1
00:00:00,844 --> 00:00:03,306
So this is a talk about gene drives,

2
00:00:03,330 --> 00:00:05,883
but I'm going to start
by telling you a brief story.

3
00:00:06,621 --> 00:00:09,844
20 years ago, a biologist
named Anthony James

4
00:00:09,868 --> 00:00:12,187
got obsessed with the idea
of making mosquitos

5
00:00:12,211 --> 00:00:14,790
that didn't transmit malaria.

6
00:00:15,683 --> 00:00:19,739
It was a great idea,
and pretty much a complete failure.

7
00:00:20,596 --> 00:00:23,256
For one thing, it turned out
to be really hard

8
00:00:23,280 --> 00:00:25,287
to make a malaria-resistant mosquito.

9
00:00:26,286 --> 00:00:29,913
James managed it, finally,
just a few years ago,

10
00:00:29,937 --> 00:00:32,008
by adding some genes
that make it impossible

11
00:00:32,032 --> 00:00:34,746
for the malaria parasite
to survive inside the mosquito.

12
00:00:35,551 --> 00:00:37,373
But that just created another problem.

13
00:00:38,043 --> 00:00:40,885
Now that you've got
a malaria-resistant mosquito,

14
00:00:40,909 --> 00:00:44,495
how do you get it to replace
all the malaria-carrying mosquitos?

15
00:00:46,107 --> 00:00:47,580
There are a couple options,

16
00:00:47,604 --> 00:00:49,627
but plan A was basically to breed up

17
00:00:49,651 --> 00:00:52,738
a bunch of the new
genetically-engineered mosquitos

18
00:00:52,762 --> 00:00:54,088
release them into the wild

19
00:00:54,112 --> 00:00:56,079
and hope that they pass on their genes.

20
00:00:56,673 --> 00:00:58,787
The problem was that you'd have to release

21
00:00:58,811 --> 00:01:03,111
literally 10 times the number
of native mosquitos to work.

22
00:01:03,135 --> 00:01:05,222
So in a village with 10,000 mosquitos,

23
00:01:05,246 --> 00:01:07,325
you release an extra 100,000.

24
00:01:08,229 --> 00:01:09,379
As you might guess,

25
00:01:09,403 --> 00:01:12,157
this was not a very popular strategy
with the villagers.

26
00:01:12,181 --> 00:01:13,331
(Laughter)

27
00:01:14,963 --> 00:01:18,667
Then, last January,
Anthony James got an email

28
00:01:18,691 --> 00:01:20,642
from a biologist named Ethan Bier.

29
00:01:21,400 --> 00:01:24,259
Bier said that he
and his grad student Valentino Gantz

30
00:01:24,283 --> 00:01:27,227
had stumbled on a tool
that could not only guarantee

31
00:01:27,251 --> 00:01:29,925
that a particular genetic trait
would be inherited,

32
00:01:29,949 --> 00:01:32,028
but that it would spread
incredibly quickly.

33
00:01:32,734 --> 00:01:35,400
If they were right,
it would basically solve the problem

34
00:01:35,424 --> 00:01:37,805
that he and James had been
working on for 20 years.

35
00:01:38,450 --> 00:01:43,108
As a test, they engineered two mosquitos
to carry the anti-malaria gene

36
00:01:43,132 --> 00:01:45,291
and also this new tool, a gene drive,

37
00:01:45,315 --> 00:01:46,806
which I'll explain in a minute.

38
00:01:47,506 --> 00:01:49,839
Finally, they set it up
so that any mosquitos

39
00:01:49,863 --> 00:01:51,801
that had inherited the anti-malaria gene

40
00:01:51,825 --> 00:01:55,824
wouldn't have the usual white eyes,
but would instead have red eyes.

41
00:01:56,682 --> 00:01:58,732
That was pretty much just for convenience

42
00:01:58,756 --> 00:02:01,293
so they could tell just at a glance
which was which.

43
00:02:02,192 --> 00:02:04,949
So they took their two
anti-malarial, red-eyed mosquitos

44
00:02:04,973 --> 00:02:07,632
and put them in a box
with 30 ordinary white-eyed ones,

45
00:02:07,656 --> 00:02:08,830
and let them breed.

46
00:02:09,383 --> 00:02:13,240
In two generations, those had produced
3,800 grandchildren.

47
00:02:14,217 --> 00:02:16,010
That is not the surprising part.

48
00:02:16,716 --> 00:02:18,638
This is the surprising part:

49
00:02:18,662 --> 00:02:21,569
given that you started
with just two red-eyed mosquitos

50
00:02:21,593 --> 00:02:22,927
and 30 white-eyed ones,

51
00:02:22,951 --> 00:02:25,712
you expect mostly white-eyed descendants.

52
00:02:26,514 --> 00:02:29,538
Instead, when James opened the box,

53
00:02:29,562 --> 00:02:32,903
all 3,800 mosquitos had red eyes.

54
00:02:33,315 --> 00:02:35,355
When I asked Ethan Bier about this moment,

55
00:02:35,379 --> 00:02:38,894
he became so excited that he was literally
shouting into the phone.

56
00:02:39,886 --> 00:02:42,049
That's because getting
only red-eyed mosquitos

57
00:02:42,073 --> 00:02:45,021
violates a rule that is the absolute
cornerstone of biology,

58
00:02:45,045 --> 00:02:46,287
Mendelian genetics.

59
00:02:46,912 --> 00:02:48,081
I'll keep this quick,

60
00:02:48,105 --> 00:02:50,869
but Mendelian genetics
says when a male and a female mate,

61
00:02:50,893 --> 00:02:53,535
their baby inherits half
of its DNA from each parent.

62
00:02:53,559 --> 00:02:57,154
So if our original mosquito was aa
and our new mosquito is aB,

63
00:02:57,178 --> 00:02:58,980
where B is the anti-malarial gene,

64
00:02:59,004 --> 00:03:01,391
the babies should come out
in four permutations:

65
00:03:01,415 --> 00:03:03,988
aa, aB, aa, Ba.

66
00:03:04,884 --> 00:03:07,173
Instead, with the new gene drive,

67
00:03:07,197 --> 00:03:08,709
they all came out aB.

68
00:03:09,900 --> 00:03:12,452
Biologically, that shouldn't
even be possible.

69
00:03:12,476 --> 00:03:13,780
So what happened?

70
00:03:14,709 --> 00:03:16,153
The first thing that happened

71
00:03:16,177 --> 00:03:19,447
was the arrival of a gene-editing tool
known as CRISPR in 2012.

72
00:03:20,701 --> 00:03:22,811
Many of you have probably
heard about CRISPR,

73
00:03:22,835 --> 00:03:26,136
so I'll just say briefly that CRISPR
is a tool that allows researchers

74
00:03:26,160 --> 00:03:28,739
to edit genes very precisely,
easily and quickly.

75
00:03:29,533 --> 00:03:33,262
It does this by harnessing a mechanism
that already existed in bacteria.

76
00:03:33,286 --> 00:03:35,912
Basically, there's a protein
that acts like a scissors

77
00:03:35,936 --> 00:03:37,170
and cuts the DNA,

78
00:03:37,194 --> 00:03:39,677
and there's an RNA molecule
that directs the scissors

79
00:03:39,701 --> 00:03:41,419
to any point on the genome you want.

80
00:03:41,443 --> 00:03:44,133
The result is basically
a word processor for genes.

81
00:03:44,157 --> 00:03:46,783
You can take an entire gene
out, put one in,

82
00:03:46,807 --> 00:03:49,252
or even edit just a single
letter within a gene.

83
00:03:49,646 --> 00:03:51,599
And you can do it in nearly any species.

84
00:03:53,328 --> 00:03:57,042
OK, remember how I said that gene drives
originally had two problems?

85
00:03:57,835 --> 00:04:00,955
The first was that it was hard
to engineer a mosquito

86
00:04:00,979 --> 00:04:02,366
to be malaria-resistant.

87
00:04:02,390 --> 00:04:04,517
That's basically gone now,
thanks to CRISPR.

88
00:04:05,117 --> 00:04:06,879
But the other problem was logistical.

89
00:04:07,307 --> 00:04:09,109
How do you get your trait to spread?

90
00:04:10,196 --> 00:04:11,672
This is where it gets clever.

91
00:04:12,943 --> 00:04:16,545
A couple years ago, a biologist
at Harvard named Kevin Esvelt

92
00:04:16,569 --> 00:04:17,919
wondered what would happen

93
00:04:17,943 --> 00:04:21,570
if you made it so that
CRISPR inserted not only your new gene

94
00:04:21,594 --> 00:04:24,435
but also the machinery
that does the cutting and pasting.

95
00:04:25,233 --> 00:04:29,073
In other words, what if CRISPR
also copied and pasted itself.

96
00:04:30,145 --> 00:04:33,276
You'd end up with a perpetual
motion machine for gene editing.

97
00:04:34,458 --> 00:04:36,192
And that's exactly what happened.

98
00:04:37,037 --> 00:04:39,915
This CRISPR gene drive that Esvelt created

99
00:04:39,939 --> 00:04:43,582
not only guarantees
that a trait will get passed on,

100
00:04:43,606 --> 00:04:46,044
but if it's used in the germline cells,

101
00:04:46,068 --> 00:04:48,644
it will automatically copy and paste
your new gene

102
00:04:48,668 --> 00:04:51,314
into both chromosomes
of every single individual.

103
00:04:51,743 --> 00:04:54,140
It's like a global search and replace,

104
00:04:54,164 --> 00:04:57,385
or in science terms, it makes
a heterozygous trait homozygous.

105
00:04:59,045 --> 00:05:01,656
So, what does this mean?

106
00:05:01,680 --> 00:05:04,404
For one thing, it means we have
a very powerful,

107
00:05:04,428 --> 00:05:07,085
but also somewhat alarming new tool.

108
00:05:08,576 --> 00:05:11,465
Up until now, the fact that gene drives
didn't work very well

109
00:05:11,489 --> 00:05:12,972
was actually kind of a relief.

110
00:05:13,425 --> 00:05:16,107
Normally when we mess around
with an organism's genes,

111
00:05:16,131 --> 00:05:18,480
we make that thing
less evolutionarily fit.

112
00:05:18,504 --> 00:05:21,337
So biologists can make
all the mutant fruit flies they want

113
00:05:21,361 --> 00:05:22,614
without worrying about it.

114
00:05:22,638 --> 00:05:25,645
If some escape, natural selection
just takes care of them.

115
00:05:26,750 --> 00:05:29,973
What's remarkable and powerful
and frightening about gene drives

116
00:05:29,997 --> 00:05:31,757
is that that will no longer be true.

117
00:05:33,092 --> 00:05:36,656
Assuming that your trait does not have
a big evolutionary handicap,

118
00:05:36,680 --> 00:05:38,759
like a mosquito that can't fly,

119
00:05:38,783 --> 00:05:42,236
the CRISPR-based gene drive
will spread the change relentlessly

120
00:05:42,260 --> 00:05:45,266
until it is in every single individual
in the population.

121
00:05:47,101 --> 00:05:49,990
Now, it isn't easy to make
a gene drive that works that well,

122
00:05:50,014 --> 00:05:52,188
but James and Esvelt think that we can.

123
00:05:53,339 --> 00:05:57,163
The good news is that this opens
the door to some remarkable things.

124
00:05:57,187 --> 00:05:59,091
If you put an anti-malarial gene drive

125
00:05:59,115 --> 00:06:01,315
in just 1 percent of Anopheles mosquitoes,

126
00:06:01,339 --> 00:06:03,091
the species that transmits malaria,

127
00:06:03,488 --> 00:06:07,607
researchers estimate that it would spread
to the entire population in a year.

128
00:06:08,178 --> 00:06:11,047
So in a year, you could virtually
eliminate malaria.

129
00:06:11,455 --> 00:06:15,327
In practice, we're still a few years out
from being able to do that,

130
00:06:15,351 --> 00:06:18,069
but still, a 1,000 children
a day die of malaria.

131
00:06:18,093 --> 00:06:20,283
In a year, that number
could be almost zero.

132
00:06:20,966 --> 00:06:23,862
The same goes for dengue fever,
chikungunya, yellow fever.

133
00:06:25,211 --> 00:06:26,616
And it gets better.

134
00:06:27,215 --> 00:06:29,691
Say you want to get rid
of an invasive species,

135
00:06:29,715 --> 00:06:31,838
like get Asian carp
out of the Great Lakes.

136
00:06:32,434 --> 00:06:34,458
All you have to do is release a gene drive

137
00:06:34,482 --> 00:06:36,799
that makes the fish produce
only male offspring.

138
00:06:37,386 --> 00:06:41,584
In a few generations,
there'll be no females left, no more carp.

139
00:06:41,608 --> 00:06:44,687
In theory, this means we could restore
hundreds of native species

140
00:06:44,711 --> 00:06:46,378
that have been pushed to the brink.

141
00:06:47,410 --> 00:06:50,697
OK, that's the good news,

142
00:06:50,721 --> 00:06:52,013
this is the bad news.

143
00:06:53,070 --> 00:06:55,098
Gene drives are so effective

144
00:06:55,122 --> 00:06:58,947
that even an accidental release
could change an entire species,

145
00:06:58,971 --> 00:07:00,161
and often very quickly.

146
00:07:01,178 --> 00:07:03,304
Anthony James took good precautions.

147
00:07:03,328 --> 00:07:05,494
He bred his mosquitos
in a bio-containment lab

148
00:07:05,518 --> 00:07:08,106
and he also used a species
that's not native to the US

149
00:07:08,130 --> 00:07:09,685
so that even if some did escape,

150
00:07:09,709 --> 00:07:12,724
they'd just die off, there'd be nothing
for them to mate with.

151
00:07:12,748 --> 00:07:16,771
But it's also true that if a dozen
Asian carp with the all-male gene drive

152
00:07:16,795 --> 00:07:20,724
accidentally got carried
from the Great Lakes back to Asia,

153
00:07:20,748 --> 00:07:24,177
they could potentially wipe out
the native Asian carp population.

154
00:07:25,930 --> 00:07:28,842
And that's not so unlikely,
given how connected our world is.

155
00:07:28,866 --> 00:07:31,437
In fact, it's why we have
an invasive species problem.

156
00:07:31,912 --> 00:07:33,080
And that's fish.

157
00:07:33,428 --> 00:07:36,016
Things like mosquitos and fruit flies,

158
00:07:36,040 --> 00:07:38,000
there's literally no way to contain them.

159
00:07:38,024 --> 00:07:40,135
They cross borders
and oceans all the time.

160
00:07:41,754 --> 00:07:43,936
OK, the other piece of bad news

161
00:07:43,960 --> 00:07:46,278
is that a gene drive
might not stay confined

162
00:07:46,302 --> 00:07:48,115
to what we call the target species.

163
00:07:48,556 --> 00:07:50,055
That's because of gene flow,

164
00:07:50,079 --> 00:07:52,676
which is a fancy way of saying
that neighboring species

165
00:07:52,700 --> 00:07:53,948
sometimes interbreed.

166
00:07:53,972 --> 00:07:57,123
If that happens, it's possible
a gene drive could cross over,

167
00:07:57,147 --> 00:07:59,674
like Asian carp could infect
some other kind of carp.

168
00:07:59,698 --> 00:08:03,413
That's not so bad if your drive
just promotes a trait, like eye color.

169
00:08:03,437 --> 00:08:05,684
In fact, there's a decent
chance that we'll see

170
00:08:05,708 --> 00:08:08,279
a wave of very weird fruit flies
in the near future.

171
00:08:09,390 --> 00:08:10,653
But it could be a disaster

172
00:08:10,677 --> 00:08:13,581
if your drive is deigned
to eliminate the species entirely.

173
00:08:14,351 --> 00:08:17,879
The last worrisome thing
is that the technology to do this,

174
00:08:17,903 --> 00:08:21,594
to genetically engineer an organism
and include a gene drive,

175
00:08:21,618 --> 00:08:24,952
is something that basically any lab
in the world can do.

176
00:08:24,976 --> 00:08:26,316
An undergraduate can do it.

177
00:08:27,209 --> 00:08:30,491
A talented high schooler
with some equipment can do it.

178
00:08:32,526 --> 00:08:34,835
Now, I'm guessing
that this sounds terrifying.

179
00:08:35,351 --> 00:08:37,557
(Laughter)

180
00:08:37,581 --> 00:08:40,471
Interestingly though,
nearly every scientist I talk to

181
00:08:40,495 --> 00:08:44,376
seemed to think that gene drives were not
actually that frightening or dangerous.

182
00:08:44,400 --> 00:08:47,028
Partly because they believe
that scientists will be

183
00:08:47,052 --> 00:08:49,256
very cautious and responsible
about using them.

184
00:08:49,280 --> 00:08:50,294
(Laughter)

185
00:08:50,318 --> 00:08:51,548
So far, that's been true.

186
00:08:52,302 --> 00:08:54,998
But gene drives also have
some actual limitations.

187
00:08:55,022 --> 00:08:58,331
So for one thing, they work
only in sexually reproducing species.

188
00:08:58,704 --> 00:09:02,132
So thank goodness, they can't be used
to engineer viruses or bacteria.

189
00:09:02,156 --> 00:09:05,346
Also, the trait spreads
only with each successive generation.

190
00:09:05,370 --> 00:09:07,363
So changing or eliminating a population

191
00:09:07,387 --> 00:09:10,713
is practical only if that species
has a fast reproductive cycle,

192
00:09:10,737 --> 00:09:13,537
like insects or maybe
small vertebrates like mice or fish.

193
00:09:14,438 --> 00:09:16,699
In elephants or people,
it would take centuries

194
00:09:16,723 --> 00:09:19,103
for a trait to spread
widely enough to matter.

195
00:09:20,079 --> 00:09:25,221
Also, even with CRISPR, it's not that easy
to engineer a truly devastating trait.

196
00:09:26,079 --> 00:09:27,808
Say you wanted to make a fruit fly

197
00:09:27,832 --> 00:09:30,387
that feeds on ordinary fruit
instead of rotting fruit,

198
00:09:30,411 --> 00:09:33,044
with the aim of sabotaging
American agriculture.

199
00:09:33,068 --> 00:09:34,598
First, you'd have to figure out

200
00:09:34,622 --> 00:09:37,394
which genes control
what the fly wants to eat,

201
00:09:37,418 --> 00:09:39,965
which is already a very long
and complicated project.

202
00:09:40,489 --> 00:09:43,793
Then you'd have to alter those genes
to change the fly's behavior

203
00:09:43,817 --> 00:09:45,355
to whatever you'd want it to be,

204
00:09:45,379 --> 00:09:48,180
which is an even longer
and more complicated project.

205
00:09:48,204 --> 00:09:49,522
And it might not even work,

206
00:09:49,546 --> 00:09:52,040
because the genes
that control behavior are complex.

207
00:09:52,064 --> 00:09:54,072
So if you're a terrorist
and have to choose

208
00:09:54,096 --> 00:09:56,454
between starting a grueling
basic research program

209
00:09:56,478 --> 00:10:00,005
that will require years of meticulous
lab work and still might not pan out,

210
00:10:00,029 --> 00:10:01,376
or just blowing stuff up?

211
00:10:01,400 --> 00:10:03,055
You'll probably choose the later.

212
00:10:03,498 --> 00:10:05,918
This is especially true
because at least in theory,

213
00:10:05,942 --> 00:10:09,076
it should be pretty easy
to build what's called a reversal drive.

214
00:10:09,100 --> 00:10:12,798
That's one that basically overwrites
the change made by the first gene drive.

215
00:10:12,822 --> 00:10:15,116
So if you don't like
the effects of a change,

216
00:10:15,140 --> 00:10:17,973
you can just release a second drive
that will cancel it out,

217
00:10:17,997 --> 00:10:19,147
at least in theory.

218
00:10:21,374 --> 00:10:23,096
OK, so where does this leave us?

219
00:10:24,604 --> 00:10:28,199
We now have the ability
to change entire species at will.

220
00:10:29,017 --> 00:10:30,183
Should we?

221
00:10:30,552 --> 00:10:31,917
Are we gods now?

222
00:10:33,972 --> 00:10:35,305
I'm not sure I'd say that.

223
00:10:36,094 --> 00:10:37,321
But I would say this:

224
00:10:38,329 --> 00:10:40,187
first, some very smart people

225
00:10:40,211 --> 00:10:43,067
are even now debating
how to regulate gene drives.

226
00:10:43,598 --> 00:10:46,162
At the same time,
some other very smart people

227
00:10:46,186 --> 00:10:48,297
are working hard to create safeguards,

228
00:10:48,321 --> 00:10:51,873
like gene drives that self-regulate
or peter out after a few generations.

229
00:10:52,644 --> 00:10:53,844
That's great.

230
00:10:54,313 --> 00:10:56,860
But this technology still requires
a conversation.

231
00:10:58,059 --> 00:10:59,789
And given the nature of gene drives,

232
00:10:59,813 --> 00:11:01,503
that conversation has to be global.

233
00:11:02,131 --> 00:11:04,838
What if Kenya wants to use a drive
but Tanzania doesn't?

234
00:11:05,241 --> 00:11:08,608
Who decides whether to release
a gene drive that can fly?

235
00:11:10,874 --> 00:11:12,828
I don't have the answer to that question.

236
00:11:13,618 --> 00:11:15,626
All we can do going forward, I think,

237
00:11:15,650 --> 00:11:18,484
is talk honestly
about the risks and benefits

238
00:11:18,508 --> 00:11:20,730
and take responsibility for our choices.

239
00:11:21,817 --> 00:11:25,730
By that I mean, not just the choice
to use a gene drive,

240
00:11:25,754 --> 00:11:27,753
but also the choice not to use one.

241
00:11:29,111 --> 00:11:32,088
Humans have a tendency to assume
that the safest option

242
00:11:32,112 --> 00:11:33,904
is to preserve the status quo.

243
00:11:34,905 --> 00:11:36,477
But that's not always the case.

244
00:11:37,711 --> 00:11:41,112
Gene drives have risks,
and those need to be discussed,

245
00:11:41,136 --> 00:11:44,286
but malaria exists now
and kills 1,000 people a day.

246
00:11:44,977 --> 00:11:48,541
To combat it, we spray pesticides
that do grave damage to other species,

247
00:11:48,565 --> 00:11:50,135
including amphibians and birds.

248
00:11:51,668 --> 00:11:54,532
So when you hear about gene drives
in the coming months,

249
00:11:54,556 --> 00:11:56,810
and trust me, you will
be hearing about them,

250
00:11:56,834 --> 00:11:58,378
remember that.

251
00:11:58,402 --> 00:12:00,422
It can be frightening to act,

252
00:12:00,446 --> 00:12:02,583
but sometimes, not acting is worse.

253
00:12:04,746 --> 00:12:12,566
(Applause)