00:44
with a different letter,
such as T, G or A.
such as T, G or A.
10
32497
3241
00:48
In any day, the cells in your body
will collectively accumulate
will collectively accumulate
11
36744
3373
00:52
billions of these single-letter swaps,
which are also called "point mutations."
which are also called "point mutations."
12
40141
4836
00:58
Now, most of these
point mutations are harmless.
point mutations are harmless.
13
46147
2531
01:00
But every now and then,
14
48702
1158
01:01
a point mutation disrupts
an important capability in a cell
an important capability in a cell
15
49884
3993
01:05
or causes a cell to misbehave
in harmful ways.
in harmful ways.
16
53901
3355
01:10
If that mutation were inherited
from your parents
from your parents
17
58099
2999
01:13
or occurred early enough
in your development,
in your development,
18
61122
2660
01:15
then the result would be
that many or all of your cells
that many or all of your cells
19
63806
2966
01:18
contain this harmful mutation.
20
66796
1912
01:21
And then you would be one
of hundreds of millions of people
of hundreds of millions of people
21
69153
3270
01:24
with a genetic disease,
22
72447
1611
01:26
such as sickle cell anemia or progeria
23
74082
3003
01:29
or muscular dystrophy
or Tay-Sachs disease.
or Tay-Sachs disease.
24
77109
3121
01:34
Grievous genetic diseases
caused by point mutations
caused by point mutations
25
82225
3182
01:37
are especially frustrating,
26
85431
1993
01:39
because we often know
the exact single-letter change
the exact single-letter change
27
87448
2904
01:42
that causes the disease
and, in theory, could cure the disease.
and, in theory, could cure the disease.
28
90376
4200
01:47
Millions suffer from sickle cell anemia
29
95268
2849
01:50
because they have
a single A to T point mutations
a single A to T point mutations
30
98141
3071
01:53
in both copies of their hemoglobin gene.
31
101236
2361
01:57
And children with progeria
are born with a T
are born with a T
32
105529
3132
02:00
at a single position in their genome
33
108685
2168
02:02
where you have a C,
34
110877
1399
02:05
with the devastating consequence
that these wonderful, bright kids
that these wonderful, bright kids
35
113125
3439
02:08
age very rapidly and pass away
by about age 14.
by about age 14.
36
116588
3976
02:14
Throughout the history of medicine,
37
122358
1683
02:16
we have not had a way
to efficiently correct point mutations
to efficiently correct point mutations
38
124065
3060
02:19
in living systems,
39
127149
1769
02:20
to change that disease-causing
T back into a C.
T back into a C.
40
128942
3200
02:25
Perhaps until now.
41
133482
1968
02:27
Because my laboratory recently succeeded
in developing such a capability,
in developing such a capability,
42
135474
4190
02:31
which we call "base editing."
43
139688
1800
02:35
The story of how we developed base editing
44
143277
2024
02:37
actually begins three billion years ago.
45
145325
2674
02:41
We think of bacteria
as sources of infection,
as sources of infection,
46
149055
2660
02:43
but bacteria themselves are also
prone to being infected,
prone to being infected,
47
151739
3314
02:47
in particular, by viruses.
48
155077
1907
02:49
So about three billion years ago,
49
157871
2151
02:52
bacteria evolved a defense mechanism
to fight viral infection.
to fight viral infection.
50
160046
3880
02:57
That defense mechanism
is now better known as CRISPR.
is now better known as CRISPR.
51
165649
2785
03:01
And the warhead in CRISPR
is this purple protein
is this purple protein
52
169008
2825
03:03
that acts like molecular
scissors to cut DNA,
scissors to cut DNA,
53
171857
3778
03:07
breaking the double helix into two pieces.
54
175659
2428
03:11
If CRISPR couldn't distinguish
between bacterial and viral DNA,
between bacterial and viral DNA,
55
179323
3976
03:15
it wouldn't be a very useful
defense system.
defense system.
56
183323
2239
03:18
But the most amazing feature of CRISPR
57
186315
2785
03:21
is that the scissors can be
programmed to search for,
programmed to search for,
58
189124
5037
03:26
bind to and cut
59
194185
2423
03:28
only a specific DNA sequence.
60
196632
2738
03:32
So when a bacterium encounters
a virus for the first time,
a virus for the first time,
61
200911
3397
03:36
it can store a small snippet
of that virus's DNA
of that virus's DNA
62
204332
3373
03:39
for use as a program
to direct the CRISPR scissors
to direct the CRISPR scissors
63
207729
3644
03:43
to cut that viral DNA sequence
during a future infection.
during a future infection.
64
211397
3536
03:47
Cutting a virus's DNA messes up
the function of the cut viral gene,
the function of the cut viral gene,
65
215778
4913
03:52
and therefore disrupts
the virus's life cycle.
the virus's life cycle.
66
220715
2702
03:58
Remarkable researchers including
Emmanuelle Charpentier, George Church,
Emmanuelle Charpentier, George Church,
67
226059
4801
04:02
Jennifer Doudna and Feng Zhang
68
230884
2653
04:05
showed six years ago how CRISPR scissors
could be programmed
could be programmed
69
233561
3969
04:09
to cut DNA sequences of our choosing,
70
237554
2587
04:12
including sequences in your genome,
71
240165
2369
04:14
instead of the viral DNA sequences
chosen by bacteria.
chosen by bacteria.
72
242558
3343
04:18
But the outcomes are actually similar.
73
246550
2534
04:21
Cutting a DNA sequence in your genome
74
249606
2468
04:24
also disrupts the function
of the cut gene, typically,
of the cut gene, typically,
75
252098
4127
04:28
by causing the insertion and deletion
of random mixtures of DNA letters
of random mixtures of DNA letters
76
256997
4467
04:33
at the cut site.
77
261488
1153
04:36
Now, disrupting genes can be very
useful for some applications.
useful for some applications.
78
264625
3881
04:42
But for most point mutations
that cause genetic diseases,
that cause genetic diseases,
79
270005
4301
04:46
simply cutting the already-mutated gene
won't benefit patients,
won't benefit patients,
80
274330
4357
04:50
because the function of the mutated gene
needs to be restored,
needs to be restored,
81
278711
3968
04:54
not further disrupted.
82
282703
1615
04:57
So cutting this
already-mutated hemoglobin gene
already-mutated hemoglobin gene
83
285259
2882
05:00
that causes sickle cell anemia
84
288165
2523
05:02
won't restore the ability of patients
to make healthy red blood cells.
to make healthy red blood cells.
85
290712
3516
05:07
And while we can sometimes introduce
new DNA sequences into cells
new DNA sequences into cells
86
295631
4341
05:11
to replace the DNA sequences
surrounding a cut site,
surrounding a cut site,
87
299996
3421
05:15
that process, unfortunately, doesn't work
in most types of cells,
in most types of cells,
88
303441
4324
05:19
and the disrupted gene outcomes
still predominate.
still predominate.
89
307789
2441
05:24
Like many scientists,
I've dreamed of a future
I've dreamed of a future
90
312297
2182
05:26
in which we might be able to treat
or maybe even cure
or maybe even cure
91
314503
2774
05:29
human genetic diseases.
92
317301
1371
05:31
But I saw the lack of a way
to fix point mutations,
to fix point mutations,
93
319135
3801
05:34
which cause most human genetic diseases,
94
322960
3024
05:38
as a major problem standing in the way.
95
326008
2388
05:41
Being a chemist, I began
working with my students
working with my students
96
329434
2668
05:44
to develop ways on performing chemistry
directly on an individual DNA base,
directly on an individual DNA base,
97
332126
4935
05:49
to truly fix, rather than disrupt,
the mutations that cause genetic diseases.
the mutations that cause genetic diseases.
98
337085
5619
05:56
The results of our efforts
are molecular machines
are molecular machines
99
344522
2548
05:59
called "base editors."
100
347094
1388
06:01
Base editors use the programmable
searching mechanism of CRISPR scissors,
searching mechanism of CRISPR scissors,
101
349618
5475
06:07
but instead of cutting the DNA,
102
355117
2936
06:10
they directly convert
one base to another base
one base to another base
103
358077
2941
06:13
without disrupting the rest of the gene.
104
361042
2253
06:16
So if you think of naturally occurring
CRISPR proteins as molecular scissors,
CRISPR proteins as molecular scissors,
105
364674
4158
06:20
you can think of base editors as pencils,
106
368856
2786
06:23
capable of directly rewriting
one DNA letter into another
one DNA letter into another
107
371666
3496
06:28
by actually rearranging
the atoms of one DNA base
the atoms of one DNA base
108
376098
3803
06:31
to instead become a different base.
109
379925
2334
06:35
Now, base editors don't exist in nature.
110
383513
2176
06:38
In fact, we engineered
the first base editor, shown here,
the first base editor, shown here,
111
386683
3230
06:41
from three separate proteins
112
389937
1357
06:43
that don't even come
from the same organism.
from the same organism.
113
391318
2230
06:46
We started by taking CRISPR scissors
and disabling the ability to cut DNA
and disabling the ability to cut DNA
114
394151
5097
06:51
while retaining its ability to search for
and bind a target DNA sequence
and bind a target DNA sequence
115
399272
4539
06:55
in a programmed manner.
116
403835
1534
06:58
To those disabled CRISPR
scissors, shown in blue,
scissors, shown in blue,
117
406351
2837
07:01
we attached a second protein in red,
118
409212
2508
07:03
which performs a chemical reaction
on the DNA base C,
on the DNA base C,
119
411744
4301
07:08
converting it into a base
that behaves like T.
that behaves like T.
120
416069
3333
07:12
Third, we had to attach
to the first two proteins
to the first two proteins
121
420958
3142
07:16
the protein shown in purple,
122
424124
1350
07:17
which protects the edited base
from being removed by the cell.
from being removed by the cell.
123
425498
3600
07:22
The net result is an engineered
three-part protein
three-part protein
124
430466
2842
07:25
that for the first time
allows us to convert Cs into Ts
allows us to convert Cs into Ts
125
433332
4118
07:29
at specified locations in the genome.
126
437474
2163
07:33
But even at this point,
our work was only half done.
our work was only half done.
127
441490
3032
07:36
Because in order to be stable in cells,
128
444546
2626
07:39
the two strands of a DNA double helix
have to form base pairs.
have to form base pairs.
129
447196
3659
07:44
And because C only pairs with G,
130
452125
3658
07:47
and T only pairs with A,
131
455807
3002
07:51
simply changing a C to a T
on one DNA strand creates a mismatch,
on one DNA strand creates a mismatch,
132
459752
4846
07:56
a disagreement between the two DNA strands
133
464622
2849
07:59
that the cell has to resolve
by deciding which strand to replace.
by deciding which strand to replace.
134
467495
4268
08:05
We realized that we could further engineer
this three-part protein
this three-part protein
135
473149
4341
08:10
to flag the nonedited strand
as the one to be replaced
as the one to be replaced
136
478649
3866
08:14
by nicking that strand.
137
482539
1911
08:17
This little nick tricks the cell
138
485276
2529
08:19
into replacing the nonedited G with an A
139
487829
4947
08:24
as it remakes the nicked strand,
140
492800
2325
08:27
thereby completing the conversion
of what used to be a C-G base pair
of what used to be a C-G base pair
141
495149
4031
08:31
into a stable T-A base pair.
142
499204
2296
08:36
After several years of hard work
143
504585
1551
08:38
led by a former post doc
in the lab, Alexis Komor,
in the lab, Alexis Komor,
144
506160
3981
08:42
we succeeded in developing
this first class of base editor,
this first class of base editor,
145
510165
3182
08:45
which converts Cs into Ts and Gs into As
146
513371
3666
08:49
at targeted positions of our choosing.
147
517061
2159
08:52
Among the more than 35,000 known
disease-associated point mutations,
disease-associated point mutations,
148
520633
5230
08:57
the two kinds of mutations
that this first base editor can reverse
that this first base editor can reverse
149
525887
3785
09:01
collectively account for about 14 percent
or 5,000 or so pathogenic point mutations.
or 5,000 or so pathogenic point mutations.
150
529696
6143
09:08
But correcting the largest fraction
of disease-causing point mutations
of disease-causing point mutations
151
536593
4770
09:13
would require developing
a second class of base editor,
a second class of base editor,
152
541387
3635
09:17
one that could convert
As into Gs or Ts into Cs.
As into Gs or Ts into Cs.
153
545046
4086
09:22
Led by Nicole Gaudelli,
a former post doc in the lab,
a former post doc in the lab,
154
550846
3727
09:26
we set out to develop
this second class of base editor,
this second class of base editor,
155
554597
3122
09:29
which, in theory, could correct up to
almost half of pathogenic point mutations,
almost half of pathogenic point mutations,
156
557743
6127
09:35
including that mutation that causes
the rapid-aging disease progeria.
the rapid-aging disease progeria.
157
563894
3911
09:42
We realized that we could
borrow, once again,
borrow, once again,
158
570107
3167
09:45
the targeting mechanism of CRISPR scissors
159
573298
4068
09:49
to bring the new base editor
to the right site in a genome.
to the right site in a genome.
160
577390
5161
09:55
But we quickly encountered
an incredible problem;
an incredible problem;
161
583543
3092
09:59
namely, there is no protein
162
587896
2428
10:02
that's known to convert
A into G or T into C
A into G or T into C
163
590348
4052
10:06
in DNA.
164
594424
1161
10:08
Faced with such a serious stumbling block,
165
596760
2166
10:10
most students would probably
look for another project,
look for another project,
166
598950
2532
10:13
if not another research advisor.
167
601506
1740
10:15
(Laughter)
168
603270
1164
10:16
But Nicole agreed to proceed with a plan
169
604458
1942
10:18
that seemed wildly ambitious at the time.
170
606424
2667
10:21
Given the absence
of a naturally occurring protein
of a naturally occurring protein
171
609966
2339
10:24
that performs the necessary chemistry,
172
612329
2161
10:26
we decided we would evolve
our own protein in the laboratory
our own protein in the laboratory
173
614514
3436
10:29
to convert A into a base
that behaves like G,
that behaves like G,
174
617974
3835
10:33
starting from a protein
that performs related chemistry on RNA.
that performs related chemistry on RNA.
175
621833
4827
10:39
We set up a Darwinian
survival-of-the-fittest selection system
survival-of-the-fittest selection system
176
627230
3934
10:43
that explored tens of millions
of protein variants
of protein variants
177
631188
3992
10:47
and only allowed those rare variants
178
635204
2018
10:49
that could perform the necessary
chemistry to survive.
chemistry to survive.
179
637246
3221
10:53
We ended up with a protein shown here,
180
641883
2388
10:56
the first that can convert A in DNA
181
644295
2857
10:59
into a base that resembles G.
182
647176
2092
11:01
And when we attached that protein
183
649292
1603
11:02
to the disabled CRISPR
scissors, shown in blue,
scissors, shown in blue,
184
650919
2571
11:05
we produced the second base editor,
185
653514
2008
11:07
which converts As into Gs,
186
655546
3095
11:10
and then uses the same
strand-nicking strategy
strand-nicking strategy
187
658665
3841
11:14
that we used in the first base editor
188
662530
1920
11:16
to trick the cell into replacing
the nonedited T with a C
the nonedited T with a C
189
664474
5465
11:21
as it remakes that nicked strand,
190
669963
1675
11:23
thereby completing the conversion
of an A-T base pair to a G-C base pair.
of an A-T base pair to a G-C base pair.
191
671662
4171
11:28
(Applause)
192
676845
2047
11:30
Thank you.
193
678916
1170
11:32
(Applause)
194
680110
3357
11:35
As an academic scientist in the US,
195
683491
2335
11:37
I'm not used to being
interrupted by applause.
interrupted by applause.
196
685850
2147
11:40
(Laughter)
197
688021
3151
11:43
We developed these
first two classes of base editors
first two classes of base editors
198
691196
4405
11:47
only three years ago
and one and a half years ago.
and one and a half years ago.
199
695625
2774
11:51
But even in that short time,
200
699267
1548
11:52
base editing has become widely used
by the biomedical research community.
by the biomedical research community.
201
700839
3722
11:57
Base editors have been sent
more than 6,000 times
more than 6,000 times
202
705776
4365
12:02
at the request of more than
1,000 researchers around the globe.
1,000 researchers around the globe.
203
710165
3871
12:07
A hundred scientific research papers
have been published already,
have been published already,
204
715475
3516
12:11
using base editors in organisms
ranging from bacteria
ranging from bacteria
205
719015
3728
12:14
to plants to mice to primates.
206
722767
2134
12:19
While base editors are too new
207
727950
1607
12:21
to have already entered
human clinical trials,
human clinical trials,
208
729581
2885
12:24
scientists have succeeded in achieving
a critical milestone towards that goal
a critical milestone towards that goal
209
732490
5122
12:29
by using base editors in animals
210
737636
2849
12:32
to correct point mutations
that cause human genetic diseases.
that cause human genetic diseases.
211
740509
3909
12:37
For example,
212
745815
1151
12:38
a collaborative team of scientists
led by Luke Koblan and Jon Levy,
led by Luke Koblan and Jon Levy,
213
746990
3793
12:42
two additional students in my lab,
214
750807
2413
12:45
recently used a virus to deliver
that second base editor
that second base editor
215
753244
4119
12:49
into a mouse with progeria,
216
757387
2190
12:51
changing that disease-causing
T back into a C
T back into a C
217
759601
3857
12:55
and reversing its consequences
at the DNA, RNA and protein levels.
at the DNA, RNA and protein levels.
218
763482
4106
13:00
Base editors have also
been used in animals
been used in animals
219
768880
2746
13:03
to reverse the consequence of tyrosinemia,
220
771650
2924
13:07
beta thalassemia, muscular dystrophy,
221
775642
3618
13:11
phenylketonuria, a congenital deafness
222
779284
3690
13:14
and a type of cardiovascular disease --
223
782998
1939
13:16
in each case, by directly
correcting a point mutation
correcting a point mutation
224
784961
4862
13:21
that causes or contributes to the disease.
225
789847
2553
13:25
In plants, base editors have been used
226
793688
2056
13:27
to introduce individual
single DNA letter changes
single DNA letter changes
227
795768
4072
13:31
that could lead to better crops.
228
799864
1968
13:34
And biologists have used base editors
to probe the role of individual letters
to probe the role of individual letters
229
802253
4589
13:38
in genes associated
with diseases such as cancer.
with diseases such as cancer.
230
806866
2817
13:43
Two companies I cofounded,
Beam Therapeutics and Pairwise Plants,
Beam Therapeutics and Pairwise Plants,
231
811046
4567
13:47
are using base editing
to treat human genetic diseases
to treat human genetic diseases
232
815637
3825
13:51
and to improve agriculture.
233
819486
1606
13:53
All of these applications of base editing
234
821953
1966
13:55
have taken place in less
than the past three years:
than the past three years:
235
823943
3094
13:59
on the historical timescale of science,
236
827061
2364
14:01
the blink of an eye.
237
829449
1282
14:04
Additional work lies ahead
238
832657
1253
14:05
before base editing can realize
its full potential
its full potential
239
833934
3032
14:08
to improve the lives of patients
with genetic diseases.
with genetic diseases.
240
836990
3614
14:13
While many of these diseases
are thought to be treatable
are thought to be treatable
241
841244
2780
14:16
by correcting the underlying mutation
242
844048
1849
14:17
in even a modest fraction
of cells in an organ,
of cells in an organ,
243
845921
3516
14:21
delivering molecular machines
like base editors
like base editors
244
849461
2976
14:24
into cells in a human being
245
852461
1767
14:26
can be challenging.
246
854252
1169
14:28
Co-opting nature's viruses
to deliver base editors
to deliver base editors
247
856962
3373
14:32
instead of the molecules
that give you a cold
that give you a cold
248
860359
2198
14:34
is one of several promising
delivery strategies
delivery strategies
249
862581
2687
14:37
that's been successfully used.
250
865292
1659
14:40
Continuing to develop
new molecular machines
new molecular machines
251
868268
2365
14:42
that can make all of the remaining ways
252
870657
1868
14:44
to convert one base pair
to another base pair
to another base pair
253
872549
2892
14:47
and that minimize unwanted editing
at off-target locations in cells
at off-target locations in cells
254
875465
4380
14:51
is very important.
255
879869
1200
14:53
And engaging with other scientists,
doctors, ethicists and governments
doctors, ethicists and governments
256
881782
4706
14:58
to maximize the likelihood
that base editing is applied thoughtfully,
that base editing is applied thoughtfully,
257
886512
4791
15:03
safely and ethically,
258
891327
2381
15:05
remains a critical obligation.
259
893732
2000
15:09
These challenges notwithstanding,
260
897525
1611
15:11
if you had told me
even just five years ago
even just five years ago
261
899160
3655
15:14
that researchers around the globe
262
902839
1651
15:16
would be using laboratory-evolved
molecular machines
molecular machines
263
904514
3539
15:20
to directly convert
an individual base pair
an individual base pair
264
908077
2997
15:23
to another base pair
265
911098
1182
15:24
at a specified location
in the human genome
in the human genome
266
912304
2619
15:26
efficiently and with a minimum
of other outcomes,
of other outcomes,
267
914947
3825
15:30
I would have asked you,
268
918796
1168
15:31
"What science-fiction novel
are you reading?"
are you reading?"
269
919988
2474
15:35
Thanks to a relentlessly dedicated
group of students
group of students
270
923706
3460
15:39
who were creative enough to engineer
what we could design ourselves
what we could design ourselves
271
927190
4460
15:43
and brave enough
to evolve what we couldn't,
to evolve what we couldn't,
272
931674
2925
15:46
base editing has begun to transform
that science-fiction-like aspiration
that science-fiction-like aspiration
273
934623
5040
15:51
into an exciting new reality,
274
939687
1857
15:54
one in which the most important gift
we give our children
we give our children
275
942250
3231
15:57
may not only be
three billion letters of DNA,
three billion letters of DNA,
276
945505
3025
16:00
but also the means to protect
and repair them.
and repair them.
277
948554
3110
16:04
Thank you.
278
952339
1151
16:05
(Applause)
279
953514
4502
16:10
Thank you.
280
958040
1150
ABOUT THE SPEAKER
David R. Liu - Chemical biologistDavid R. Liu leads a research group that combines chemistry and evolutionary techniques to create revolutionary new medicines.
Why you should listen
During his PhD research at Berkeley, David R. Liu initiated the first general effort to expand the genetic code in living cells. As a professor at Harvard and the Broad Institute, Liu integrates chemistry and evolution to illuminate biology and develop next-generation therapeutics. He has published more than 170 papers and is an inventor on more than 65 issued US patents.
Liu's major research interests include development and use of genome editing technologies to study and treat genetic diseases; the evolution of proteins with novel therapeutic potential; and the discovery of bioactive synthetic molecules using DNA-encoded libraries. Base editing, phage-assisted continuous evolution (PACE) and DNA-encoded libraries are three technologies pioneered in his laboratory that are now widely used in the biomedical sciences. Liu has also cofounded six biotechnology and therapeutics companies, including Editas Medicine, Beam Therapeutics, Pairwise Plants and Exo Therapeutics.
Liu grew up in Riverside, California, where playing with insects in his backyard crystallized his interest in science. He also is passionate about photography and has been banned from playing blackjack at virtually every major casino in Las Vegas after developing a creative and highly advantageous card-counting system.
David R. Liu | Speaker | TED.com