Return to Video

The next software revolution: programming biological cells

  • 0:01 - 0:05
    The second half of the last century
    was completely defined
  • 0:05 - 0:07
    by a technological revolution:
  • 0:07 - 0:09
    the software revolution.
  • 0:09 - 0:14
    The ability to program electrons
    on a material called silicon
  • 0:14 - 0:17
    made possible technologies,
    companies and industries
  • 0:17 - 0:21
    that were at one point
    unimaginable to many of us,
  • 0:21 - 0:25
    but which have now fundamentally changed
    the way the world works.
  • 0:26 - 0:28
    The first half of this century, though,
  • 0:28 - 0:32
    is going to be transformed
    by a new software revolution:
  • 0:32 - 0:35
    the living software revolution.
  • 0:35 - 0:39
    And this will be powered by the ability
    to program biochemistry
  • 0:39 - 0:41
    on a material called biology.
  • 0:41 - 0:45
    And doing so will enable us to harness
    the properties of biology
  • 0:45 - 0:48
    to generate new kinds of therapies,
  • 0:48 - 0:50
    to repair damaged tissue,
  • 0:50 - 0:53
    to reprogram faulty cells
  • 0:53 - 0:57
    or even build programmable
    operating systems out of biochemistry.
  • 0:58 - 1:02
    If we can realize this --
    and we do need to realize it --
  • 1:02 - 1:04
    its impact will be so enormous
  • 1:04 - 1:08
    that it will make the first
    software revolution pale in comparison.
  • 1:08 - 1:12
    And that's because living software
    would transform the entirety of medicine,
  • 1:12 - 1:14
    agriculture and energy,
  • 1:14 - 1:18
    and these are sectors that dwarf
    those dominated by IT.
  • 1:19 - 1:23
    Imagine programmable plants
    that fix nitrogen more effectively
  • 1:23 - 1:26
    or resist emerging fungal pathogens,
  • 1:26 - 1:29
    or even programming crops
    to be perennial rather than annual
  • 1:30 - 1:32
    so you could double
    your crop yields each year.
  • 1:32 - 1:34
    That would transform agriculture
  • 1:34 - 1:38
    and how we'll keep our growing
    and global population fed.
  • 1:39 - 1:41
    Or imagine programmable immunity,
  • 1:41 - 1:45
    designing and harnessing molecular devices
    that guide your immune system
  • 1:45 - 1:49
    to detect, eradicate
    or even prevent disease.
  • 1:49 - 1:51
    This would transform medicine
  • 1:51 - 1:54
    and how we'll keep our growing
    and aging population healthy.
  • 1:56 - 2:00
    We already have many of the tools
    that will make living software a reality.
  • 2:00 - 2:02
    We can precisely edit genes with CRISPR.
  • 2:02 - 2:05
    We can rewrite the genetic code
    one base at a time.
  • 2:05 - 2:10
    We can even build functioning
    synthetic circuits out of DNA.
  • 2:10 - 2:13
    But figuring out how and when
    to wield these tools
  • 2:13 - 2:15
    is still a process of trial and error.
  • 2:15 - 2:19
    It needs deep expertise,
    years of specialization.
  • 2:19 - 2:22
    And experimental protocols
    are difficult to discover
  • 2:22 - 2:25
    and all too often, difficult to reproduce.
  • 2:25 - 2:30
    And, you know, we have a tendency
    in biology to focus a lot on the parts,
  • 2:30 - 2:33
    but we all know that something like flying
    wouldn't be understood
  • 2:33 - 2:34
    by only studying feathers.
  • 2:35 - 2:39
    So programming biology is not yet
    as simple as programming your computer.
  • 2:39 - 2:41
    And then to make matters worse,
  • 2:41 - 2:45
    living systems largely bear no resemblance
    to the engineered systems
  • 2:45 - 2:47
    that you and I program every day.
  • 2:48 - 2:52
    In contrast to engineered systems,
    living systems self-generate,
  • 2:52 - 2:53
    they self-organize,
  • 2:53 - 2:55
    they operate at molecular scales.
  • 2:55 - 2:57
    And these molecular-level interactions
  • 2:57 - 3:00
    lead generally to robust
    macro-scale output.
  • 3:00 - 3:03
    They can even self-repair.
  • 3:04 - 3:07
    Consider, for example,
    the humble household plant,
  • 3:07 - 3:09
    like that one sat
    on your mantelpiece at home
  • 3:09 - 3:11
    that you keep forgetting to water.
  • 3:12 - 3:15
    Every day, despite your neglect,
    that plant has to wake up
  • 3:15 - 3:18
    and figure out how
    to allocate its resources.
  • 3:18 - 3:22
    Will it grow, photosynthesize,
    produce seeds, or flower?
  • 3:22 - 3:26
    And that's a decision that has to be made
    at the level of the whole organism.
  • 3:26 - 3:29
    But a plant doesn't have a brain
    to figure all of that out.
  • 3:29 - 3:32
    It has to make do
    with the cells on its leaves.
  • 3:32 - 3:34
    They have to respond to the environment
  • 3:34 - 3:37
    and make the decisions
    that affect the whole plant.
  • 3:37 - 3:41
    So somehow there must be a program
    running inside these cells,
  • 3:41 - 3:43
    a program that responds
    to input signals and cues
  • 3:43 - 3:45
    and shapes what that cell will do.
  • 3:46 - 3:49
    And then those programs must operate
    in a distributed way
  • 3:49 - 3:50
    across individual cells,
  • 3:50 - 3:54
    so that they can coordinate
    and that plant can grow and flourish.
  • 3:56 - 3:59
    If we could understand
    these biological programs,
  • 3:59 - 4:02
    if we could understand
    biological computation,
  • 4:02 - 4:06
    it would transform our ability
    to understand how and why
  • 4:06 - 4:08
    cells do what they do.
  • 4:08 - 4:10
    Because, if we understood these programs,
  • 4:10 - 4:12
    we could debug them when things go wrong.
  • 4:12 - 4:17
    Or we could learn from them how to design
    the kind of synthetic circuits
  • 4:17 - 4:21
    that truly exploit
    the computational power of biochemistry.
  • 4:22 - 4:25
    My passion about this idea
    led me to a career in research
  • 4:25 - 4:29
    at the interface of maths,
    computer science and biology.
  • 4:29 - 4:34
    And in my work, I focus on the concept
    of biology as computation.
  • 4:34 - 4:37
    And that means asking
    what do cells compute,
  • 4:38 - 4:41
    and how can we uncover
    these biological programs?
  • 4:42 - 4:46
    And I started to ask these questions
    together with some brilliant collaborators
  • 4:46 - 4:48
    at Microsoft Research
    and the University of Cambridge,
  • 4:48 - 4:50
    where together we wanted to understand
  • 4:50 - 4:55
    the biological program
    running inside a unique type of cell:
  • 4:55 - 4:57
    an embryonic stem cell.
  • 4:57 - 5:00
    These cells are unique
    because they're totally naïve.
  • 5:00 - 5:02
    They can become anything they want:
  • 5:03 - 5:05
    a brain cell, a heart cell,
    a bone cell, a lung cell,
  • 5:05 - 5:07
    any adult cell type.
  • 5:07 - 5:09
    This naïvety, it sets them apart,
  • 5:09 - 5:12
    but it also ignited the imagination
    of the scientific community,
  • 5:12 - 5:15
    who realized, if we could
    tap into that potential,
  • 5:15 - 5:17
    we would have a powerful
    tool for medicine.
  • 5:18 - 5:21
    If we could figure out
    how these cells make the decision
  • 5:21 - 5:23
    to become one cell type or another,
  • 5:23 - 5:24
    we might be able to harness them
  • 5:24 - 5:29
    to generate cells that we need
    to repair diseased or damaged tissue.
  • 5:30 - 5:33
    But realizing that vision
    is not without its challenges,
  • 5:33 - 5:36
    not least because these particular cells,
  • 5:36 - 5:38
    they emerge just six days
    after conception.
  • 5:39 - 5:41
    And then within a day or so, they're gone.
  • 5:41 - 5:43
    They have set off down the different paths
  • 5:43 - 5:46
    that form all the structures
    and organs of your adult body.
  • 5:48 - 5:51
    But it turns out that cell fates
    are a lot more plastic
  • 5:51 - 5:52
    than we might have imagined.
  • 5:52 - 5:57
    About 13 years ago, some scientists
    showed something truly revolutionary.
  • 5:57 - 6:02
    By inserting just a handful of genes
    into an adult cell,
  • 6:02 - 6:04
    like one of your skin cells,
  • 6:04 - 6:08
    you can transform that cell
    back to the naïve state.
  • 6:08 - 6:11
    And it's a process that's actually
    known as "reprogramming,"
  • 6:11 - 6:14
    and it allows us to imagine
    a kind of stem cell utopia,
  • 6:14 - 6:18
    the ability to take a sample
    of a patient's own cells,
  • 6:18 - 6:20
    transform them back to the naïve state
  • 6:20 - 6:23
    and use those cells to make
    whatever that patient might need,
  • 6:23 - 6:25
    whether it's brain cells or heart cells.
  • 6:27 - 6:28
    But over the last decade or so,
  • 6:28 - 6:31
    figuring out how to change cell fate,
  • 6:31 - 6:34
    it's still a process of trial and error.
  • 6:34 - 6:38
    Even in cases where we've uncovered
    successful experimental protocols,
  • 6:38 - 6:40
    they're still inefficient,
  • 6:40 - 6:44
    and we lack a fundamental understanding
    of how and why they work.
  • 6:45 - 6:48
    If you figured out how to change
    a stem cell into a heart cell,
  • 6:48 - 6:51
    that hasn't got any way of telling you
    how to change a stem cell
  • 6:51 - 6:52
    into a brain cell.
  • 6:53 - 6:56
    So we wanted to understand
    the biological program
  • 6:56 - 6:58
    running inside an embryonic stem cell,
  • 6:58 - 7:02
    and understanding the computation
    performed by a living system
  • 7:02 - 7:06
    starts with asking
    a devastatingly simple question:
  • 7:06 - 7:09
    What is it that system actually has to do?
  • 7:10 - 7:13
    Now, computer science actually
    has a set of strategies
  • 7:13 - 7:17
    for dealing with what it is the software
    and hardware are meant to do.
  • 7:17 - 7:19
    When you write a program,
    you code a piece of software,
  • 7:19 - 7:21
    you want that software to run correctly.
  • 7:21 - 7:23
    You want performance, functionality.
  • 7:23 - 7:24
    You want to prevent bugs.
  • 7:24 - 7:26
    They can cost you a lot.
  • 7:26 - 7:28
    So when a developer writes a program,
  • 7:28 - 7:30
    they could write down
    a set of specifications.
  • 7:30 - 7:32
    These are what your program should do.
  • 7:32 - 7:34
    Maybe it should compare
    the size of two numbers
  • 7:35 - 7:36
    or order numbers by increasing size.
  • 7:37 - 7:42
    Technology exists that allows us
    automatically to check
  • 7:42 - 7:44
    whether our specifications are satisfied,
  • 7:44 - 7:47
    whether that program
    does what it should do.
  • 7:47 - 7:50
    And so our idea was that in the same way,
  • 7:50 - 7:53
    experimental observations,
    things we measure in the lab,
  • 7:53 - 7:58
    they correspond to specifications
    of what the biological program should do.
  • 7:59 - 8:01
    So we just needed to figure out a way
  • 8:01 - 8:04
    to encode this new type of specification.
  • 8:05 - 8:08
    So let's say you've been busy in the lab
    and you've been measuring your genes
  • 8:08 - 8:11
    and you've found that if Gene A is active,
  • 8:11 - 8:14
    then Gene B or Gene C seems to be active.
  • 8:15 - 8:18
    We can write that observation down
    as a mathematical expression
  • 8:18 - 8:21
    if we can use the language of logic:
  • 8:21 - 8:23
    If A, then B or C.
  • 8:24 - 8:27
    Now, this is a very simple example, OK.
  • 8:27 - 8:28
    It's just to illustrate the point.
  • 8:28 - 8:31
    We can encode truly rich expressions
  • 8:31 - 8:36
    that actually capture the behavior
    of multiple genes or proteins over time
  • 8:36 - 8:38
    across multiple different experiments.
  • 8:39 - 8:41
    And so by translating our observations
  • 8:41 - 8:43
    into mathematical expression in this way,
  • 8:43 - 8:48
    it becomes possible to test whether
    or not those observations can emerge
  • 8:48 - 8:51
    from a program of genetic interactions.
  • 8:52 - 8:55
    And we developed a tool to do just this.
  • 8:55 - 8:58
    We were able to use this tool
    to encode observations
  • 8:58 - 8:59
    as mathematical expressions,
  • 8:59 - 9:03
    and then that tool would allow us
    to uncover the genetic program
  • 9:03 - 9:04
    that could explain them all.
  • 9:05 - 9:08
    And we then apply this approach
  • 9:08 - 9:12
    to uncover the genetic program
    running inside embryonic stem cells
  • 9:12 - 9:16
    to see if we could understand
    how to induce that naïve state.
  • 9:16 - 9:18
    And this tool was actually built
  • 9:18 - 9:21
    on a solver that's deployed
    routinely around the world
  • 9:21 - 9:23
    for conventional software verification.
  • 9:24 - 9:27
    So we started with a set
    of nearly 50 different specifications
  • 9:27 - 9:32
    that we generated from experimental
    observations of embryonic stem cells.
  • 9:32 - 9:35
    And by encoding these
    observations in this tool,
  • 9:35 - 9:38
    we were able to uncover
    the first molecular program
  • 9:38 - 9:40
    that could explain all of them.
  • 9:40 - 9:43
    Now, that's kind of a feat
    in and of itself, right?
  • 9:43 - 9:46
    Being able to reconcile
    all of these different observations
  • 9:46 - 9:49
    is not the kind of thing
    you can do on the back of an envelope,
  • 9:49 - 9:52
    even if you have a really big envelope.
  • 9:52 - 9:54
    Because we've got
    this kind of understanding,
  • 9:54 - 9:56
    we could go one step further.
  • 9:56 - 9:59
    We could use this program to predict
    what this cell might do
  • 9:59 - 10:01
    in conditions we hadn't yet tested.
  • 10:01 - 10:04
    We could probe the program in silico.
  • 10:05 - 10:06
    And so we did just that:
  • 10:06 - 10:09
    we generated predictions
    that we tested in the lab,
  • 10:09 - 10:12
    and we found that this program
    was highly predictive.
  • 10:12 - 10:15
    It told us how we could
    accelerate progress
  • 10:15 - 10:18
    back to the naïve state
    quickly and efficiently.
  • 10:18 - 10:21
    It told us which genes
    to target to do that,
  • 10:21 - 10:23
    which genes might even
    hinder that process.
  • 10:23 - 10:28
    We even found the program predicted
    the order in which genes would switch on.
  • 10:29 - 10:32
    So this approach really allowed us
    to uncover the dynamics
  • 10:32 - 10:35
    of what the cells are doing.
  • 10:36 - 10:39
    What we've developed, it's not a method
    that's specific to stem cell biology.
  • 10:39 - 10:42
    Rather, it allows us to make sense
    of the computation
  • 10:42 - 10:44
    being carried out by the cell
  • 10:44 - 10:47
    in the context of genetic interactions.
  • 10:47 - 10:49
    So really, it's just one building block.
  • 10:49 - 10:52
    The field urgently needs
    to develop new approaches
  • 10:52 - 10:54
    to understand biological
    computation more broadly
  • 10:54 - 10:56
    and at different levels,
  • 10:56 - 11:00
    from DNA right through
    to the flow of information between cells.
  • 11:00 - 11:03
    Only this kind of
    transformative understanding
  • 11:03 - 11:08
    will enable us to harness biology
    in ways that are predictable and reliable.
  • 11:09 - 11:12
    But to program biology,
    we will also need to develop
  • 11:12 - 11:14
    the kinds of tools and languages
  • 11:14 - 11:18
    that allow both experimentalists
    and computational scientists
  • 11:18 - 11:20
    to design biological function
  • 11:20 - 11:24
    and have those designs compile down
    to the machine code of the cell,
  • 11:24 - 11:25
    its biochemistry,
  • 11:25 - 11:27
    so that we could then
    build those structures.
  • 11:27 - 11:31
    Now, that's something akin
    to a living software compiler,
  • 11:31 - 11:33
    and I'm proud to be
    part of a team at Microsoft
  • 11:33 - 11:35
    that's working to develop one.
  • 11:35 - 11:39
    Though to say it's a grand challenge
    is kind of an understatement,
  • 11:39 - 11:40
    but if it's realized,
  • 11:40 - 11:44
    it would be the final bridge
    between software and wetware.
  • 11:45 - 11:48
    More broadly, though, programming biology
    is only going to be possible
  • 11:48 - 11:53
    if we can transform the field
    into being truly interdisciplinary.
  • 11:53 - 11:56
    It needs us to bridge
    the physical and the life sciences,
  • 11:56 - 11:58
    and scientists from
    each of these disciplines
  • 11:58 - 12:01
    need to be able to work together
    with common languages
  • 12:01 - 12:03
    and to have shared scientific questions.
  • 12:05 - 12:09
    In the long term, it's worth remembering
    that many of the giant software companies
  • 12:09 - 12:11
    and the technology
    that you and I work with every day
  • 12:11 - 12:13
    could hardly have been imagined
  • 12:13 - 12:16
    at the time we first started
    programming on silicon microchips.
  • 12:16 - 12:19
    And if we start now to think about
    the potential for technology
  • 12:20 - 12:22
    enabled by computational biology,
  • 12:22 - 12:25
    we'll see some of the steps
    that we need to take along the way
  • 12:25 - 12:26
    to make that a reality.
  • 12:27 - 12:30
    Now, there is the sobering thought
    that this kind of technology
  • 12:30 - 12:32
    could be open to misuse.
  • 12:32 - 12:34
    If we're willing to talk
    about the potential
  • 12:34 - 12:36
    for programming immune cells,
  • 12:36 - 12:39
    we should also be thinking
    about the potential of bacteria
  • 12:39 - 12:41
    engineered to evade them.
  • 12:41 - 12:43
    There might be people willing to do that.
  • 12:44 - 12:45
    Now, one reassuring thought in this
  • 12:45 - 12:48
    is that -- well, less so
    for the scientists --
  • 12:48 - 12:51
    is that biology is
    a fragile thing to work with.
  • 12:51 - 12:53
    So programming biology
    is not going to be something
  • 12:53 - 12:55
    you'll be doing in your garden shed.
  • 12:56 - 12:58
    But because we're at the outset of this,
  • 12:58 - 13:00
    we can move forward
    with our eyes wide open.
  • 13:00 - 13:03
    We can ask the difficult
    questions up front,
  • 13:03 - 13:06
    we can put in place
    the necessary safeguards
  • 13:06 - 13:09
    and, as part of that,
    we'll have to think about our ethics.
  • 13:09 - 13:12
    We'll have to think about putting bounds
    on the implementation
  • 13:12 - 13:13
    of biological function.
  • 13:14 - 13:17
    So as part of this, research in bioethics
    will have to be a priority.
  • 13:17 - 13:20
    It can't be relegated to second place
  • 13:20 - 13:22
    in the excitement
    of scientific innovation.
  • 13:23 - 13:27
    But the ultimate prize,
    the ultimate destination on this journey,
  • 13:27 - 13:30
    would be breakthrough applications
    and breakthrough industries
  • 13:30 - 13:34
    in areas from agriculture and medicine
    to energy and materials
  • 13:34 - 13:36
    and even computing itself.
  • 13:36 - 13:40
    Imagine, one day we could be powering
    the planet sustainably
  • 13:40 - 13:42
    on the ultimate green energy
  • 13:42 - 13:45
    if we could mimic something
    that plants figured out millennia ago:
  • 13:46 - 13:49
    how to harness the sun's energy
    with an efficiency that is unparalleled
  • 13:49 - 13:51
    by our current solar cells.
  • 13:52 - 13:54
    If we understood that program
    of quantum interactions
  • 13:54 - 13:58
    that allow plants to absorb
    sunlight so efficiently,
  • 13:58 - 14:02
    we might be able to translate that
    into building synthetic DNA circuits
  • 14:02 - 14:04
    that offer the material
    for better solar cells.
  • 14:05 - 14:09
    There are teams and scientists working
    on the fundamentals of this right now,
  • 14:09 - 14:12
    so perhaps if it got the right attention
    and the right investment,
  • 14:12 - 14:15
    it could be realized in 10 or 15 years.
  • 14:15 - 14:19
    So we are at the beginning
    of a technological revolution.
  • 14:19 - 14:22
    Understanding this ancient type
    of biological computation
  • 14:22 - 14:24
    is the critical first step.
  • 14:24 - 14:26
    And if we can realize this,
  • 14:26 - 14:29
    we would enter in the era
    of an operating system
  • 14:29 - 14:31
    that runs living software.
  • 14:31 - 14:32
    Thank you very much.
  • 14:32 - 14:34
    (Applause)
Title:
The next software revolution: programming biological cells
Speaker:
Sara-Jane Dunn
Description:

The cells in your body are like computer software: they're "programmed" to carry out specific functions at specific times. If we can better understand this process, we could unlock the ability to reprogram cells ourselves, says computational biologist Sara-Jane Dunn. In a talk from the cutting-edge of science, she explains how her team is studying embryonic stem cells to gain a new understanding of the biological programs that power life -- and develop "living software" that could transform medicine, agriculture and energy.

more » « less
Video Language:
English
Team:
TED
Project:
TEDTalks
Duration:
14:47

English subtitles

Revisions Compare revisions