1
00:00:00,843 --> 00:00:02,888
I'd like to show you a video of some of the models

2
00:00:02,888 --> 00:00:04,477
I work with.

3
00:00:04,477 --> 00:00:08,015
They're all the perfect size, and they don't have an ounce of fat.

4
00:00:08,015 --> 00:00:10,553
Did I mention they're gorgeous?

5
00:00:10,553 --> 00:00:13,683
And they're scientific models? (Laughs)

6
00:00:13,683 --> 00:00:16,026
As you might have guessed, I'm a tissue engineer,

7
00:00:16,026 --> 00:00:18,475
and this is a video of some of the beating heart

8
00:00:18,475 --> 00:00:20,691
that I've engineered in the lab.

9
00:00:20,691 --> 00:00:22,573
And one day we hope that these tissues

10
00:00:22,573 --> 00:00:25,517
can serve as replacement parts for the human body.

11
00:00:25,517 --> 00:00:27,797
But what I'm going to tell you about today

12
00:00:27,797 --> 00:00:32,244
is how these tissues make awesome models.

13
00:00:32,244 --> 00:00:34,971
Well, let's think about the drug screening process for a moment.

14
00:00:34,971 --> 00:00:37,949
You go from drug formulation, lab testing, animal testing,

15
00:00:37,949 --> 00:00:40,452
and then clinical trials, which you might call human testing,

16
00:00:40,452 --> 00:00:42,717
before the drugs get to market.

17
00:00:42,717 --> 00:00:45,860
It costs a lot of money, a lot of time,

18
00:00:45,860 --> 00:00:48,670
and sometimes, even when a drug hits the market,

19
00:00:48,670 --> 00:00:52,605
it acts in an unpredictable way and actually hurts people.

20
00:00:52,605 --> 00:00:56,692
And the later it fails, the worse the consequences.

21
00:00:56,692 --> 00:01:00,876
It all boils down to two issues. One, humans are not rats,

22
00:01:00,876 --> 00:01:04,964
and two, despite our incredible similarities to one another,

23
00:01:04,964 --> 00:01:07,405
actually those tiny differences between you and I

24
00:01:07,405 --> 00:01:09,914
have huge impacts with how we metabolize drugs

25
00:01:09,914 --> 00:01:11,783
and how those drugs affect us.

26
00:01:11,783 --> 00:01:14,615
So what if we had better models in the lab

27
00:01:14,615 --> 00:01:17,885
that could not only mimic us better than rats

28
00:01:17,885 --> 00:01:21,805
but also reflect our diversity?

29
00:01:21,805 --> 00:01:25,732
Let's see how we can do it with tissue engineering.

30
00:01:25,732 --> 00:01:28,261
One of the key technologies that's really important

31
00:01:28,261 --> 00:01:31,453
is what's called induced pluripotent stem cells.

32
00:01:31,453 --> 00:01:33,971
They were developed in Japan pretty recently.

33
00:01:33,971 --> 00:01:36,418
Okay, induced pluripotent stem cells.

34
00:01:36,418 --> 00:01:38,531
They're a lot like embryonic stem cells

35
00:01:38,531 --> 00:01:40,748
except without the controversy.

36
00:01:40,748 --> 00:01:43,647
We induce cells, okay, say, skin cells,

37
00:01:43,647 --> 00:01:46,154
by adding a few genes to them, culturing them,

38
00:01:46,154 --> 00:01:47,775
and then harvesting them.

39
00:01:47,775 --> 00:01:50,482
So they're skin cells that can be tricked,

40
00:01:50,482 --> 00:01:53,266
kind of like cellular amnesia, into an embryonic state.

41
00:01:53,266 --> 00:01:55,978
So without the controversy, that's cool thing number one.

42
00:01:55,978 --> 00:01:58,527
Cool thing number two, you can grow any type of tissue

43
00:01:58,527 --> 00:02:01,082
out of them: brain, heart, liver, you get the picture,

44
00:02:01,082 --> 00:02:03,605
but out of your cells.

45
00:02:03,605 --> 00:02:07,170
So we can make a model of your heart, your brain

46
00:02:07,170 --> 00:02:09,802
on a chip.

47
00:02:09,802 --> 00:02:12,658
Generating tissues of predictable density and behavior

48
00:02:12,658 --> 00:02:15,490
is the second piece, and will be really key towards

49
00:02:15,490 --> 00:02:18,162
getting these models to be adopted for drug discovery.

50
00:02:18,162 --> 00:02:21,274
And this is a schematic of a bioreactor we're developing in our lab

51
00:02:21,274 --> 00:02:24,722
to help engineer tissues in a more modular, scalable way.

52
00:02:24,722 --> 00:02:28,121
Going forward, imagine a massively parallel version of this

53
00:02:28,121 --> 00:02:30,458
with thousands of pieces of human tissue.

54
00:02:30,458 --> 00:02:34,506
It would be like having a clinical trial on a chip.

55
00:02:34,506 --> 00:02:38,301
But another thing about these induced pluripotent stem cells

56
00:02:38,301 --> 00:02:40,850
is that if we take some skin cells, let's say,

57
00:02:40,850 --> 00:02:43,026
from people with a genetic disease

58
00:02:43,026 --> 00:02:45,282
and we engineer tissues out of them,

59
00:02:45,282 --> 00:02:47,250
we can actually use tissue-engineering techniques

60
00:02:47,250 --> 00:02:50,651
to generate models of those diseases in the lab.

61
00:02:50,651 --> 00:02:54,235
Here's an example from Kevin Eggan's lab at Harvard.

62
00:02:54,235 --> 00:02:56,525
He generated neurons

63
00:02:56,525 --> 00:02:59,240
from these induced pluripotent stem cells

64
00:02:59,240 --> 00:03:01,869
from patients who have Lou Gehrig's Disease,

65
00:03:01,869 --> 00:03:04,312
and he differentiated them into neurons, and what's amazing

66
00:03:04,312 --> 00:03:07,464
is that these neurons also show symptoms of the disease.

67
00:03:07,464 --> 00:03:09,563
So with disease models like these, we can fight back

68
00:03:09,563 --> 00:03:12,145
faster than ever before and understand the disease better

69
00:03:12,145 --> 00:03:16,108
than ever before, and maybe discover drugs even faster.

70
00:03:16,108 --> 00:03:19,488
This is another example of patient-specific stem cells

71
00:03:19,488 --> 00:03:23,497
that were engineered from someone with retinitis pigmentosa.

72
00:03:23,497 --> 00:03:25,251
This is a degeneration of the retina.

73
00:03:25,251 --> 00:03:28,008
It's a disease that runs in my family, and we really hope

74
00:03:28,008 --> 00:03:30,232
that cells like these will help us find a cure.

75
00:03:30,232 --> 00:03:33,040
So some people think that these models sound well and good,

76
00:03:33,040 --> 00:03:36,481
but ask, "Well, are these really as good as the rat?"

77
00:03:36,481 --> 00:03:39,469
The rat is an entire organism, after all,

78
00:03:39,469 --> 00:03:41,175
with interacting networks of organs.

79
00:03:41,175 --> 00:03:45,096
A drug for the heart can get metabolized in the liver,

80
00:03:45,096 --> 00:03:47,936
and some of the byproducts may be stored in the fat.

81
00:03:47,936 --> 00:03:52,463
Don't you miss all that with these tissue-engineered models?

82
00:03:52,463 --> 00:03:54,577
Well, this is another trend in the field.

83
00:03:54,577 --> 00:03:57,444
By combining tissue engineering techniques with microfluidics,

84
00:03:57,444 --> 00:03:59,608
the field is actually evolving towards just that,

85
00:03:59,608 --> 00:04:02,114
a model of the entire ecosystem of the body,

86
00:04:02,114 --> 00:04:04,514
complete with multiple organ systems to be able to test

87
00:04:04,514 --> 00:04:06,117
how a drug you might take for your blood pressure

88
00:04:06,117 --> 00:04:09,384
might affect your liver or an antidepressant might affect your heart.

89
00:04:09,384 --> 00:04:13,456
These systems are really hard to build, but we're just starting to be able to get there,

90
00:04:13,456 --> 00:04:16,760
and so, watch out.

91
00:04:16,760 --> 00:04:19,392
But that's not even all of it, because once a drug is approved,

92
00:04:19,392 --> 00:04:23,074
tissue engineering techniques can actually help us develop more personalized treatments.

93
00:04:23,074 --> 00:04:26,816
This is an example that you might care about someday,

94
00:04:26,816 --> 00:04:28,936
and I hope you never do,

95
00:04:28,936 --> 00:04:31,456
because imagine if you ever get that call

96
00:04:31,456 --> 00:04:34,664
that gives you that bad news that you might have cancer.

97
00:04:34,664 --> 00:04:37,200
Wouldn't you rather test to see if those cancer drugs

98
00:04:37,200 --> 00:04:39,960
you're going to take are going to work on your cancer?

99
00:04:39,960 --> 00:04:42,382
This is an example from Karen Burg's lab, where they're

100
00:04:42,382 --> 00:04:45,288
using inkjet technologies to print breast cancer cells

101
00:04:45,288 --> 00:04:47,759
and study its progressions and treatments.

102
00:04:47,759 --> 00:04:50,312
And some of our colleagues at Tufts are mixing models

103
00:04:50,312 --> 00:04:53,400
like these with tissue-engineered bone to see how cancer

104
00:04:53,400 --> 00:04:56,120
might spread from one part of the body to the next,

105
00:04:56,120 --> 00:04:58,504
and you can imagine those kinds of multi-tissue chips

106
00:04:58,504 --> 00:05:01,489
to be the next generation of these kinds of studies.

107
00:05:01,489 --> 00:05:03,911
And so thinking about the models that we've just discussed,

108
00:05:03,911 --> 00:05:05,824
you can see, going forward, that tissue engineering

109
00:05:05,824 --> 00:05:08,280
is actually poised to help revolutionize drug screening

110
00:05:08,280 --> 00:05:11,058
at every single step of the path:

111
00:05:11,058 --> 00:05:13,632
disease models making for better drug formulations,

112
00:05:13,632 --> 00:05:17,503
massively parallel human tissue models helping to revolutionize lab testing,

113
00:05:17,503 --> 00:05:21,728
reduce animal testing and human testing in clinical trials,

114
00:05:21,728 --> 00:05:23,420
and individualized therapies that disrupt

115
00:05:23,420 --> 00:05:27,008
what we even consider to be a market at all.

116
00:05:27,008 --> 00:05:29,552
Essentially, we're dramatically speeding up that feedback

117
00:05:29,552 --> 00:05:31,875
between developing a molecule and learning about

118
00:05:31,875 --> 00:05:34,224
how it acts in the human body.

119
00:05:34,224 --> 00:05:36,552
Our process for doing this is essentially transforming

120
00:05:36,552 --> 00:05:41,413
biotechnology and pharmacology into an information technology,

121
00:05:41,413 --> 00:05:44,392
helping us discover and evaluate drugs faster,

122
00:05:44,392 --> 00:05:47,608
more cheaply and more effectively.

123
00:05:47,608 --> 00:05:51,688
It gives new meaning to models against animal testing, doesn't it?

124
00:05:51,688 --> 00:05:58,503
Thank you. (Applause)