MIT Tech Day 2005: Bioengineering – Building Bridges Between the Sciences, Engineering & Medicine

MIT Tech Day 2005: Bioengineering – Building Bridges Between the Sciences, Engineering & Medicine


[MUSIC PLAYING] VINCENT: Folks, my name is Doug
Vincent, chair of the Tech Day committee. I’m really excited today. It’s a beautiful day outside. There’s not a lot of things
that would draw me inside today. But the program we have for you
today is one of those things. We have had so much fun
putting this program together, working with these five
outstanding faculty members. Couldn’t be happier than to
share that with you all today. We have a little bit
different format this year. We’ve delayed lunch
till after 12:45, so our five speakers
will all be before lunch. We’ll hear from
three of them, have a break, including
some fruit and granola, then rejoin here
for the last two. And at the end, we’ll have
some questions and answers. President Hockfield
will moderate that. We have a live feed
downstairs, to Little Kresge, and we’ll go through some more
of the details for questions and answers a little bit
later in this program. The goal of that is
so this afternoon, folks are free to
explore the campus, participate in other class
activities, and so forth. So bear with us for a little
bit lengthened morning program. We think you will find
it worth the while. I need to say thank you
to the Alumni Association for the faith that they
continue to have in the Tech Day committee. Thank you to the committee. Outstanding year. And of course, thank
you to Lou Alexander. Outstanding. Just a pleasure
to work with you. If folks would reach to
wherever they have their cell phones right now and find the
button on it that says off and turn that off,
I appreciate that. Thank you. Just like to introduce
now Beth Garvin, Executive Vice President and CEO
of the MIT Alumni Association. Beth? [APPLAUSE] GARVIN: Good morning. It’s a pleasure to see
so many familiar faces, and even more exciting to
see so many unfamiliar faces. I hope I get a chance to
say hello to each of you personally and welcome
you back to MIT. Doug and I were just chatting
before the program started and I said, don’t give me
much of an introduction, because nobody wants to
hear from you or me, Doug. They’re here to hear
from the president and to hear from these
marvelous faculty. He noted, however,
that, yeah, we’re just the glue that
holds it together. And I’d like to acknowledge
that glue in the Technology Day committee and all
the reunion committee volunteers, the reunion gift
committee volunteers, reunion row volunteers, tech
challenge game volunteers, all the hundreds of
people who have helped make Tech Reunions
2005 a fabulous success from all accounts. It’s been my joy this year
to be able to introduce our 16th president, Susan
Hockfield, to all of you, and to introduce
the alumni to her. It has been just a remarkable
journey as we first announced Susan’s naming. And any of you with
an email address that we know have heard from
me multiple times about how I think this has been such
a wonderful choice for MIT and what a wonderful
fit Dr. Hockfield is for MIT at this point in time. So without any further
ado, all of the background is in your program book. And I’m sure you’ve
already read of Susan’s remarkable accomplishments, so
I do nothing more than turn it over to President Hockfield. [APPLAUSE] HOCKFIELD: Thank you. Thank you. It’s a great pleasure
to be here this morning. Thank you, Beth. It’s been a real treat for me
to travel the country with you and to have you by my
side introducing me to alumni from coast to coast. One of the things that has
impressed me enormously about MIT is the
kind of energy that just emanates from the place. You walk down the
corridors and the place is just bursting with energy. But that energy is carried
inside the people of the place. And so as I travel
the country with Beth and her band of just
tireless troopers, having them by my side and
having them introduce me to many of you over the
last, what is it now, almost six months, has just
been an incredible delight and a treat. The MIT alumni are, indeed,
a very special bunch, special in so many ways, special
that you could have finally brought spring to Cambridge. We thank you for that. It’s been a grim
spring, but this weekend is the first real
burst of sunshine. It makes us feel
optimistic that we actually might have a summer here. It’s just been wonderful
to sample the treats of this reunion weekend. My only regret– people
have asked me what I regret. What are the things I
don’t like about my job? And my response is
always the thing I don’t like is only
having 24 hours a day. And that’s the way I feel
about these several days. I mean, I wish I could live
every hour of these days three times over so that I could
be at all of the events and meet more of you than I’m
going to unfortunately have a chance to meet this weekend. I’ve met many of
you, and I will be meeting several more of
you, many more of you, over the next several hours
through the end of this reunion weekend. But I promise you in the
months and years to come, I will be getting around
to meeting all of you. So it’s great to be here. This morning’s
program is wonderful. I can take absolutely
no credit for it. There are many happy
coincidences that I’ve seen over the last six months. And certainly the topic
for this morning’s program is a very happy coincidence. Those of you who read or heard
my inauguration address know that one of the things I
drew out– among the very, very many exciting opportunities
and challenges for MIT going forward– is this fantastic convergence
between the life sciences and engineering. And this morning’s
program exemplifies that in just startling ways. In addition, the content of
this program that you will hear exemplifies the intellectual
rigor and the work at the frontiers of
knowledge that really do define MIT as an institution. I know that the
devotion of alumni is one of our absolutely
greatest strengths. And keeping MIT
strong and vibrant is a team effort that involves
all of you as well as all of us on campus. But this great
community of MIT that reaches across the
country, across the world, is perfectly in line with
the Institute’s fundamentally collaborative spirit. At MIT, something
that I’ve learned is that we have a gift for
learning from one another. We excel at the kinds of
intellectual interaction that push conventional
boundaries and establish new
fields of study, new industries, and new ways
of thinking about the world. The collaborations
that MIT fosters depend on our remarkable
culture of openness. I’ve been struck by how the very
architecture of the main group symbolizes this. The main group was imagined
and then constructed, imagined by Bosworth, as
one huge but interconnected building with corridors
that link the entire expanse of these enormous buildings. Our historic campus embodies
the vision of an institution without internal
boundaries, a place where people who are interested
in similar problems can work together freely
across what might otherwise be disciplinary divides. In its design, the main group
has actually helped invent MIT and helped MIT
invent the future. In the years ahead,
we will continue to capitalize on our
spirit of openness to create new partnerships
across our own schools and departments, and
with other institutions in the public and
private sectors. Collaboration is
absolutely essential if we are to fulfill MIT’s
mission of bringing knowledge to bear on the world’s
great problems. The challenges facing us
today from climate change to the future of Social
Security to the startling spread of contagious diseases are
inherently multidisciplinary. Now, MITs distinct
ability to catalyze work at the interfaces of
existing disciplines also depends on a single
uniform standard of excellence. Over the course
of this year, I’ve found an uncompromising
dedication to excellence in all of our
departments and schools. We see it in the
faculty, in the students, and also in the staff
around the Institute. I think this morning’s program
will demonstrate what I mean. Today’s faculty presenters are
absolutely stellar researchers. But they’re also stellar
teachers and members, devoted members, generous
members of the MIT community. And they will be
talking about some of our era’s most
exciting advances in science and technology. This generation
is bearing witness to a fascinating convergence
of engineering and the life sciences. This convergence
holds the promise of transforming our lives. But this kind of convergence
is not unprecedented. It’s a convergence that
we saw not so long ago, and we know this precedent. 70 years ago,
president Karl Compton insisted that the
physical sciences must play a critical role in
education and research at MIT. And the result of bringing
in the physical sciences into conversation
with engineering produced nothing less than
a new era for engineering which MIT pioneered. To appreciate the power of
the kinds of collaborations that Compton helped spawn
through the 20th century’s big convergence, you only have
to think about the radiation lab that helped develop radar
and helped us win World War II. Today engineering is
making the same kind of fertile connections
with the life sciences. And I believe we can expect
equally revolutionary results. Combining MIT’s historic
strength and engineering and our newer strengths
in biology in the brain and cognitive
sciences, we’re already opening up unprecedented
opportunities for educational innovation, for
invention, and for discovery. We’re pioneering new
educational areas as well as new research directions. We have biological variants
of several of the engineering majors. And next fall, we’re going
to be initiating a new major in biological engineering. I think this Tech Day
program will show you how MIT is leading the way in
this new field of all fields, just as we have led the
way in disciplines that define the information age. I think it also demonstrates
the tremendous contributions MIT can and will be making
to this nation and the world. I’ve been looking forward
to hearing these faculty presenters talk about
their research for months, and I am delighted
to have the luxury of sitting in the audience
with all of you today. I’m anticipating an
exciting program, and I’m sure that MIT will prove
as it has done over the last several months, that my
anticipation cannot even come close to what we are
actually going to experience. So enjoy the morning
and I will return– I promise you it’s only
going to be the morning. But I’m going to return at the
end to moderate the question period. So enjoy this morning’s
presentations. Thank you. [APPLAUSE] VINCENT: This is great. Every once in a while, I think
a group like the Tech Day committee gets lucky. And we selected last fall,
bioengineering, and then to be honored with the
president Hockfield’s first time presiding over today,
sometimes you get lucky. And about that, last fall when
we were looking at this topic, trying to understand how do
we construct this program, we sat down with Professor
Doug Lauffenburger, and he just gave
us an education. And we were so thrilled
and learned so much, we said, geez,
what do you think? Can you do that on
a broader scale? He said absolutely. Remarkably, he thanked us this
morning for being a part of it, when in fact,
really, we thank he and all the faculty for their
time to put together today’s program. Doug is the director of MIT
bioengineering division. And at this point, just
turn it over to Doug. [APPLAUSE] LAUFFENBURGER:
Thank you very much, Doug, and it’s a
privilege to be part of this today with all the
alumni and friends coming back. We’ll see if there’s any
continuing adjustment that needs to be done. And I’ll be darned, there was. What I want to give you is not
the research excitement itself. I have to lay out the context
for what’s happening at MIT and why it’s happening,
why it’s happening now, why it’s happening here. And that’s the
purpose of this talk. So I’m going to talk about the
landscape of bioengineering at MIT and how we’re breaking
new ground in this landscape. Now, those of you who
haven’t been back for a while know that this landscape has
been changing dramatically, too. Breaking new ground all over the
local blocks with new buildings and new excitement. And so that’s been
changing, as well. But that’s not the
landscape that I mean. I want to talk about the
intellectual landscape of the interface between
engineering and biology. Now, there has been a
traditional biomedical engineering landscape, both
at MIT and around the country. And this has been around
for about 40 years. There’s been a field known
as biomedical engineering, bioengineering, medical
engineering at many places. And what it’s been all
about is essentially, the marriage of the
classical engineering disciplines in which students
are trained in mathematics and physics and chemistry,
married with medicine to solve medical problems
in ways that technology is based on, mathematics and
physics and chemistry could. Now, one important
feature of this is that it’s a multidisciplinary
application field. All the different
engineering fields could bring what they had to
bear in contact with medicine and solve problems from
their own unique disciplinary backgrounds. Accordingly, from the
MIT point of view, this was never appropriate as
an undergraduate major degree of study because
you were better off studying in any of the
other classical engineering disciplines and applying your
interests to medicine, if you so wished. At the graduate
level, of course, since you were taking
these other engineering disciplinary
backgrounds and using them to solve medical
problems, that was a very appropriate
type of graduate study to do within any of the MIT
engineering departments– electrical engineering
and computer science, mechanical engineering, material
science, chemical engineering, and so forth. And many of these programs
benefited from interaction with the health sciences,
technology and its connection to Harvard Medical School. The other feature is that
the bio in the bioengineering in the biomedical
engineering has, for those 40 years
in these programs, essentially been organ
level physiology, which is, of course,
crucial for medicine, but is essentially
descriptive that’s not the mechanisms by which
living systems operate. So you could describe what was
going on in engineering terms and try to work at
it from the outside, but you couldn’t control
the underlying mechanism. So that’s what’s been
around for 40 years. And it’s very important, had
very valuable contributions, and will continue to have
very valuable contributions, and will continue at MIT within
all these other departments. These kind of
contributions have been manifested in a variety of
career opportunities carried out by MIT alums. Probably many of you
have been working in these very fields,
medical devices of all sorts, diagnostics like imaging
and prosthetics, implants, extracorporeal devices, maybe
the pharmaceuticals industry and processing,
manufacturing, and delivery. And in all these very
important contributions of biomedical
engineering careers, the one interesting
feature is that training of engineers in the actual
biological science fundamentals really wasn’t crucial. You have to be really good
at your math and your physics and your chemistry and interface
with a physiological problem. That was what was
crucial, rather than the biological science itself. All right, with that
as a 40 year backdrop, now what is new besides
many buildings going up? What is new in the
bioengineering landscape? Well, as president
Hockfield told you, we have now created
something brand new, and in fact, I will
assert, unique– a biology based
engineering discipline that we call
biological engineering. Instead of being an engineering
discipline rooted solely in math and physics or math
and physics and chemistry, this one is rooted in the
molecular life sciences. What’s at the core of it
is genetics, biochemistry, molecular biology, cell biology. This engineering
discipline is rooted in the mechanisms by which
living systems operate, the very molecules
and cells that make things happen for good or ill. But these sciences are fused
with engineering and analysis and synthesis, just like all the
other engineering disciplines fuse analysis and synthesis
with physics and chemistry and so forth. So it’s focused on taking
these life sciences and understanding
them in a quantitative and systems oriented way and
designing new technologies based on these
biological, molecular, and cellular components
and mechanisms. So once you have
a discipline now that’s just like
electrical engineering, mechanical engineering,
chemical engineering, but is now rooted in
a different science, now MIT has agreed that
this is appropriate for standalone degree programs,
both a new PhD program that we started in 1999,
and what’s quite exciting is a new bachelor’s
degree program that starts this very fall. Right. But why has this
change happened? Why was this traditional
landscape for 20 years and now something very
radically new is happening? Why is engineering and
biology interaction changing so dramatically? Because the biological
science itself has changed so dramatically. The engineering can’t change
in a revolutionary way unless the science is changed
in a revolutionary way. And that’s what’s happened. In the past 20 years, there
have been two major revolutions in biology– the molecular revolution
roughly in the ’80s, the genomic revolution
roughly in the ’90s. So if you took biological
science before the ’80s, you did not learn these
types of biological science. This has all happened
in the last 20 years. I went to college in the 1970s. And in the 1970s, I
did not learn this kind of biological science. So what was the engineering
biology interface, quote, then, before these revolutions, back
when I took biological science in the 1970s? It honestly was not amenable
to engineering analysis and synthesis and design. Engineers could
not access biology before these revolutions because
biological sciences didn’t understand the
actual mechanisms. Few of the actual components,
the parts, had been identified. You could only quantify
things at these higher organ levels of hierarchy, which
is why engineers worked at the organ physiology level. You certainly couldn’t
manipulate any of the molecular components. Even if you
identified them, there was nothing you
could do about them. A gene was a gene and a
protein was a protein. There wasn’t enough data
to build models and design. And it was hard to generalize
from humans to mice to rats to worms and so forth. So the modern era of
bioengineering really began with the molecular biology
revolution in, essentially, the 1980s, because
that permitted mechanistic components to be
identified and manipulated. And you can’t engineer
anything unless you know what the components are
and you can manipulate them. And I just show this
for illustration. This is a bone. And for all the
traditional years of biomedical engineering, bone
could be perceived and dealt with only on the basis of being
this macroscopic load bearing mechanical object. For instance, my mother has
had six hip replacements. Many years ago she crushed
her hip falling off a bicycle. And what could biomedical
engineers do about that? Well, pre the revolution,
all they could do was replace that hip as
a mechanical load bearing substitute. First wood, crutches,
then metals, the implants became metals. Then plastics. The implants became plastics. And none of them worked very
long for a variety of reasons, mainly that they
really weren’t bone. They didn’t interact
with the surrounding tissue and the immune response. They didn’t respond properly
to stress and so forth. But with the molecular
biology revolution, now we know this isn’t just
some macroscopic load bearing object. What has been identified
as the cells that are dynamically breaking
it down, building it up. The proteins by which
the cells are stimulated to break down bone and build
it up and the genes that encode for the proteins
that tell the cells what to break down and to build up. And so now we have the hope of
regenerating not wood, metal or plastic, but
regenerating living bone. And you’ll hear a
little bit of that from Professor
Griffith in a minute. So the very fact of identifying
what the mechanistic components are and being able to do
something to manipulate them was crucial. Now, this modern era
has been catalyzed by the genomic biology
revolution in which, just depicted here are a large
number of sequencers cranking out the identities of
genes and your chromosomes and what kind of proteins they
might encode for and so forth, because this has
accelerated the rate at which molecular
compounds can be identified. Instead of one by
one in a hard slog, we can now find tens and
hundreds and thousands of them at once. So it’s the ability of
finding the parts has just been dramatically accelerated. The concomitant has been showing
even how much more complex these mechanisms are than
are previously appreciated. People had hoped with the
onset of molecular biology that if you could find maybe
one gene or one protein, you could just fix something
and a disease would go away. The genomic revolution
looking at tens and hundreds of thousands of the components
and their interactions at once now have made us realize that’s
going to be really difficult. These are very complex
networks and machines. So these 20 years of
two biology revolutions have led us to a now engineering
biology interface that’s just night and day different. Not only is biology amenable
to engineering analysis and synthesis– we have parts,
we can manipulate them– but again, I would assert
that biology benefits from and maybe even requires
the engineering analysis and synthesis approach. Now molecular components
are being identified. The mechanisms involved
in molecular cell actions are being elucidated. Quantitative analysis
is possible all the way from DNA structure,
protein DNA interactions, up through molecular
networks to cell functions to tissue functions. You can quantitatively analyze
every step along the way, not just at the organ level. And molecular components can
now be easily manipulated. It’s very simple to go
in and change a gene, change a protein, change
any piece you want. What’s the hard part? The hard part is to
understand and predict what happens when
you’re going to do that. Right now it’s almost
trial and error. And that’s what we’ve
got to get beyond. And that’s where
engineering comes in. So we believe that
this is the time for a new fusion of biology
with engineering because of these two revolutions. The molecular biology revolution
finally, for the first time, permitted engineers to access
the science of biology. And the genomic revolution
now requires biology to be addressed by engineering. So we now have the parts. We can manipulate them. And it’s darn complex
and engineers really can get involved and
have to get involved. Just as a few
illustrations, if we think it’s something
like cells migrating the cells of your
immune response seeking out cancer cells or
viruses or healing a wound, how would you make that happen? How would you enhance the rate
at which wounds were healed or immune response operated? Or if this was a tumor
cell metastasizing to an improper organ,
how would you stop that? Well, we have to understand
how this cell migration works in terms of what’s depicted here
schematically, it’s a machine. This is the cell
and the membrane and it’s connected molecularly
to the extracellular matrix and allows the cells to crawl
and pull their way through. And the cell has motors inside
that are generating the forces in just the right places. It’s an exquisite
molecular machine. And we need to understand it in
terms of molecular structures and properties and
biophysics and biomechanics. And these machines
aren’t autonomous. They’re not spontaneous. They’re highly regulated. The cells will migrate
under some conditions and not under other conditions. They’ll proliferate
under some conditions and not other conditions. What’s regulating them? Well, there’s a fascinating
set of biomolecular signal processing circuits
that take information from the environment and
say, what should I do now? Should I migrate? Should I not? Should I proliferate? Should I not? Should I die from a stem cell? Should I differentiate
into something else? How do I decide? So they take their information
from their surroundings and process those signals and
change their gene expression, metabolism inside a skull,
and so forth, and carry out different functions. They do signal processing. So we need to understand this
regulation of the machines by these signal processing
networks the way, perhaps, electrical engineers,
computer scientists and so forth might think
about regulatory circuits. And a fascinating thing that’s
really becoming realistic to think about– especially at MIT
because we realistically think about things that
most places really only will speculate about– there’s an opportunity of a
field called synthetic biology, where folks on this
campus and others are trying to design
these machines and networks from scratch. Instead of just intervening
in them one molecule here or there, they’re saying, can
we design our own DNA sequences that will produce the
right proteins that will interact in the right ways
to build a network or a machine with the right sort
of topology then that would carry out a
dynamic function of one sort or another? People are building their
own machines and networks from designing their own
DNA sequences from scratch. A fascinating field
and it’s starting here. There’s a young professor
named Drew [? Wendy ?] who’s really the– it’s very interesting. He’s assistant
professor and he’s known as the father
of the field. So that doesn’t
happen very often. Now at the same time, we
start with these mechanistic underpinnings. We say, for an engineer
to access biology, you’ve got to have the parts. You’ve got to know
what the parts are. You’ve got to be able
to manipulate the parts. But of course, you don’t
do that in a vacuum. You wanted to think about what
higher level functions you’re aiming at, whether it’s
physiology, pathology, therapeutic interventions,
brand new devices, like I suggested on the previous
slide, brand new materials. Professor Belcher will talk
about brand new materials from biomolecular mechanisms. [INAUDIBLE] will talk about
new types of therapeutics from these molecular mechanisms. So you’ve got to think about
the complexity of biology in one dimension, which
is many, many components at once in another dimension
of all the information you know about it– sequence, structure,
thermodynamics, kinetics, mechanics. And in the other dimension of
the physiological complexity– individual cells, cells and
population, cells and tissues, and so forth. So the underlying mechanisms
govern the higher level functions. And to an engineer, this
is the way we think. You think you’re thinking about
higher level system function and how you’re
going to design that or how you’re going
to intervene in it from the molecular and
cellular components. So there is something that
we call this discipline of biological
engineering that is both the discipline important
for science and for technology. And it looks a lot, in
these first few bullets, it should look just like
any of the other engineering disciplines maybe you studied– chemical engineering,
mechanical engineering, electrical engineering,
and so forth– because you and it will analyze
complicated many component hierarchical systems. You do that in all those
other engineering disciplines. You’ll synthesize
designed technologies, not trial and error. We don’t build any
other technologies by trial and error. It’s time that we really
stop doing medicine by trial and error. And I think Martha Gray will
talk about that, as well. And all engineers operate
with this paradigm. There’s different
ways to pose it. But measure and
model and manipulate with design principles
and parameters, computational
models, and so forth, construct your own components
to have the right properties. The only difference between
biological engineering and all the other engineering
disciplines is what are our
components and mechanisms? In other cases, it might
be organic chemistry, inorganic chemistry,
solid state physics. In our case, it’s biological
molecules, biological machines and networks, and the cells
that comprise, essentially, the organizing
principles for these. So just like any other
engineer, except the science is different. But the science is still
quantitative manipulable components. So it’s clear to us
that this is time to have created a new
biology-based engineering discipline to sit alongside
its sibling disciplines in the School of Engineering. And the idea of creating a
new engineering discipline isn’t new. MIT does this every few decades. Electrical engineering,
mechanical engineering, at one point in
time didn’t exist. At one point in time they were
established as now physics based engineering disciplines. 100 years ago, chemical
engineering didn’t exist. Materials science
and engineering really didn’t exist. And they came into
being as chemistry based engineering disciplines. So now we are in the 21st
century creating another one. It just happens to be a biology
based engineering discipline. And the biology is
molecular life sciences. And it comprises both
technology and science facets. We are designing and
developing biology based technologies, as you’ll
see in the ensuing talks. And at the same
time, I think we’re facilitating the advance
of the science itself, turning engineering
back to the science and how can we
understand it better? So it’s a two way street between
the science and the technology, as any good engineering
discipline, really, is. I won’t go into
this in much detail, but we are very excited about
the new biological engineering major that starts this fall. Professor Griffith,
who speaks next, was really the principal
architect of this. He spent an immense
amount of time constructing this in partnership
with many other departments. And all I want to show
you is the science core beyond the
typical freshman year. It requires organic chemistry,
biochemistry, molecular cell biology, genetics,
just as I said, along with the mathematics
that one would require. And then what’s listed here
are nine brand new courses. We didn’t just sort of
cobble together, oh, let’s borrow a course from– let’s borrow a subject from
this department, a the subject from that department, a
subject from that department. These are nine new
courses all designed from scratch saying,
if we’re going to understand and manipulate
this kind of biology of genetics and biochemistry
and molecular cell biology, what is the way an
engineer would do that? What’s the right thermodynamics
to study this science? What’s the right mechanics
to study this science? What’s the right signal
processing or fields or control or computation? You can’t just borrow the same
old and sew them together. It’s brand new engineering,
because it’s brand new science. So these are brand new subjects. And a lot of faculty
effort, but the faculty are very enthusiastic
about this. What’s the point of learning
this new discipline? We’re very convinced and we
have lots of confident input, likewise, from our
friends out in industry that there’s going to
be new types of places that engineers have never been
hired before because engineers haven’t been capable of
knowing this kind of a science. So, yes, medical devices, but
instead of building the boxes these folks are going
to design and manipulate the interactions of these
devices with their environment. So maybe you still
have a prosthetic hip. Maybe we can’t quite
regenerate my mother’s hip from cells and genes
and proteins quite yet. But her hip may fail
less if we can figure out how to interface that
synthetic material with the immune response,
the inflammatory response, the surrounding
tissue and so forth so that it doesn’t fail there. Medical diagnostics but
now will be in terms of these biological mechanisms. Analysis will be at the
gene and molecule and cell levels, the appropriate
information technology. We can imagine cell and
tissue based therapeutics. Engineers will be involved in
the pharmaceutical and biotech industries, not just to
manufacture and to deliver, but to actually design and
to develop, to discover. These kind of
biological engineers will help facilitate
the discovery of where the right
targets should be, what the right
therapeutics should be, because of their
ability to think in a quantitative
systems oriented manner about these very, very
complex systems, and all the way along the way to
predict pharmacology effects, toxicology effects,
and so forth. And finally, but maybe
even more importantly, in terms of the scope. Biological engineers
will not only contribute to medicine
and human health in a very powerful new way. They, we believe, will transform
many of the other non health care associated fields,
creating new types of materials and devices in a perhaps
environmentally benign manner with much more controlled
properties, things you’ll hear from
Professor Belcher, understanding better toxicology
environmental health, the pathogens, the toxins,
carcinogens in our environment, national defense, new
types of biological sensors and actuators that can
carry out functions in far more exquisite ways than
physics and chemistry might by themselves. So let me end the
sort of pictorially. The traditional biology
medicine engineering landscape looks something like this. You can imagine problems
driven by applications in health care– maybe clinical, hospital,
aerospace, military, maybe the pharmaceutical
device diagnostics industries. And I’m speaking only about the
engineering disciplines here, certainly the
science disciplines would be analogous situation. You could take any one of
these engineering disciplines– chemical engineering,
electrical engineering, mechanical engineering,
material science– based on chemistry,
physics, and math, the way it has been for 40 years,
interface with medicine be a wonderful
biomedical engineer and solve problems in
the health care field. That’s been the
traditional pathway and it’s an important pathway. You can also contribute
to biotechnology perhaps in the
processing manufacturing and delivering or so forth. So what’s new, what’s new– and some of you over here
won’t be able to see this– there will be something
new showing up in the bottom left
hand corner, which is the sibling discipline
of biological engineering, that according to
the curriculum I just showed you a
couple of slides ago now takes as a
central science based genetics biochemistry
molecular biology, cell biology and
can very powerfully interface with medicine and
solve biomedical engineering problems for the
health care industry, can also be applied to novel
types of biotechnologies, both for the health
care industry in terms of therapeutics and in other
types of diverse industries as new materials, the
environment, and so forth. And so we’re just placing
this new discipline alongside the old ones because
it has a new science space. You notice in parentheses in
the middle I have bioengineering because at MIT what
we’ve decided to do is called bioengineering
the whole scope. So biomedical engineering,
medical engineering, you can think about as
the precise application, regardless of the engineering
discipline itself, to medical problems. Biological engineering is this
new biology based discipline. And then bioengineering
just covers the whole scope. So you can major in
electrical engineering, minor in biomedical engineering. You can major in biology, minor
in biomedical engineering. Or now you can major in
biological engineering per se and have a whole world of
applications open to you. What do I want to say
about the landscape, this is another
landscape that’s changed. These are the faculty
that are affiliated with the biological
engineering department at this point in time. And if you can see
the colors, you can see how the
landscape has changed. The names in blue, many of them
who might be familiar to you, were members of the department
when we started in 1998. The names in green have
all been hired since then. And they give great credit to
the president and the provost and the deans of these
last few years of investing in this brand new
opportunity for MIT, which I truly believe
were at the forefront of, were defining, and
the Institute has been investing in in terms
of its most precious asset, which is people. Finally, I may say, to go back
to this landscape, that it’s crucial to emphasize that
we do this in partnership for the Department of Biology. We have a wonderful– I won’t say unique, but
it’s really unusual– partnership between
biological engineering in the school of
engineering and biology in the school of science
in terms of faculty members who have appointments in both,
courses we’ve co-developed and co teach, parts of the
undergraduate curriculum that are now transparent
research programs and so forth. And so biology is mainly
located in this brown box. Biological engineering is
mainly located in this blue box. But we have this wonderful
handshake partnership. And I really think this
could have happened only at MIT, because of the world
class biology department, because of its own
role in defining the biology of the
molecular biology revolution and the
genomic biology revolution that they could
appreciate what we wanted to do from the
engineering school in having a similar revolution
in biological engineering. So I thank you for your
patience with this overview. The real excitement is going to
come from the next few people because they’re going to show
you the fascinating things that we’re actually doing. But I hope this painted you
a picture of what is new, why we’re doing it,
how we got there, and why this is really such an
extraordinary point in time. Thank you very much. [APPLAUSE] VINCENT: Wow. That’s pretty incredible. What a remarkable turn of
events these last few years. And what a great setup. Thank you, Doug,
for really teeing up the rest of the
excitement this morning. I mean, it is a remarkable story
about the courage, the vision, the leadership, to
look to do this. And this is I think what I know
I certainly love about MIT, and I hope this sort
of thing continues. Next we’re going to hear from
Professor Linda Griffith. As Doug talked about,
Professor Griffith was instrumental
in pulling together a new undergraduate major
here, bioengineering, at MIT. She’s also going
to share with us some of the really fascinating
work in tissue engineering. [APPLAUSE] GRIFFITH: I’ve been
giving a lot of talks at MIT in the past year,
mainly about our new education program. And it’s really wonderful to
be able to share with you some of the research that
drove a lot of faculty at MIT to try to create
the kind of student who could push these research
frontiers forward. I’d like to acknowledge
at the beginning before I start the research
talk the wonderful support we’ve gotten from the
Alumni Association in helping develop
our curriculum. Several of the faculty
received a class of 1960 Teaching Award for
some of the very early ideas that we had in moving toward
this kind of program that would bring biology and
engineering together. And it allowed us to
work with students a lot more effectively. And we’re very, very
grateful for that support for the education program. What I’d like to
do in describing a little view of the field
of tissue engineering today is give you two
sort of vignettes. One is where the field is going
in terms of what we classically think of of tissue engineering
in building organs and tissues to treat patients. And so that’s the bench to
bedside part of the talk. In the second part
of the talk, I want to give you a
glimpse into what may seem like a more mundane,
but ultimately, I think, is a far more
powerful application of tissue engineering, and
that’s coming back to the bench and using tissue engineering
for a range of applications and drug development
that will hopefully help us put all the surgeons
who do organ transplants out of business someday. Except for, of course,
my postdoc mentor. So to get us
oriented, let’s think about what generated
enormous excitement in the field starting almost
a couple of decades ago. And I call this tissue
engineering version 1. A common paradigm
in the field is that we have some kind of
source of cells that might be used to regrow a tissue. And in this case, we
could imagine a patient who’s lost the outer
part of their ear could have some cells taken
from their rib cartilage if only we could grow
them into the proper shape to put in the patient. So this is just
an example of how you might do that and
make a scaffold that’s made of a porous
biodegradable polymer, actually common surgical
polymer, combine it with cells. And here this is a
demonstration that you can implant this kind
of cell polymer scaffold beneath the skin and
have the tissue grow from the cells that were
implanted onto the polymer. This was an experiment
done using a mouse model. And you can see the shape
of a human ear growing on the back of this mouth
after a couple of months. And so this kind of application
generated lots of excitement because it actually works,
and it works fairly well. And so these kinds of things
are now moving into the clinic. We can ask the
question, though, if it works for us an
application like that which is fairly constrained in
the number of patients, will it work for
applications where there are lots of patients
and more serious problems? So the answer, happily, is yes. This shows a radiograph
from the patient of one of my collaborators,
George Marshall, at the Cleveland Clinic. This patient has a very
large tumor in his bone. You can see the diffuse area
here is cancer of the bone. And it needs to be removed. So what will happen in
a patient like this, and happen in this
patient, typically, is this part of the
bone gets cut out. And what’s done right
now is, because it’s a very large defect in the
bone, a piece of cadaver bone will be placed there in order
to give the patient support following the surgery. Well we all can imagine
that cadaver bone doesn’t heal so well. There are potential
problems with infection. It’s not quite the same
as having the real bone. Another application area that
some of you may be facing arises from injuries to joints. You can see, perhaps,
pretty clearly that this soccer player
here is experiencing an injury that will later
lead to great problems with his joint. And, in fact, people
who experience injuries like this and even lesser
ones will go on offense to develop osteoarthritis. That’s a wearing away of
the surface of the cartilage that lines the joint, in
this case, in the knee. And osteoarthritis is a
leading cause for hip implants. My mother also just
had a hip implant, and it was due to
the wearing away of the cartilage in her knee. So wouldn’t it be better than
having a joint replacement that goes on to lead to
complications later and never works perfectly
as the original joint? Wouldn’t it be better
if we could somehow repair this lesion, and
grow cartilage back, and, have it secured firmly
to the underlying bone just as it is in
the normal tissue? So one of the great areas of
activity at MIT, at my lab and and many others,
is trying to combine these concepts of
understanding the facets that govern cell behavior, the
underlying molecular mechanisms that cause cells to migrate
into a wound bed and cause, say, stem cells that are present
there to proliferate and turn into bone or turn
into cartilage. We’re trying now to
combine these factors the molecular factors that
govern the properties of cells with engineering approaches
that would create environments that cells could be
transplanted into these sites and go on to form new tissues. So in work in our lab, we
said, okay, what we really need are two aspects of this. We need to be able to
build a large scaffold that would be like the house that
these cells would live in. And at the same
time, we need to be able to give those
cells cues for what they should be doing when we
put them in an implant site. So a common situation
in the clinic is an orthopedic surgeon
would have available, if he’s doing a
joint replacement or wants to repair a
joint or repair a bone, he can harvest bone
marrow from that patient. And the bone marrow
has in its stem cells that can go
on to form bone. So we ask the question, can we
build a scaffold that will help those cells go on to form bone. And so let me just
take you through one part of solving this problem. In order to build
a device that could be used to put the
cells on, we had to invent a whole new way
to process materials that would be appropriate for
implanting into people. So when I started
out doing this, materials processing wasn’t my
particular area of strength. I knew a lot about
the cells and how to design polymers to
interact with cells, but making those
into a big device wasn’t something that
I was that talented at. So it turns out that
MIT, as we all know, is a place where you run
into people all the time from different disciplines. I ran into a mechanical engineer
who had invented a process called three
dimensional printing. And he’d invented it to do rapid
prototyping of airplane engine parts, essentially. So what we decided
to do together is say, gee, when you’re
making airplane engine parts, you need to make
complicated shapes, you need to build them from
some kind of computer model. It sounds like that’s the
same kind of challenge we face in building scaffolds
for bone regeneration. We might want to start with an
MRI image of a patient’s jaw, for example, and build
a new jaw just exactly to match the other half
of the patients jaw. So can we take a process
that’s computer controlled and that builds complex
three dimensional objects up from scratch, can we
take that and adapt it to tissue engineering? So starting about 10 years ago,
this is exactly what we did. So we could start with
an idea saying, okay, we have a computer model
for what we want to build. We want to build a scaffold that
has lots of pores and channels that will facilitate
tissue in growth. And we want to build it
in a particular shape. And the way that we can
do that is to build it up as a series of very thin,
two dimensional layers, step by step. So the three dimensional
printing process is one of many kinds of
manufacturing and prototyping processes that are now
endemic in many areas that are called solid freeform
fabrication methods. They build up complex
objects as a series of thin two dimensional slices. So you start, in this
case by spreading, a powder of the
material you want to build an object from, in
our case, a bio material. You spread a powder
on top of the piston, so you just roll out a
very thin layer of powder. Into that powder,
using a computer model of what you want to build,
you print a binder or glue that holds the powder particles
together everywhere you want a solid part of
the object to be formed, so then you can build
many, many devices at once. You drop the piston and– You finish printing, you
drop the piston down. And that gives you the
chance then to roll out another layer of powder. So this has now been licensed
and is being commercialized. There’s actually some early
stage bone regeneration products and patients now. But some of the
frontiers are actually solving those
extreme challenges, such as joint repair. So Therics, one of the companies
that licenses that technology, teamed up with a tissue
engineering company, Advanced Tissue Sciences,
to design and test a scaffold for joint repair that
takes into account the number of challenges you have in
building a complex scaffold that will allow a region
to have both bone grow and to have cartilage grow. So the idea here is that
we can, at the top part of the scaffold,
build a region that will favor the seeding
of cartilage cells into the scaffold and their
growth into cartilage tissue. And we can do that by arranging
the pores and channels in a way that let the
cartilage cells easily get seeded in here. And then we can
create a second region that has different materials
and different structure that facilitates the
end growth of bone. So you can take
these in the lab. And it’s quite straightforward,
actually to build these. And this shows a joint. In this case, this
was done in sheep as a test for how this
process would work. So the scaffold was
seeded and cultured with cartilage cells,
implanted into the joint. And you could later
see that the cartilage formed in vitro and allowed
the joint not to move. So these sorts of
applications where we combine cells and
scaffolds are really moving into the clinic now
in somewhat incremental ways. But you’re starting to see
them appear in applications, particularly, in
orthopedics and in places where we have a connective
tissue repair that are needed. But you can imagine,
with the power of combining cells and
scaffolds, what if we could build organs that would replace
much more life threatening functions? So if we can do this for
tissues like bone and cartilage, could we do it for diseases
that are truly life threatening? So you could ask
the question, well, gee, we have an incredible
shortage of organs for heart transplants,
for liver transplants, for kidney transplants. Can we take this same
concept and maybe take one donor organ,
one liver, and use it to cure 100 patients who
need an organ transplant? Could we do that by taking a
tissue engineering approach, and taking the cells
from one donor organ, and constructing a whole series
of smaller organs that could be transplanted into patients? So many people in
the field have had a vision that
since liver is such an amazingly
regenerative organ– I think that some of you may
remember from your high school mythology class the
story of Prometheus. And he stole fire from the gods,
so he got chained to the rocks. And the eagle would come every
day and eat his liver out. And then it would
grow back overnight. And the whole process
would repeat the next day. So even long ago, it was
known that liver regenerates. [LAUGHING] So all we need is the eagle
to come and have Prometheus chained to the rock, and we
can solve the organ transplant problem. Well, maybe not. No, but that makes people
think, especially transplant surgeons who see organs grow
back after you excise a tumor. They ask a question,
could you then say, okay, maybe we could
get regeneration and culture if we had a very
complex scaffold that would let cells assemble
into a liver-like structure with an artery and a vein? Maybe what we could
do is take one organ and make a bunch of mini
organs that we could then transplant into patients. And so this captured the
imagination of a lot of people. And it’s something that I think
is still envisioned as possibly one day coming to fruition. But that one day is
pretty far in the future. I think saying 10 years
is incredibly optimistic, 20 years is optimistic. And keeping the attention
of NIH and so forth to fund this kind of research
for that long may be the biggest challenge. So we can take a step
back and say, gee, where will we be in
a couple of decades with this kind of
organ replacement? And option 1, and
one that certainly gets the attention
of popular press, is that we’re going to be able
to have more organ transplants. So all those people now
who die because they don’t get a transplant will
be able to get liver in a box. Hopefully, it won’t
be made in my lab, but MIT isn’t a good
manufacturing process approved place, I don’t think. But the idea is that you would
have these organs available for transplant. So that’s one option and one
that I think a lot of people get excited about. It would certainly
be very dramatic. You get brought back
from the brink of death by ordering your liver off
livers.com or whatever, and there it is,
made just for you. Option 2 doesn’t give
the press to have so many dramatic lifesaving stories. Option 2 is that we just
get rid of transplants altogether and,
maybe, only do them under unusual
circumstances so that we put all those organ transplant
surgeons out of work. And now they get to do
other things, like golf or whatever they would maybe
be doing if they had more time. So we could ask ourselves– I actually see a show of hands. Who would prefer option
1 if you’re the patient? Okay, option 2, that
we cure the disease much earlier so people
actually never even, maybe, have to go to the hospital? They may go to their doctor,
get diagnosed and take a drug, and be able to cure the
disease that would ultimately lead to the organ transplant. So we can ask, what
are the bottlenecks in realizing option 2? Why aren’t we there with more
diseases and more organs? And I’ll say that my own
epiphany about this vision of why aren’t we
pursuing option 2 harder was really brought
about by the creation of the biological
engineering division. In fact, several
faculty who became an integral part of the
division were focused on the area of toxicology. And when we think about
toxicology, a lot of us want to maybe yawn. I certainly felt
that way initially, but then there was
the aha moment. Preventing things is much
better than fixing them later on when they’re
much more serious. And so we really started
to take a hard look at what it was that prevented
us from going and preventing diseases. We certainly do have a
lot of disease models that we use very, very
productively to understand and help develop treatments
for human disease. So Bob Horvitz and MIT won
the Nobel Prize, in part, for showing that c
elegans, a nematode worm, could be used to understand
some facets of cancer to understand how cells
undergo programmed cell death. And we certainly can
use yeast to understand aspects of our own biology. Mice and all of
these animals here can reveal different
facets of human disease and help us enormously in
developing therapies and cures for these diseases. But despite the utility
of all these models, we know that mice are
not little people. Worms are certainly
not little people. Mice are not little people, and
even chimpanzees and orangutans are not people. And what we would be
really excited to have, if you’re excited– the “we” being perhaps
a drug company– is a model of a human that could
be used to really understand a complex disease or
use to predict how a human will respond to a drug. So we can do pretty
good right now in looking at the individual
molecular interactions that govern the cell behavior. As Doug described
earlier, we’re starting to put together the molecular
interactions that transmit signals from outside
the cell to inside the cell that govern gene
regulation, for example. And we’re somewhat
good at understanding how different cell types
interact with each other to create a more
complex response. But as we get up to the tissue
level and certainly the organ level, humans start
to diverge enormously from animals in how they
respond in most cases. What we would love to do is
have readily accessible models, therefore, of how human cells
and tissues interact to come up with a complex response. So about seven years
ago, we started to really refocus efforts
in our lab toward this goal. We still continue to develop
some exciting approaches in treating connective tissue
diseases by tissue engineering. But we decided that trying
to build models of humans so that we could study
disease, and hopefully cure it, would be an enormous
contribution ultimately, to health care. And certainly, liver
probably exemplifies the need for this approach
more than any other organ. So you– I don’t
know any of you who know anyone who has
hepatitis C. It turns out, hepatitis C was discovered
around the same time as AIDS, but there really aren’t great
therapies for hepatitis C right now. There are therapies, and there’s
some new ones coming out, but it’s been incredibly
hard to study this disease, because it only affects
humans and chimps, and it targets liver cells. And it turns out that when you
take liver cells out and put them in culture, they become
refractory to infection by hepatitis C. So
you can’t really study the whole viral
lifecycle in culture. And guess what? Hepatitis C is the
single-leading cause of liver transplants
in the Western world. So if you eliminated
hepatitis C, you would have a lot more organs
available for other people who need transplants. And in fact, if you take
this approach and say, gee, what are some of the
other causes of transplants? Well, there’s toxicity of drugs. Even Tylenol kills about
100 people a year in the US. And right now, one of the
leading causes for drugs to fail in phase I clinical
trials is liver toxicity. Very difficult to
predict liver toxicity, because human livers metabolize
drugs, in most cases, different than animals you
use to test those drugs. So there are all
of these problems that we face for
which there are no really good accessible models
to study the human condition. And so why is that? We know we can take cells
out and put them in culture and get them to
behave, in some cases, like they do in the body. And so with liver,
we can do that. We can take liver tissue,
dissociate it down into the individual cells,
and put them in culture, and get them to
look very beautiful. So this is just an example. If we take– in our
lab, we use rats for these kinds of studies,
because they’re commonly used in toxicology. So we can isolate the
cells from a rat liver by digesting it with
enzymes, and put it in something that makes
it think that it’s a little bit– that it’s
like it’s back in the body. We use an extracellular matrix. And then we can take a picture
of these cells a couple of days after they’re in culture,
and they look very beautiful. Their nuclei are
very round, they have a cytoskeleton organized,
and they look quite happy. But looks are deceiving. When we go in and start to
measure the level of functions that those cells have that are
liver specific, it turns out, they really aren’t acting
like liver anymore. So this just illustrates that. So those same cells I just
showed you– if we go in now and break them open and
get their messenger RNA out and do a microarray
analysis to look at all the levels of
most of their genes using a microarray
experiment, we see that the cells are
losing function quickly. So let me orient you. So this is a scale that
everything above the line means that the gene
expression level has gone up. Everything below the
line means that it’s gone down relative
to liver in the body. And this is over a
few days in culture. So what you can
see here is we are looking at the genes
involved in drug metabolism. So there’s about 100
or so of these probes on this particular
microarray chip. And what’s really obvious is
that, really soon after cells are placed in culture, they
lose their liver-specific gene expression. And this is what bedevils the
efforts of the drug industry to do a better job
predicting toxicology. It’s loss of these
kinds of genes that bedevil our efforts to
culture hepatitis in culture, and to study many
facets of liver disease. So what can we do? We asked the question, can we
recapitulate function of liver by recapitulating the
structure of liver? Instead of doing
a 2D cell culture, maybe what we should
be doing is trying to create environment in culture
through tissue engineering that really replicates this 3D
microscopic environment. So for example, this is
the liver capillary bed. It’s the business
end of the liver. It’s about 1/2 a
millimeter across. And in that little environment,
hepatocytes, the cells that carry out most of
the functions of liver, are in constant communication
with other cells. And in fact, there
are mechanical forces on these cells arising from
flow of blood that give rise to signals that tell the cells
yes, you’re in the right place. Keep doing all the
things that liver does. And in fact, David Darnell
won a special Lasker Award for showing that when you take
cells out of this environment, they lose their message
level and all the things that make them like liver. So I’m going to fast forward
through all the development we did to come up with a design
for how we could create this three-dimensional
structure, and show you what we’re doing
now in our efforts to build, essentially, liver
on a chip and use it for a number of applications. So what we reasoned
is that if we could create a little
microenvironment that would give the cells
some cues, they could organize themselves into
a capillary bed-like structure. And it turns out, using silicon
microfabrication technology was a very rapid route
to doing that. We have a wonderful Microsystems
Technology Lab at MIT that lets you prototype
different designs. So we went over there. And in a period of
about three years with a lot of
support from DARPA, who was interested in
using little livers to detect viruses that may be
warfare agents, what we did is we came up with a design
based on silicon chip technology that would
foster the formation of three-dimensional
tissue structure from isolated liver cells. So what you see here is a
chip that has bored into it a whole bunch of holes. So this chip is about 2/10
of a millimeter thick. And when we bore holes in it,
and then into those holes, we seed liver cells. And the dimension here
is 3/10 of a millimeter. And so what happens
is the liver cells attach to the walls of
the holes in the chip. Now, we modify the
walls of the holes in this chip with molecules
that the liver cells recognize, and they form these
tissue-like structures. The structures are
stained in green, and that actually indicates the
cells are healthy and viable. And so what you see is formation
of tissue-like structures. That happens very rapidly,
and then the tissue appears stable over time. The whole thing lives
in a bioreactor, so we place that chip, the
scaffold with the cells, in here. And then we can flow fluid
through the whole system. So you can see a
set of pumps here that are pumping fluid
through, and we essentially have liver on a chip. Now, this kind of
arrangement allows the cells to capture features of what
liver architecture looks like. So you can see vessels
here, endothelium– and in fact, we can
follow this using what’s called two-photon
microscopy that lets us optically section
throughout the tissue. So what we’re seeing here
is sequential sections through the tissue over 10 days. And what’s labeled here are just
the blood vessel lining cells. The blood vessel
lining cells are green against– the rest of
the cells aren’t labeled. You can see that over a
period of a couple of weeks, they’re forming vessel
structures within this chip so that we truly are creating
a three-dimensional tissue that could be used for a
lot of applications. So now we’ve got this
tissue, and we’ve gone through a lot of
different kinds of assays to show that it has a very
high level of function. What can we use it for? Well, for one thing,
we can use it to try to predict drug toxicity. And what Doug
Lauffenburger didn’t show on his campus changes
were the things associated with the MIT campus changes– Novartis Research Headquarters,
Pfizer Research Technology Center. All these
pharmaceutical companies are looking to MIT
for new methods that can help with drug development. And in fact, one of the
huge applications of this is to build little
livers that could be used in high-throughput
screening of new drugs. One other application–
there are a number of applications listed here. I’m just going to show
you briefly some results from one that we’re working on
in the Biotechnology Process Engineering Center. And that is this issue of
why gene therapy hasn’t quite panned out as much as everyone
hoped about 10 years ago. There was, a little
over five years ago, a very significant event in the
whole field of gene therapy, and that was the
death of a patient, Jesse Gelsinger at the
University of Pennsylvania. He was undergoing a trial
to correct a genetic defect in liver, and he died from the
effects of the gene therapy vector, which happened
to be an adenovirus. So one of the huge
challenges in predicting how well these kinds of
gene therapy vectors work is that what you see
in a 2D culture then doesn’t get
replicated in animals. Very huge differences
between what you see in mice and what you see in
cell culture when you’re designing these vectors. And then huge differences
between what you see in a mouse and what you see
in a person when you go to use it clinically. So we can now use this new tool
of a three-dimensional liver to start asking questions on
why these kind of gene therapy vectors are toxic,
and try to build in safeguards against that. And address those as
we design new vectors. So just as an
example, we can use the classic adenoviral
vector– in fact, this is a vector we get from
the University of Pennsylvania– and monitor the whole
process in a model way by having cells express a
green fluorescent protein. So just to give you an
idea of how this pans out, we can measure many, many
things about how these gene therapy vectors
interact with real 3D tissues in this new system. So here, we’re showing
a 3D reconstruction of a single capillary
bed in our little unit. So again, the dimensions here
are about 3/10 of a millimeter, and this is stained. It’s a 3D reconstructed image
from optical sectioning. So we’ve stained
it so that cells expressing the gene that we’ve
transfected in are green. So these cells have taken up
the gene and are expressing it. All of the cells
are stained blue, and cells that have been
killed are stained red. So now we can start to build
quantitative models of how we should design gene
therapy vectors to be effective and
actually also safe. And we’re working with a lot of
other people in that project. Another application area
that is hugely challenging is understanding
how to stop cells at the very earliest
stages of tumor metastasis. So often, we can’t
see a metastasis until it’s about a
centimeter in diameter. The resolution methods
are getting better, but what we really
would love to do is figure out, if you
remove a primary tumor, how do you keep the cells that
have moved to distant organs, and our single cells invading
the tissue from growing into larger tumors? So we can use this little
microreactor system. Because we control the
fluid flow locally, we can build a model of how
that tumor cell actually grows into a tumor, and we
can watch the whole thing as it happens. So this is just an example
of putting prostate tumor cells into the liver. And so this is the silicon
chip, this is the tissue, and you can just see it there. And if we look a
couple of weeks later, we’ve just put a
few tumor cells in. They’ve grown into large
tumors that we can see. You can actually see
them with your naked eye. These little white
areas are tumors, and they’ve been growing
very large, because there’s fluid flow through controlled
at the microscale level. And so I’ll finish up here just
showing this is, again, imaging to show the tumor taking over. The liver starts out
as green, and you can see the tumor taking over in
a period of a couple of weeks. The tumor is red in
this case, and it’s being invaded by
connective tissue from the underlying cells. So all of this is great, and I
have a lot of graduate students in the lab who are working very
hard with our prototype system. But clearly, what
we need if we’re going to move this thing
to high-throughput assays, is to adapt this design
and these design concepts into a high-throughput format. So we’re working with a
local pharmaceutical company to do exactly that,
and figure out how we can take this concept
and put it in a multiwell plate format. And this little
movie just shows you how we can use
microfluidics now to replace that whole big reactor system. Now we’ve crunched it down. Actually, this plate
has 24 little reactors. This one’s got five. And you can see the fluid
being pumped through here using microfluidics. And so we’re trying to move
this now out of the lab and actually have
people use it so that we can have physiological
models of humans, and actually put those
transplant surgeons out of business. And even though it’s not
as exciting for the news media, what we
hope is that we’ll be able to develop
drugs better and faster, and cure diseases
that right now, we really don’t have
a great handle on. So the parting
thoughts are that– two big messages. There are some
great applications of tissue engineering
moving into the clinic. They mostly have to do
with connective tissues. It’s very hard to
move into the– go from animals up to people. But that the big future
of tissue engineering, though it sounds a lot
more mundane, not quite as scintillating, is actually to
have a human body on a chip so that we can ultimately
predict how humans will respond and cure diseases. So with that, I’ll close. And I’ll talk to you at
the panel discussion. [APPLAUSE] DOUG VINCENT: Thank you. Let’s see. I’ve used “wow,” so I guess I’ll
go for “spectacular” this time. Just terrific, terrific stuff. And you guys having fun? AUDIENCE: Yes. AUDIENCE: Yeah. DOUG VINCENT: Yeah. This is neat, neat stuff. Professor Angela
Belcher is up next. Now, Angela, the last fall, was
one of MIT’s newest MacArthur Foundation Genius
Grant recipients. Pretty exciting news. Delighted to share that
with you this morning. And she’s going to
spend some time with us now, as she’s working to develop
materials on a nanoscale, and how she’s using
nature as a guide. Angela? Thank you. [APPLAUSE] ANGELA BELCHER: Okay, thanks. Well, thank you very much. I really appreciate
the invitation to come here and speak
to you today and tell you a little bit about
what we’re thinking about at the Biological
Engineering interface. And as Professor Lauffenburger
said in the first talk, that this is a relatively
new discipline. So I came to MIT
because I thought, this is the place where
it’s all going to happen. And it’s a really
exciting place to be. But one thing that I think
is a little bit different is that I think that biological
engineering has been around for quite a long time, and
I brought an example with me today. This is an abalone
shell that was a marine gastropod
that was grown off the coast of Santa Barbara. This is an organism that
has an incredible ability to make materials
at the nanoscale. What it does is it uses
materials from its environment. It uses the chemicals
it has in the ocean to build this really
exquisite structure. This structure is very,
very tough and very strong. It’s 3,000 times tougher than
its geological counterpart. I didn’t tell you, it’s made
out of calcium carbonate, which is basically chalk. But the organism figured
out how to make it in a way that was
incredibly strong. The organism also
figured out a way to make it using
non-toxic processing. And so I think that one of
the keys is to look at nature and how nature has
constructed materials over millions of years, and see
what we can learn from that. And what can we
apply it to, what we like to say, nature hasn’t
had the opportunity to work with yet? During the Precambrian
geological time period, when the chemicals in the
ocean started changing, the organisms had
to learn how to deal with the changes in the
chemicals in the ocean. And being clever,
they actually decided to use it to make materials. So the kinds of materials
that nature has worked with are really limited to
what’s found in the ocean, and we’ll talk about that today. This structure also was
built based on millions of years of evolution. At MIT, the lifespan
of a graduate student is about four to
five years, and so we have to figure out how
to develop materials on the time scale– not geological time
scale, but the time scale of an MIT graduate student. So starting with that, I want to
show my really incredible group that I have at MIT. This is a really
multidisciplinary group. We have students from
Biological Engineering, Biology, Chemistry, Chemical
Engineering, Material Science and Engineering, and Electrical
Engineering and Physics. We’re all working
together to see if we can understand how
nature makes materials and convince organisms to
work with materials that they haven’t already
worked with before. So about a year ago, I was asked
by the National Nanoscience Initiative to say what I
thought the biggest challenges in material science were today. And I took this very seriously,
and so I put together a list of what I thought
the biggest challenges were. Maybe we can actually make
that fit a little better. I said, I like a material that
can actually self assemble. I like a material
that’s self-correcting. So if you have a computer
component and it breaks, it can fix itself– or self-healing. I’d like to have a material that
can grow its own template, that can recycle its own
template, and only goes to a desired length in
diameter, and then stops. I’ll show you a little bit
about how abalone’s shells grow, but the abalone shell–
the structure that makes it so exquisite and
makes this color– you can come see it afterwards–
this color so beautiful– this is the same
color of pearls– is that the thickness of
this inorganic material– the thickness of this chalk is
controlled at the genetic level to be precise. And that precise control gives
it many wonderful properties. So I’d also like to
have a material that’s environmentally benign
and uses environmentally benign precursors and solvents. It would be great to
have a material that is grown at room temperature
and pressure– that generates little waste. The organisms in the
ocean, they don’t produce a lot of toxic
waste, because that would be bad for the organisms. We’d like to span
multiple length scales. For the most part I’m
a self-assembling, self-organizing material that
is organized from the nanoscale to the macro length scale. So we’d like to be able to
take those kinds of properties and be able to apply it to
other kinds of materials. Interface– organic and
inorganic interfaces– interface with biology. But my dream is to
have a material that’s genetically controllable
and genetically tunable. I’d like to have a
DNA sequence that codes for the production of
any kind of material you want. You want a solar cell? Here’s the DNA sequence for it. You want a battery? Here’s the DNA sequence for it. I’d also like one that’s
responsible to external cues, inexpensive, of
course, and scalable. [LAUGHTER] So it’s a pretty long
list, and your response is an appropriate response. But at the same time,
biology has already figured out how to master
many of these properties. And so, us as
material scientists and biological
engineers, we think it’s our job to figure out
how they solve these problems and guide them to work
on other problems. And so what we’d
like to do is order, with genetic control, the atomic
scale to the macroscopic scale to where we can control
atoms and unit cells and lattices to basically
self-assemble devices. And I’ll show you a
couple devices today. I’ll show you the first
virus-assembled nanoelectrode. I’ll show you the first
virus-assembled rechargeable battery as two examples. So as I said before, biology
has already figured out how to do a lot of this. They can communicate
within a cell. They can correct themselves. They have genetic code,
which codes for the synthesis of everything they need. This is a eukaryotic cell, and
this is a prokaryotic cell. We’d like to have the
same kinds of properties. The way that we’re
going to do this is, we’re going
to actually evolve organisms to work with materials
we want them to work with. We’re going to force organisms
to live with semiconductor materials, and force them to
live with electronic materials, so that they can start to
use them and process them. But that’s pretty challenging. When you think about biological
organisms, which are soft, and they’re in aqueous
conditions– they’re in water, and they’re basically
squishy organisms, how are you going to
think about integrating it into a semiconductor fab line? Or something that’s done
under really controlled atmospheric conditions. So there’s robustness questions. How robust is it? Do you have to work with
the conditions under which organisms normally live? And then if you want to
evolve these organisms, how far can you push them? And where do you start? When we first started
thinking about this problem, we thought about
antibodies, which are natural proteins in the
body that bind small molecules. These small molecules
can be organic molecules, or they could be other proteins. And they do it with
an incredible amount of specificity, so these
proteins that are in your body can recognize an addition
of a carbon to a molecule. Or they can recognize
chirality of a molecule. We said, we want that. We want a protein
that can recognize an electronic component–
that can recognize a unit cell of an inorganic material. Here’s another. I never really knew snowflakes
until I moved to New England, but this is an incredible
self-assembling system. You think of this really
exquisite structure. And in the middle,
it has a nucleus that allows for the
nucleation of this really complex structure– really exquisite structure. So what we’d like to
do is have a nucleus that codes for, or tells
an electronic component how to grow. This is a slide that
talks about scale. And so we’re interested in
nanoscale, and I had a– a reporter asked me yesterday. Did you always know you wanted
to be a nanotechnologist? And– [STAMMERS] [LAUGHTER] I can tell you that I haven’t
thought a day about it. What I thought was biology
makes incredible materials. What length scale does biology
choose to grow materials? For the most part, it
chooses the nanoscale. So it’s worked well for biology. Let’s see if we
can harness that. And so here’s a–
this is a length of a typical component
here of a piece of inside of your computer. Now, here’s a drawing– a schematic of the
length scale of certain biological molecules. And so this is a
DNA strand here. This is a ribosome. It’s a molecular
machine that processes all the proteins in your body. And here’s some other
kinds of proteins as well. So we said, if you want
to make things small, why don’t you start
with the building blocks of biological molecules
that can process these? Again, the fact that there’s
these already naturally existing molecular
machines that have evolved to be very specific– this
is our diagram of what a ribosome looks like. Basically, the biological
machine that reads messenger RNA and builds proteins. We said, wouldn’t it
be great if, as it’s going along and reading
its messenger RNA, instead of building a
protein, why don’t you have it build a semiconductor? Is that possible? Can you have each tRNA
molecule that is specific for an amino acid
have a codon that’s specific for
electronic material? And we’ve recently accomplished
that in a relatively small way in the lab this year. Well, there’s many kinds
of naturally occurring biomaterials. This is a coccolithophorid. It’s a unicellular algae. It’s made out of
calcium carbonate. This is an abalone shell, like
the one that I showed before. This is a fracture
of an abalone shell. And looking at it in scanning
electron in a micrograph, you can see these tablets, which
are basically chalk tablets. But they stack on top of
each other to form this brick wall-like structure. Now, proteins that are
coded at the DNA level decide what kind of material to
grow– what thermodynamic phase of this material to
grow, how thick it is, and how it stacks up to make
this really tough structure. This is a diatom. It’s made out of SiO2. I’ll show you a little
bit more about it. It’s basically
made out of glass. And these are
magnetotactic bacteria, which have inside of them small,
single-domain magnets which are used for navigation. They make these really
incredible magnets. All of these are
processed in the ocean, basically at ocean temperatures,
using nontoxic materials, and pressures in the ocean. So what’s the key? How does this work? Well, this is a mantle
epithelial– basically, epithelial tissue from
the abalone shell. An abalone is a gastropod. It basically has a big foot
and a big piece of tissue that comes across here. These cells are the cells
that are pushed right up against the abalone
shell, and their job is to secrete proteins
and to secrete ions have actually builds the shell. So in between this
tissue and this shell is basically a
little reaction vessel that grows the shell. And the key is is that there’s
proteins in these cells that are pumped out into
this space that guide the growth of the
inorganic materials. Here’s a diagram
of how it works. Proteins are made
out of amino acids. Amino acids differ
from each other based on chemical functionality. These particular proteins
are very negatively charged, and their job is to grab calcium
ions out of solution, out of the ocean, and to start to
use them to build structures. And so they do this in
a very exquisite way. The distance between
the negative charges is very important. It causes the distances between
the inorganic material– the chalk, basically. It causes it to grow up
in this particular way. This is just showing
some more pictures. This is the abalone shell here. These are individual tablets. This is the top. This is actually
a part of a pearl that it looks like it hasn’t
grown out its full length. But this is an
organic component. This is the protein that
actually directs it and helps it grow these
kinds of materials. And this is a mechanism. Basically, you have these
proteins in solution. They’re binding calcium,
binding calcium carbonate, starting to build these
nanostructure materials that start to grow up to be a
full shell or a full pearl. And this is just a fun
picture, because this down here is the older shell. And this is the
new components that are just growing, and haven’t
grown out to their full length yet. These are diatoms
made out of silica. This is a scanning
electron micrograph, and this is on the micron scale. But the reason I show you
this is that diatoms are– there’s at least 10,000
different species of diatoms, and they’re classified
based on their morphology. And you can see how
different this morphology is. These are made out of glass,
but they’re all this– they’re different species. And so what it says– they
also have all the same reaction vessel, which is the ocean. So it says that this overall
macroscopic structure of these organisms is
coded at the DNA level. So if you look at the kinds of– here’s this periodic table. And if you look at
the kinds of elements that nature has to
work with, it mostly has calcium in the form of
calcium carbonate, like shells. It has calcium
phosphate, like bones. It has the silicon that
I showed you for silica, and it has iron. But what about the rest
of the periodic table? Why hasn’t nature used
the whole periodic table to build materials out of? And the answer to that our
group is that just, they haven’t had the opportunity yet. Let’s give the
organisms opportunity to work in this part
of the periodic table. Let’s give them the opportunity
to make semiconductors. Let’s give them the
opportunity to make materials for magnetic storage
and materials for batteries. So let’s take the idea of a
protein-protein interaction or protein-antibody interaction,
and let’s combine it with fabrication to make what we
call evolved hybrid materials. And the question is,
how are you going to think about doing that? If we can take proteins
from the shell, and we can extract
them out, and we can start making synthetic
shell in the lab. But how do you figure
out which sequences to use to grow a semiconductor? Because we don’t have
any data for that. You could probably do modeling,
but Professor Lauffenburger says they don’t use a
lot of trial and error. We actually use– sorry, Doug. We use a lot a lot of trial
and error to begin with. And then we take what
we learned from that, and then we try
to make it better. And so we didn’t know how to
come up with these sequences, so we decided to borrow an
idea from the drug discovery industry that’s called phage
display, which uses a virus. This is a virus that
is a non-toxic virus. It actually has
a bacterial host. As a material scientist,
I look at this is an object that’s about one
micron long in this direction, and about six nanometers
across in this direction. And it only has a
couple of genes that code for a couple of proteins. So it’s actually
easy to manipulate. We can do additions
in the DNA sequence that code for the
additions in the protein on one end of the virus. Why do we want to do this? Well, we want to take a billion
possibilities simultaneously as a library, and we want to
search through this library to see if we can
find any chemical groups on the protein that match
the semiconductor of interest. Just like the abalone
shell found a way to grow calcium
carbonate, we want to use a virus because
it’s quick and easy, and it takes only
a couple of weeks to go through this whole
process and use this in our lab. So here’s the idea. The idea is, now
you have a library of viruses that
are all genetically identical to each other. They only differ from each other
based on a small amino acid protein on– a small
protein on each end. Now, in about a one
microliter sample, you can have a billion
different viruses. And you can force them to
interact with a semiconductor wafer. Most of them won’t have
any chemical specificity, and you wash those off. Some will have a
chemical specificity, and you want to keep those. No, they can’t make
copies of themselves. And so you have to infect them
into their bacterial host, which now acts as a factory and
makes a million copies for you. And now you have a population
of these proteins that have some affinity for this surface. Now, we think of us as kind
of a Darwinian process. We want to take that
population and force them to interact again,
and go through this process about five to seven
times, looking for the survival
of the fittest– the one that works best
with a semiconductor. And here’s a movie
that my students made that was on
MSNBC, where you have this population
of viruses, and you’re throwing it at a semiconductor. And you’re looking
for the ones that have a chemical functionality
that matches the material. Most of them don’t, and
they’re washed away. This is also done to music,
but I didn’t play the music. And then you can manipulate
and change the surface charge on these, and remove
them from the surface. And then you can infect them
into their bacterial host. The bacteria now makes a
million copies for you, and then gives them back again. And so this is
our amplification. So now we’ve worked with
about 40 different kinds of materials in my group. And it takes about two to three
weeks to get to the answer– get to the sequence that
can control a material. So we look at a virus
as just a material that has DNA that’s
easy to manipulate, that you can actually
change the DNA and change the
coat of the virus– the major protein
coat of the virus– to have it be
specific for anything you want it to be specific for. Here’s a list of the
kinds of materials we’ve worked with in my group. These are semiconductor
materials over here that we’re looking at
for laser technology. Right now, we’re really
interested in gallium nitride– a laser material that might
be a very efficient material for solar cells. We’ve worked in this part
of the periodic table here for magnetic storage. I’ll show you some
materials over here that we’re using to make
rechargeable batteries. And over here, that we’re
using to make nanoelectrodes. So this is just in a–
this is the real data. This is an example of how
you can select a virus to be specific for something
like a gallium arsenide– a semiconductor wafer. What we did was,
through this process of throwing the
viruses at the material and asking them to bind and
getting rid of the ones that weren’t specific,
we came up with some that bound this material
gallium arsenide, which is a semiconductor material that
biological organisms would have never had to come in contact
with and learn to live with, or to use. And then– So this is a gallium
arsenide wafer. These are 1-micron lines
with 4-micron spaces. And we basically use these
viruses as just a tag to– we put an organic molecule–
a dimolecule on it. We dipped our wafer in
the solution of the virus and we pulled it out. And what you can see is that
it pulled out just the gallium arsenide. It didn’t recognize
the material that it wasn’t evolved to recognize. Well, we’ve been doing–
that work was first published– we first
thought about this in 2000, and we published this. But we’ve got– we want to
go down that rest of the list that I told you, so
we can get an organism to work with a material it
normally doesn’t work with. But so what? We want to grow the material. We want to use it for some
kind of important technology. And this is a work by a
graduate student of mine, Asher, who had this very clever idea. It was a very bold
idea, which was, we’d already shown
that we could select viruses to be specific
for different kinds of semiconductors. But he said, let’s select
viruses to find mistakes in semiconductors. Let’s find them to find
defects in semiconductors. And so what he did was he
took a population of viruses and forced them to interact
with defects in semiconductors– single atomic defects
in semiconductors. And then he negatively
selected them to not recognize non-defects. And so then what
he does is now he has a fluorescent signal
where you can take a device or you can take a wafer, and
you can dip it in your virus. And it can count the
number of mistakes on that, and you can optically
read it out, which is pretty interesting. But now we’re looking
at applying this to other materials. We do a lot of
work with the Army. We’re looking at applying
this to other materials where you might be able to
find defects in instrumentation in the field. So what we want to do is
basically have a solution that you spray onto something,
like the wing of an airplane, where the organism– a virus,
or in this case, it’s a yeast– actually sticks to it and
gives out an optical signal. So if you’re able to have
a biological organism that recognizes a defect,
it’ll light up. And so you can just spray it on. Yes, there’s a defect in this. You need to pull the plane over. No, there’s not
a defect in this. You’re okay to deploy. So in addition, we’re looking
at being able to grow things. We can tag things. We can look for defects. But what if you wanted
to self-assemble a battery from scratch? How are you going to do that? This is some real sequence. This is protein
sequence that comes from third round, fourth round,
and fifth round selection in part of your
evolutionary process. And what you can see–
these are color coded for the functional groups. What I hope you can
see is that there’s a lot of similarities
in the colors of groups. But eventually, as you take
it through additional rounds, the sequences fall out. They come to what’s called
a consensus sequence. These are the best
sequences to use. So before, we
tagged the material, but we’d like to
grow a material. Shells do it. Bones– we do it. Bacteria do it. Here’s our cartoon
of what it would look like if a virus could
grow four semiconductor particles on its head. This was our idea. So our idea was to
have the proteins on the tip of a virus– have a sequence that allows
them to grow a semiconductor. And this is a
low-resolution TEM image of a virus that’s grown a
collection of semiconductor particles called quantum
dots on its head. And this is magnified
about 800,000 times, showing you the lattice– showing you the atomic
structure of that semiconductor. So this is where
we used a virus. Through genetic
engineering and selection, we convinced it to grow
five semiconductor particles on its head. So we thought that
was very interesting, but I’m not sure how
practical that is. And so we decided to go back
into the genome of the virus and do more engineering. So that now, instead of growing
things on just five copies on one end of the
virus, let’s grow them all across the
length of a virus. So let’s make viruses that
are electronic components– semiconductors and metal
materials that could be used for wiring up a device. And here’s a picture. So the idea now is to have
the virus act as a scaffold to grow lots of little
semiconductor particles along its length,
like this picture. And I forgot to mention
to you that we lucked out when we picked this
organism to work with, because the code of the virus,
the major body of the virus, actually self-assembles itself. And it self-assembles itself
in a crystalline manner so that all the proteins are
all crystographically related to each other. So we picked a perfect scaffold
already made by nature. And what we say is,
now let’s manipulate it to do something
we want it to do. Let’s have it grow
semiconductor wires. And this is a special kind
of transmission electron micrograph. This is a one-micron
scale bar here of the first virus-grown
semiconductor wires. This is an elemental
map that shows you that most of the wires, the
inorganic materials, the zinc and the sulfur,
actually map back to this particular
semiconductor. So our idea was that if
you could get a virus to grow lots of
different particles, and you could get it
do it in high packing, could you then get rid of
the biological organism and be left with nice wires
that are all the same length and all the same diameter that
could be used in electronics? And here’s some of our
first examples of those. These are very high-quality
semiconductor wires that are all grown based on the
genetic engineering of a virus. This is a 200-nanometer
length scale here. We’ve been doing a lot
of further engineering. I’ll show you the first
virus-based nanotube. This is a silver-based nanotube. This part here is all
inorganic material. It’s all silver. This part here– this is
where the DNA is in the virus. And the virus actually has the
code to build this silver wire. And we’ve been using
this technology. This is work by a very talented
student of mine, Ki Tae Nam, to see if you can direct
viruses to build electrodes for batteries. And so this is going towards
the first virus-based battery. The thing that we think
is good about this is that we could try to
grow it at room temperature. We could try to use
nontoxic materials. We could make it flexible. And we think that by adding
the nanoscale regularity that nature already
gives you, that we could take advantage of that. And so this was Ki Tae’s
dream about six months ago. And this is the first
virus-based lithium ion rechargeable battery. It was made in our group, and
this was done in collaboration with Professor Paula
Hammond in Chem E and Professor Yet-Ming
Chang in Materials. So over here, you
basically have a flask that has the sequences encoded
to grow the anode material– the electrode for the battery. You grow it up in the lab. You pour in the precursors. The organism grows it. You put it on a surface. And we were able to actually
reach the theoretical capacity for energy density
based on this material, and it’s all made very rapidly. It’s made at room temperature. The thing that’s really nice
is that the viruses actually self-organize
themselves as well. And so we’re looking at this as
a flexible, high-energy density battery that could be
made very inexpensive. We’re also looking at actually
integrating it into textiles. We do a lot of work with Natick
Soldier Center, which actually tries to make textiles
and better uniforms for the soldiers in the field. And a good portion of
the weight of a soldier is actually batteries,
and so we’re seeing, can we make these flexible
batteries into textiles, and weave them into the
soldier’s uniform directly? Now, you can do other
fun things with viruses. Like you can engineer
the head of a virus to bind to the tail of a virus
to make virus-based nanorings. That’s mostly just for fun. And we’ve been working on
trying to assemble these into electronic components–
either nanoelectrodes that are self-assembling. We want to have a sequence that
tells the electronic component where to sit, what to grow,
and how to grow it back in case it gets damaged. And this is a work by Chiang
and Yu Muang in my group that have been trying to figure
out new ways of engineering to have multiple kinds
of functionalities. And here are some examples
of the first virus-based nanoelectrodes grown in our lab. We convinced the
virus to bind gold on this side, gold on this side. So this is six nanometers
in this direction. I covered up the
scale bar, but this must be 500 nanometers across. And then we poured
in the solution and the viruses
grew the gold wires. These are very small
wires, actually, which have all made
contact and have some reasonable
electronic properties. They’re definitely
not perfect yet, but the organism has the
ability to bind the material on this side, bind the
material on this side, and grow the
material in between, all coded at the DNA level. This is going towards our first
virus-based transistor, where we’ve encoded the
organism to bind to two electrodes on each
side, and then encoded it to bind a
semiconductor in between to connect these two electrodes. So this is a semiconductor
material grown between two gold electrodes. I think I have a few
more minutes here. So I’ve reached
some of those goals that I told you that we’d
like to be able to do, but we have a long
way left to go. But the other thing in
picking this organism– again, we’re really lucky in
the fact that biology chooses things to be
actually of defined lengths. If you have an enzyme in
your body that’s too short, if an important
part is cleaved off, then it could be non-functional
and you could have a disease. Well, the same
thing with viruses. They always build their
structures at the same length, and so we can predict that
and we know how they are. But the thing that’s really
nice about having materials all at the same length is that
they can assemble themselves. And so think about taking
a box with crayons, and you throw a
couple of crayons in and you shake the box, and
they’re randomly oriented relative to each other. But the more you throw
in and keep shaking it, they actually start
to self-assemble. Well, viruses do this too. And they actually
make liquid crystals. And so in this case,
what we’ve done is we’ve engineered our virus
to grow a material of interest, and then self-assemble
itself into a liquid crystal for things like liquid
crystal displays. And so here’s the
movie for that. So the idea is through
this genetic selection, you’ve selected
these organisms to be specific to grow
a semiconductor. You put in the precursors,
just like happens in the ocean. They grow the materials. You keep adding more viruses. And they actually start
to assemble themselves into very defined structures. They line up. So virus, semiconductor,
virus, semiconductor, virus, semiconductor. And they form really
exquisite patterns. And if you look at them in a
cross-polarized microscopy, you can see that
they’re actually liquid crystal and materials. The thing that’s
really fascinating about these materials, though– if you take these
liquid crystals, these high
concentrations of viruses and you basically plate
them out on a surface, they actually form films. This is a structure that
looks like a piece of tape. You can pick it up with forceps. And it’s 99% virus,
1% semiconductor. And you can pick
it up with forceps and you can move it around. So we look at this
as maybe having materials that are
like tape rolls, where you could actually roll
out your electronic materials, or roll out different kinds of
materials based on this ability to self-assemble. This is just showing you– it’s interesting to think
of a virus as a unit cell. And these are the unit
cells that actually stack on top of each other. They have really
interesting structures, which we’ll skip over. But you can actually
engineer and put anything you want on
them– a gold, an enzyme, an organic molecule, and
have them self-assemble into these film
kind of structures. I just want to introduce
one more idea, which is looking to nature
again and thinking about how spiders spin silk. Silk is a material that’s
obviously naturally evolved. What happens is
that the organisms– the spiders take a high
concentration of protein, and that’s a liquid
crystal phase. And they push it through a
small hole called the spinneret, and they spin out silk fibers. So our idea was, wow. If you can make liquid
crystals out of viruses, and spiders spin silk
out of liquid crystals, maybe we can spin viruses. And we were able to do that. These are viruses that you can
grow for meters in length that are basically using the same
principles of how spiders spin silk. But we spin viruses. Now, we spin viruses that have
optical materials in them, and magnetic materials,
and semiconductor materials in them, so they could be used
for things like optical fibers. We’re using them for
tags for integrating them into soldier uniforms, and we’re
using them for biodetectors, as well. And this shows you that you can
actually make mats out of them, and we’re trying to use
them for non-woven fabrics. There’s other organisms
you could use, too. We chose viruses. Yeast are incredible materials. We like yeast because we work
with Professor Wittrup, who’s a yeast expert. Because we used the
same yeast that you use to make beer
and bread, and we try to use them to
make semiconductors. And so you know from the beer
industry and the beverage industry that
yeast are scalable. And so we look at yeast as
factories as a possibility of growing materials. And so instead of Budweiser, we
think of Nano-weiser factories. Okay. So I’m out of time, but
I’ll leave you with that. And I thank you
for your attention. [APPLAUSE] DOUG VINCENT: –started
up here again. Thank you for working
quickly back to your seats. So are you guys
still having fun? You ready to hear some– two more of the
outstanding faculty here– AUDIENCE: Can’t hear you. DOUG VINCENT: –at MIT? You can’t hear me? Let’s see. Are we getting any better
with the acoustics? AUDIENCE: No. DOUG VINCENT: No? We can get more volume? They’re saying they can. Oh, there we go. Hello. Okay. I think we’ve got it. We’re going to get going. We have two more presentations,
and we may be a little bit late for lunch. My sense of the
energy in the room is that folks
will, in fact, want to have the panel Q&A. We’ll
gauge that a little bit later. What I’d like to do is get
moving with Professor Ram Sasisekharan, extraordinarily
talented researcher that has won numerous
awards for his creative ways that he’s had some innovation
here in the Life Sciences. Going to talk a little
bit about the emerging field of glycomics. [APPLAUSE] RAM SASISEKHARAN:
Thank you very much. I’d like to begin with saying
that it’s really a pleasure and privilege to be here sharing
with you some of our work here at MIT. And I’d like to specifically say
that the Alumni Association is one that I’ve paid
attention to, because one of the past presidents of
the Alumni Association, Brian Hughes, actually
did in part support some of the areas that I’m going
to briefly touch upon today. So I really know
how important it is in terms of how the
alumni give back to MIT. But having said that, the
title of my talk today is “Glycomics–
the Sweet Science.” I know all of you
took a break right now and you had some sugar. And in the simplest
form, glycomics is the study of sugars. But not the sugars
that you or I are talking about right
now in terms of the one that you put in your coffee. CREW: I’ve got to stop
you here for a second. RAM SASISEKHARAN: Okay. Now, that’s the
energy level at MIT. And so the fascinating
thing is that life has exploited simple sugars
in the most extraordinary way to generate energy that you and
I know the reason why we take a break and get some carbs. But that’s not what
I’m talking about, because I know there’s good
and bad things about carbs, and so it’s not
the low-carb diet. But I’m going to tell you about
how life systems have actually exploited carbohydrates,
going beyond the energy system in some fundamental
ways, that it’s been both a challenge in many
ways in terms of the science. Doug talked about
the two revolutions– the molecular revolution
and the genomic revolution. The field of carbohydrate–
studying carbohydrates to understand its place
in biology was a struggle. But engineering, in many ways,
with the help of technology, has demystified that. And part of what I’m
going to do today is to share with you some of
those exciting technological innovations that have
really opened up this field. So it’s not about Atkins diet. So I’d like to
connect to a point that Doug made earlier,
which is, much of biology then focused on how you
could look at a cell, or how a cell regulates
variety of cellular processes. And it was driven by the central
dogma, the Bible of life, how the blueprint DNA makes RNA. RNA makes protein,
and everything was beautiful with life. And technology, in terms
of ability to sequence DNA and proteins, and the
recombinant DNA technology that you just heard Angie
Belcher talk about– how you can manipulate
DNA to basically create any kinds of systems
that will enable you to address in a cell. And historically, everything
outside the cell– and you heard some of my colleagues
refer to this word “extracellular matrix.” In very simple terms, you
can view that as a junkyard, or the backyard of the house. You throw all the garbage. You ignore it. Sometimes you have somebody
come and take care of it. The outside of the
cell is no different. And in large part,
what it is made up of is actually these complex
glycans, complex carbohydrates, where the simple sugars
are taken further into much higher
level of complexity than I’m going to get to. And in many ways, how the
cells perceive its environment, regulates signal, so
that the cells now know they need to
divide, migrate, or die, is controlled by this
external environment. And the biology
then ignored this, and many tools that were
developed to study proteins in DNA actually were developed
in a way that you could butcher the sugars, because that
was the stuff that you wanted to get rid
of, because you wanted to access things here. Things have changed now. We’re now into a
very exciting area of how we’re trying to take a
more systems or an integrated view. It’s no longer the
cell in isolation. Cells are in the context
of this microenvironment. There’s the cell, the
extracellular matrix, how the cells come together to
form a tissue so that you have a liver cell that’s coming
to form a liver tissue, or the lung cells coming
to form as a lung tissue. And each of these cells
package themselves in unique and different ways
to give you the property associated with the tissue. And in that is this
very important space– the extracellular
environment that brings these pieces together. And the polysaccharides,
or the carbohydrates, hydrate this environment
so a liberalized variety of signaling molecules
that come from the outside and then send
signals to the cells. In many ways, the
reason why we’re able to access
these things are to several high-throughput
technologies, such as the genomics
and proteomics. It’s not looking at
one single component. It’s looking at the
component in a concerted way so that you can not only
understand how a cell behaves, but how the cells behave in
the context of the tissue and the microenvironment,
to give you the property of the organ
or the organ system leading the organisms. And in many ways, we did
need the help of genetics– the whole organism genetics,
once you knocked out a gene or knocked in a gene– to really understand how that
particular gene transmitted information, not only in
the context of a cell, but all the way into the
tissue to the organism level so as to appreciate
the complexity of biology. And with regard to glycomics,
the study of sugar, this became a very
important piece. Because when you knocked
out genes that made these carbohydrates, at the
single cellular level, you cannot not really
demystify that. We needed whole
organism genetics. And what was very
interesting, historically, is that when people studied,
whether it’s a fly or a mouse or any other sort of
animal model system, that when they saw some
very strange phenotypes– the wings in the wrong place,
the eyes in the wrong place, or the wrong kinds of limbs,
the lengths of the limbs, so on and so forth– they eventually found
out that several of the genes that were
affecting these processes going beyond the cell into the
tissue level actually were related to enzymes that
made these complex sugars. So in very simple
terms, the way I’d look at it as, you can view
every– every cell that is in our body has a
sugar coat on them. It’s like the way
we wear jackets. In winter, we wear a fur
coat or a thick jacket. In summer, we wear a t-shirt. In simple way, the
way you could look at how a cell functions is
that the sugar coat on the cell dramatically influences
how the cell behaves and how the cellular
processes are regulated hence. So I want to be able
to go back and put it in the context of the– am I okay? CREW: Yeah. One second. RAM SASISEKHARAN:
So sorry about that. So I want to be able to go
back to the molecular picture– the molecular revolution that
Doug talked about, and put sugars in context,
and then take you through this journey
of how it translates to the more complex systems. As I said earlier,
the dogma was that DNA made RNA made proteins
that truly dictated the principles of life. We now know that’s no longer
the only part of the story. These proteins are
seriously and heavily decorated by these sugars,
or glycans, that dramatically influence the behavior and the
properties of these proteins. So in many ways,
what you can get is a huge amount of functional
diversity in terms of this one gene to one protein. And with this diversity,
you’re able to take a small set of proteins and
get an extraordinary amount of functional possibilities. So if you take one
particular protein system– this is erythropoietin,
which is actually a very important
biotech drug which is used for anemic patients. This particular protein has
sugars decorated on them so that one
particular protein has greater than 150 unique
sugar structures that can give you a plethora
of biological properties. And part of what
we’re trying to do is to truly demystify
that– understand how do these distinct sugar
structures on this protein influence function. And therefore, get to the heart
of what we call as structure function relationships. So in many ways, the
dogma has been revisited. Because if you truly get from
the genotype to the phenotype, we need to really
understand what are the principles of
these sugars in terms of how they regulate the
functions of these proteins. Therefore, we can understand
biology in the truest sense. So if I were to use
one slide to drive home the point about the importance
of glycans, it would be this. Where the only
difference, in many ways, has been the kinds
of sugars that are displayed on these
cells that I talked about, to just give you an
opposite phenotype in terms of the same locus. So this is a very important
field, as you can see, but it has had its challenges. In large part, it’s
because of the fact that there have
been lack of tools to study these molecules–
the point that I made earlier. Several of the tools
that we developed was actually developed in
terms of getting rid of sugars. So in many ways, sugars
were a hassle to deal with. They were more
this inert material that was outside the cell. Probably was, by
and large, useless. But we now know that there are
numerous functions in search of what I call a structure or
sequence activity relationship. In a large part, it
is because of the fact that there are these
heterogeneous, polydispersed, but highly information
contained material that has really stymied the field. And in very simple
terms, I can say that they come in two flavors. They are these linear sugars and
then these sugars that branch. And that structural
diversity truly enables us to get
this very diverse set of functions that are possible. And I’ll give you some specific
examples to go through that. Another important point, to
put it in context with regard to DNA and proteins, is
that you can amplify DNA and you can amplify proteins. You saw my colleague,
Dr. Belcher, talk about how you can
basically take a virus and get the bacteria
to copy many numbers, but that’s not the
case with sugars. You cannot amplify them. So what you isolate them
from these cells are tissues, is what you have. So you really need
innovation that enables you to access this kind
of low amounts of material. And keep in mind, they’re
not only low amounts, but you have all this
polydispersed, highly complex set of structures. Another thing that
complicate the analysis is that, unlike DNA and proteins
where there’s a template– there’s a reading
frame that enables you to make these copies– they’re a non-template
complex biosynthesis. A bunch of enzymes come
together and decorate the cells or the proteins
with these sugars. So in many ways, it
complicates our ability to understand how some of the
structure activity relationship truly drive the
biological processes. And if I were to
step back and put it in the context of a
broader framework, you can view sugar as
a regulating function in an analog fashion. That is, it’s not a switch of
a turn on and off of a light. It’s like a rheostat. It’s like a dimmer,
where there’s a plethora of
structures that give you a diversity of functions. And when I go through
these examples, I’ll illustrate what that
analog function means. But the important
thing is the fact that, with the complexity
and the way they are made, the lack of a template
for biosynthesis, you truly need a more complex
systems approach, because it’s right at the
cell-tissue interface, and it’s not one structure. It’s a set of structures, or
an ensemble of structures, that I’ll come to in a minute. So it is a challenging problem. But nonetheless, an
exciting one from the point of view of how a technologist
or an engineer will look at it. So I just would like to very
quickly introduce the two families at a higher level– the branch sugars. They’re very complex
structures with a core, with several radiating chains. Virtually all
therapeutic proteins– about 80% to 90% of them have
sugars attached to them– several proteins in our body. By and large, the reason
is that you can view this as the most extensive
posttranslational modification that a protein undergoes. And in many ways, they perform
the function of recognition– how two proteins can interact
the way the sugars regulate that, the way the proteins fold,
the way you target proteins. How does an EPO know, once
it’s made in the bone marrow, to go to the brain? Or how do proteins know
where they need to go? What kinds of
information are there? There’s a lot of
information that’s coded in the glycans that tell
them how they get targeted. And very important part–
when you move away from a cell and look at it in the context of
an organism, is the half-life. How long is the
drug in the blood? How long is the particular
protein in the blood? The half-life of these
molecules are very dramatically regulated by the sugars
that are attached to them. And obviously, so
on and so forth in terms of stability
and specificity. So I’d like to very quickly
introduce, using the schema, the other family of sugars,
which are the linear sugars. These are the sugars
that virtually coat every eukaryotic human cell. And here’s the schema of a cell. These ant-like structures
are the proteins to which the sugar chains radiate. Our own president has worked
on some of these proteins in the nervous system in
terms of how the neurons grow. These are fundamental
protein molecules, as I said, that regulate
how extracellular molecules or signaling molecules bind to
cognate receptors on the cell surface so that they get the
necessary information in terms of what happens. So in many ways,
you could view them as they’re found on
the cell ECM interface. They act as a reservoir to
store, sequester, and present several of these signaling
molecules to the cells so that the cells
know what to do. And also, when you zoom in
to these ant-like structures, here are the sugar
chains that are attached. They modulate the
signaling molecules in a very specific way. In other words, you can
have these long chains to which proteins attach. And that’s presented to a
surface receptor on the cell to initiate signals or not. And what’s fascinating is a
point that I’ll come back to, is several pathogens– viruses, bacteria,
protozoans– use these glycans on the cell surface
in our bodies as a way to infect and
achieve these troposomes. How a flu virus infects–
in a very specific way, it affects the upper airways
as against hepatitis, which affects the liver
cells, for instance. So I’ll touch upon that briefly. But this is a high
level introduction to how the linear
sugars look like. And if you actually
zoom in, the way it’s able to perform these
diverse functions in terms of regulating how a signaling
molecule binds to these sugars to be presented to
a cognate receptor, is the fact that there’s
numerous sequences that are possible. As an example, when you
have four bases that make DNA and 20 amino
acids that make proteins, there are about 48
different building blocks that gets made in
a linear fashion to give you this complexity. Therefore, you can
get the diversity that you need to bind to
a whole host of proteins to mediate the
signaling processes. So what I’m going to do
is to, through an example, give you the idea of, when
you look at these structures– I said there are these
ensemble of structures that are present– how do you look at this problem? And how do you demystify
the sugar structure sequence through the function? And I’d like to use one
specific example that’s going to be relevant,
but then I’ll go through a variety
of applications. So with regard to
the technologies to understand how sugars
look, if you will, on the surface of a
protein or a cell, first you need to understand
what does ensemble means, and how do you go
about doing that? And this one particular,
very relevant, practical problem that basically
frames this question, which is several biotech drugs. After the small molecule
pharmaceutical drugs, such as the statins, which are
about 11 Angstroms in size, several of the
protein-based drugs, like erythropoietin, insulin,
human growth hormone, and so on, which
have been a product of the biotech revolutions,
are pretty complex proteins. But what’s fascinating
is that when you look at the top 10 proteins
that are used therapeutically in the clinic, nine of the
10 are glycosolated proteins. So the sugars on these
molecules dramatically influence the biological
properties of these. And it’s expected that these
protein-based therapeutics are going to be the mainstay of the
pharmaceutical biotech kinds of drugs that are
going to come out. So to name a few, such as
the Epogen and Factor VIII and so on. And part of the
challenge that we took on was to try to understand, how do
we take this to the next level? We looked at the proteins. Now, these
glycoproteins on sugars, and how do they affect
the biological functions of these molecules? And I just want to put it
in the context of the story that I said earlier, is that
sugars have been around, and historically, people
have viewed it as something that it’s a hassle to deal with. Get rid of it. In fact, one of the reasons
that I stayed on here at MIT– actually graduated
from the Harvard MIT division of Health Sciences
and Technology program. But the energy level
that was there at MIT, I was basically told not
to work in this area. They said this is carbohydrates. This is the last thing
you want to touch. It’s the most
difficult things to do. But a point that
I’ll come back to, I drew upon the energy that
was in this environment. I actually played tennis
with a chemical engineer. And it was the fusion of the
biology and the engineering in terms of the way we were
thinking about the problem that enabled us to tackle of how
to sequence sugars in a way that people were able to
sequence DNA and proteins so that we could really open
the Pandora’s box, if you will, in trying to understand
what sugars do. So with proteins, the
mindset that people had when I started working
in this area is, forget it. Just get rid of it. It’s not important,
because it’s useless. But now we know that
they are not impurities. They affect several important
properties, particularly the way these drugs behave. So the question to me then
was– this is very fascinating. How do you look at the system? How do you look at this
ensemble of structures? Because it is the polydispersed,
heterogeneous ensemble of structures that
we need to visualize. We can visualize DNA. We can visualize proteins. But how do we look at the
sugar coat on a protein or in a cell to really
understand what’s going on? And I’d like to use this
analogy to illustrate the idea. In many ways, looking at sugars
on a protein or in a cell is like looking at
image processing. It’s looking at, how do
you look at an image? And to me, the analogy that
I’ve used as the six blind men and the elephant. Everybody was trying to
figure out what this was, and each person got it wrong. But the idea was,
you really need to understand what that
whole picture looked like. And here is the elephant. So what we did was, we brought
in several different orthogonal analytical techniques. Each technique
had its strengths, but there are limitations
or weaknesses in terms of studying sugar, because
you’re looking, as I said, as an ensemble or a
set of structures. Then we’d make measurements
associated with these. Then if you do not
link these measurements in a meaningful way, and allow
this to just slap on raw data during image processing,
if the data doesn’t talk to each other, you
get a blurry picture. The more you’re
able to integrate the various measurements,
you’re able to then get, there’s an entire picture. This is the concept behind it. What would we really do is we
took a variety of different analytical techniques–
mass spectrometry, NMR, and different kinds of other
analytical techniques– exploited the strengths
of these techniques, understood the
limitations and weaknesses of those in such a
fashion that we then developed a
number-based approach to code information associated
with the various building blocks. Why is this important? The point that I made earlier. I really needed an electrical
engineer, a chemical engineer, a biologist, a
carbohydrate chemist as a team to really take
on this problem in a very interdisciplinary fashion. And what we were
then able to do is solve this as a
puzzle, where we were able to look at not
just one structure, but a set of structures that
existed either on a protein or on a cell. And we were able to
build beyond that. And once we got the
sugar structures, we eventually linked
it to the genome and the proteome database. So what we’re now able to do– the example that I used
earlier about EPO– here’s the EPO molecule with
the four sugar sites in it. And I said earlier that
they affect folding, they affect the way you
look at the half-life of it. What’s also interesting
is, if you let these cells recombinantly making
EPO, the human cells that make even process
condition dramatically affect the kinds of sugars
that the EPO molecules display on their surface. So part of our approach is
to take a very integrated way of understanding, what are
the different sugars that attach to what sites
of these proteins? How much are they
present [INAUDIBLE] that is their abundance
and their location? You can begin to see
one equal molecule now has been translated into these
150 different unique structures with the sugars that are
attached, and what they do. So the next question
obviously is going to be, once you
have this information, what are you going to do? How do you understand
the biology? And I’ll come back
to that in a minute. So this is giving you a
flavor with regard to how we did this for the branch sugars. Then we also took on the problem
of trying to demystify existing sugar drugs in the market. One that you might
know– heparin and low-molecular-weight
heparin. It’s the drug of choice
as an anticoagulant, especially for things like
deep vein thrombosis, which is DVT, where you want to
make sure that you don’t clot. But what’s also
interesting is, this is the drug that’s used as
the mainstay for anybody who goes into the hospital
with a chest pain. Anything to do with
acute coronary syndrome, in terms of ACS. But what’s fascinating
is that this drug, though it’s been around
for the last 85 years, it’s the [INAUDIBLE]. Essentially, it’s a complex
mixture that’s been so poorly characterized, it’s– the analogy is like what
was the case with insulin in the early ’30s. When insulin was isolated
from the pancreatic organ from a pig, it was
always used as a mixture. And then people saw
this beneficial effect, and it was a
serendipitous discovery. The story with heparins and
low-molecular-weight heparins is no different. And what’s fascinating is,
until we approached this problem in using the
technology, as we said, it was still viewed as
this complex mixture. It had these very interesting
set of biological activities. But the fact remained
that this was a big product in the
pharmaceutical industry with poor understanding. The FDA really did not know
how to look at the system and how to really understand
the complex set of structures correlating to its function. And again, to put
it into context, here’s a small molecule drug,
like your statins or your Cox 2 inhibitors and so on. Just to put it in context,
here is your heparins and low-molecular–weight
heparins. There are thousands of chains
with many, many structures and sequences that really
pose this challenge in terms of how you understand that. So part of what we did was– the proof of the pudding was
in the eating in the sense of, could we take the technology
that we developed, do the so-called
image processing, and try to be able to
give you a picture of what this complex
mixture looked like? Therefore, you can begin to
understand how these structures correlate to functions. So what I’m now going to do
is change gears and walk you through some applications. We took these set of different
technologies we developed. I’m going to walk you through
a handful of very specific examples of what we did. And to put it in
context with regard to why sugars are important,
going beyond the energy molecule, that is,
that I talked about. So one interesting
thing we found was, once we could start
sequencing these sugars, we could understand what
was going on with regard to the sugar structures. Is that tumor cells,
on their surface, contain sugar
sequences that either kept the tumors dormant
so that they were not rapidly dividing–
the point that I had made earlier
that how they respond to the environment with
regard to signaling. Or they could displace
sugar structures that actually
enhance their ability to respond to growth
signals, and then rapidly start dividing. And it was in balance. So the red sugar sequences
kept these cells quiet and would not permit signals
to go into the cells. But the blue sequences
could activate signals and enable these tumor
cells to rapidly divide. And this balance was critical. And then what happens is that
during the process of tumor progression, you get more
of the green stuff growing on these cells, so that
they can more rapidly respond to these
signaling molecules and take off in terms
of their growth. So part of what we
then began doing is, once we began to understand
the biology behind it, we began generating
various of these sugar fragments to try to
really understand, could we then take
these red guys, understand them in
terms of mechanisms not only at the cellular level,
but in terms of animal models? And we have generated
various different kinds of sugar compounds,
and a long list of these, one of them which
is actually plant-derived that I talked about Brian Hughes
being very excited to support. And the other ones that are in
the general heparin families, and they are going under
phase I testing right now. The fact that sugars
change on these cells then begs the
question– could this be a marker for distinct states
that a particular cell type is in? So here’s a set
of healthy cells. But then as you progress
towards the disease– in this case, it’s cancer. We now know– a simple
example I can give you is PSA for prostate cancer. We all know that it’s a
very controversial thing. Just looking at the
protein levels of PSA doesn’t really tell
you what stage you are in in this cancer progression. What we found, and we and
others did, was the fact that the sugars on PSA actually
subtly change and give you a beta handle or a signature
to the distinct state that you are in during the
progression of prostate cancer. So what we’re now
doing is we’re dealing with several human
samples to really try to translate what
we found in the lab to something
clinically meaningful. And there’s a whole
diagnostic platform that’s being developed
to really harness or leverage and
understand, could sugars be a better marker? Because there are enough
evidences to support that. The sugar structures
change much more subtly than those analog
regulation of function that I talked about earlier. One of the fascinating
things is the point that I made earlier about
how pathogens infect. So if you look at
epithelial cells, these are the cell
lines that line our upper airways, our gut,
the blood-brain barrier. These are the cells that
are exposed to the outside, and the way the body
sees the outside. And here’s a typical
epithelial cell. It’s polarized. It has a head and a tail. Very simple. It’s an apical side
and a basolateral side. And here’s the blood vessel. Normally, molecules
are transported across in an active way. There’s a receptor
that binds and then things are transported
through what’s called as a transcellular transport. What’s fascinating is that
several normal biological processes exploit what’s called
as a paracellular transport. Between cells, there
are these junctions that hold these cells together. And then solutes and proteins
and others from this side go into the other
side, and there’s a way that there’s a gatekeeping
mechanism going on. What we found is the fact that
the apical glycans in a very specific way regulate whether
these junctions open or shut, and therefore what goes
in or what doesn’t. In many ways, the analogy is
like the Alibaba and the 40 thieves. You need the right card to
swipe, and if you do that, you have access in. And what’s even
fascinating is the fact that these glycans can be
used as ways to open and close these junctions. But pathogens– several of
these not only bind to these glycans, but use it as a
way to mimic the system and open these functions as
a way to get into the system, if you will. So we’re beginning to see
several bacterial and virus species actually use the
sugars on the cell surface as a way to get entry into
distinct cellular compartments. And you can begin
to see that then you have these glycans which are
different for the upper airways and the blood-brain
barrier and the low gut. The kinds of sugars that
are there in these cells uniquely regulate what
gets accessed in and not, and how the access is regulated. So this is a pretty
exciting area. So the point that I
made earlier that, when you have these sugars
that regulate proteins on the outside, they store
them, they release them, the kinetics become
important– what we began to do is to address the issue
of combination therapy. We said, when you look at a
cell– here’s a tumor cell, there’s the blood
vessels around them. The endothelial cells
grows in one way. The tumor cells grow in one. The sugars are different. The way they regulate this
compartment is different. So if you really want
to understand what’s going on as tumors grow with
regard to the blood vessels, then you need to understand how
these two compartments relate to each other. And if you take
off-the-shelf drugs– off-the-shelf
chemotherapy drugs, off-the-shelf
anti-angiogenesis drug. And so just combining
them in the standard way of mixing A and B together,
part of our thinking was, if there is a very
important component of how these sugar molecules regulate
the entire microenvironment in very distinct ways, then
drugs need to be delivered, or drugs act in similar ways. To cut a long story short,
what we then decided to do was took off-the-shelf
drugs and decided to develop two compartments. We call this delivery
device a nanocell. There’s an inner core
and an outer core. The inner core has the
chemotherapeutic drug and the outer core has the
anti-angiogenesis drug. In many ways, like the
smart bomb approach, where you want to put the
chemotherapeutic agent inside so that it doesn’t escape,
and the anti-angiogenesis. And The most important
thing is the fact that you need to regulate
the kinetics of the release of these two compartments
in distinct ways which reflect back to the
way these cells grow in terms of how the
tumor cells grow and the endothelial cells grow. So cutting a long
story short– and I don’t want to give you too
much of a data dump here. But what we found
is when we looked at the ability for
these tumor cells to grow in animals, the
number of days here, and compared it with
not only the controls, but the various just physically
mixed combination therapy– it’s like taking
just a chemotherapy, and an anti-angiogenesis
as two different regimens. But then doing it in this
controlled release fashion, so that you’re affecting
the kinetics, really. What we found was some
very dramatic effects in terms of the
survival outcomes. This is the normal, so the
controls drop dead in 20 days. The standard current
combination therapies of taking combined
effects give you a little bit extra survival. But a strategy where
you’re beginning to look at this spatial temporal
release kinetics and the way cells come together to
form tissues and organs, and how these glycans
conceptually play a role, you can then begin to
really not only tease apart the system in an integrated
way, but you could leverage it in terms of these kinds of
drug delivery strategies. And where we’re
basically taking this is trying to figure out how
we can use combination therapy in the area of treating stroke. In the interest of
time, I’m not going to go through that,
but what I wanted to show you is the fact that
we are really leveraging this to the extent that we possibly
can, in terms of novel approaches to
treatment of diseases. The last point that I
want to quickly touch upon is how this field has integrated
itself to the genomics and proteomics area. As I said, since glycans
are very important, we now have large
high-throughput data in the area of glycomics. MIT is one of the centers
for an international effort for the consor– it’s called
the Consortium for Functional Glycomics, where we basically
house the largest datasets that’s there in this field. We have all the different
glycan structures. We go all the way
from the molecule to the mouse, or the
molecule of the human, and correlate the link
between the genome, proteome, and glycome, and
generate different kinds of data, whether it’s data from
the cells to the target organ systems. And then our goal
was to really be able to use relational
databases and use a simple port, like a molecule page that truly
integrates these diverse kinds of information. So I think this field– this is emerging to be
an important one where it has had its challenges
from the molecular revolution the genomic revolution– has been possible because
of the technology. And the way that we have looked
at this, the more systems fashion, but it has
begun to address some of the several
complex and challenging issues that we have had. So the summary. It’s an emerging field. Fundamental roles in
biological processes. We’re limited by several
tools that were really needed. Two kinds– the
branched and the linear. Complex and information dense. And we’ve developed several
different tools to address, at the end of day, how
structure correlate to activity and function
with many different exciting applications. And last but not the
least, funding from NIH. And this is my group. Several different
members of my group were obviously
responsible for this, but I did touch upon some of the
collaborators that we have had. And the question of “so what?” Could we really take
this and go beyond? What we were able
to do was really leverage these
different technologies that were developed
at MIT and take on looking at
these glycoproteins and low-molecular-weight
heparin. And a company has been
spun off that has really taken this problem in terms of
trying to really translate it into the real world. Thank you very much. [APPLAUSE] DOUG VINCENT: I
think you guys can see why we had so much fun
putting this panel together. I’m running out of adjectives. Holy cats? Something like that. This is not science fiction. This is just hardcore science
and engineering taking place right here. I’d like to introduce
to you Martha Gray. Martha is the director
of the HST Program– the Health, Science,
and Technology Program– a joint alliance that
has existed for the last 35 years between Harvard and MIT. In that role as
director, she is part of and helped guiding some
marvelous research that’s taking place there, to take much
of what we’ve seen here today and to help translate that
into medical practice. So– Martha? [APPLAUSE] MARTHA GRAY: He was just
fiddling with my mic. Can you hear me? Sound all right? So this is most likely what
George’s coronary artery looked like when he was
admitted to the hospital. And when you consider that
where that dotted line is is where the edge of the
artery normally should be, I think you can appreciate that
this extra tissue inside there most likely caused a reduction
in the amount of blood that could flow
through that artery. If I can figure out
how to turn this on. And that was the ultimate
cause of his symptoms. Now, what I remember
most about George is not what his diseased
coronary artery looked like. And it’s not even that both his
father and his elder brother died of sudden cardiac death. What I remember is this
big, almost jolly guy with a very deep, booming
voice who was willing– excited, even– to be
a guinea pig for me while I was learning to
do a history and physical. Now, I remember that history
and physical so well. It took me more than two hours. I’d like to think it’s not
because I was new and slow, but because he was so
busy asking me questions. But I met him when I was
doing my clinical experience at Mount Auburn hospital while
I was a graduate student at MIT. I went to see George the
day he left the hospital, and he handed me a card with
his address on it and said, you need to let me know
what happens over the years. He was very taken with the
idea of putting engineers in a clinical setting– having them have hands-on
experience in medicine. And he thought that “you
techies,” you guys at MIT, will figure out what to do
for guys with hearts like his. Now, I’m embarrassed
to say, I have no idea where that card is now. And it’s also the case that
I haven’t done anything directly to help guys
with hearts like his. But many of our
faculty and students have, and I wanted to share
some of those stories with you, as well as some other stories. And I’ll dedicate
these to George, and if I knew where
he was, he could come. This would be the card. So the therapeutic
concept in his case is really conceptually
straightforward. You remove the blockage,
and the blood flows. Now in practice, it
doesn’t work like that. Now I’ve got to figure out
how to use this pointer. Can somebody help me? Well, I’m sure there’s
a button here somewhere. Ah, here it is. A day of sitting
here, I found it. [LAUGHS] The nerves get you everywhere. Okay, so you probably
have the idea by now. There’s a catheter that’s
threaded up through the groin to the heart. And in the region of the
lesion or of the blockage, which you can see there in
yellow, a balloon is expanded. And in this particular instance,
there’s a metal casing– a wire mesh around that balloon. It’s called a stent. And that holds the vessel open. And George’s era, actually,
was before the days of these stents. And it was in the
early days of what’s called balloon angioplasty,
where just the balloon was put in place. Though sometimes there
was the problem of it re-collapsing on itself. Now, the good news is that
this Roto-Rooter approach works pretty well. Initially, because
of that collapse, the balloon would fail
under 10% the time. And with stents, it’s
really relatively rare. The bad news is, in
about one to six months, these things both fail. And with by failure here, I
mean needs another angioplasty or needs a coronary
artery bypass graft. So some of the work
I’m going to describe is work of Professor Elazer
Edelman, who is an alum. And he focused on, why is
it that those vessels become blocked again after
they’re opened? And can the procedure and
the devices be improved? Now, if you look at a vessel
after this six month period– this is one such example. This is the edge of the artery. These little black
marks in this example are where the stent
was, and you can see there’s this massive
regrowth into the lumen. And that, again,
blocks blood vessels. And the thinking
at the time was, this response was
really due to the fact that you’ve taken this
diseased artery, blown it open, all kinds of biological
things happen, and that’s it. He said, well, that’s
probably all true, but what about the stent itself? And could that play a role
in why these things fail? So these are the first
generation stents and the so-called slotted tube– or that’s his name for them. And if you put
those in an artery and look at the
lining of the vessel, 60% of the cells
that line that vessel were scraped off, if
you will, or denuded from the vessel wall. So that can’t be good. And the possible reasons
for them are things like, the stent, as it expanded,
the way it was constructed, it would shorten. And you can see, it
buckles a bit here. And also, there’s balloon
contact with the vessel wall. And any of those, and
perhaps more things, could induce the
biological response that led to restenosis. Fast forward some years
with some new designs that were developed based on
some data I won’t show you. The corrugated ring
design differs in that there is no
shortening, no buckling– or at least, not measurable
buckling– and much, much less balloon contact, though it
has exactly the same amount of wire. Now, the biological measures
are what you need to look at. And if you compare, or if he
compared, the slotted tube to the corrugated
ring on every measure, and these are four
of them, there was a marked improvement
with the corrugated ring, like the denudation, the
clotting, and so forth. These were the first data to
show that the stent design itself, never mind the
disease and everything else– the stent
design itself played an important role in the
response to this therapy. It also changed the
intellectual property or commercial landscape,
because knowing that the stent design mattered
gave companies a different axis with which to compete. Now, you’ve probably
heard recently, and hopefully not experienced
directly, drug eluting stents. And there, the concept is– well, another way in which you
could prevent this regrowth is to put a drug on the stent. Use a drug that prevents
the proliferation of those cells that grow in. And looking at
some clinical data, people have tried this
now for a number of drugs. And without naming the drugs,
if you look at clinical data and compare the restenosis
rate relative to a bare stent, you can see that some
drugs work and some don’t. The thought is, well,
this is the difference in biological activity. It’s the difference
in what happens in clearance in the
bloodstream and so forth. These three drugs are
equally effective in vitro– in the test tube, if you will,
at preventing proliferation. Again, that was our
challenge to the paradigm. It’s like, well all
that might be true, but we should look at really,
where does that drug go? What’s it interacting with? And maybe we can
get some insight into how to best both
build these things and tell a priori whether
these devices will work. And so this is one example of
one drug that’s fluorescently labeled, and you can
see at this time point that the drug is very close
to where the stent is. And I’m not showing
you, you can also look at the depth of the tissue. And it turns out, if you
look at all this data– and again I’m
collapsing many years– that there’s a
good correspondence between the properties of
the drug and its interaction with extracellular matrix– that stuff that Ram
was talking about– and the transport of that drug
into the depths of the tissue that corresponds very well
with the clinical outcome. So over the past 20 years,
30 years since George, there’s been a dramatic
improvement in therapies. That really came about from
challenging the paradigm and understanding the tissue
device and drug interaction. That’s restoring blood flow. In the realm of treating
cardiac problems, another basic
therapeutic concept is maintaining
electrical signaling, or restoring
electrical signaling. And one exemplar of this
issue is the problem of sudden cardiac death. Out of every seven
of you, one will die of sudden cardiac death. That’s more than
will die of cancer. And the cause of that death is
arrhythmia in the ventricles. You might have underlying
cardiac disease, but what kills you is
the arrhythmia that prevents the heart
from beating properly. There is a treatment
that’s extraordinarily effective– an
implantable defibrillator. The problem is, who do you
give the defibrillator to? If you gave it to everybody who
had cardiac risk, only 4% or so would actually need it, and
that’s a very invasive therapy for that kind of population. And I should say that the– as the slide says, sometimes
the first indication of cardiac disease
is sudden death. And obviously, it’s
too late by then. [LAUGHTER] So I’m sharing a little bit of
the work of Professor Richard Cohen, who’s been
thinking about this issue in various dimensions
for several decades. He’s also an alum of MIT. And he’s looked at
stochastic behavior of heart muscle and
experimental models, and is really focused on
strategies for predicting the risk of arrhythmia. Now, what you can
actually measure at the surface of the body
is an electrocardiogram. And what that’s detecting
is a whole collection of a whole bunch of small
currents within the heart muscle itself. And if you were to be able to
visualize current in a working muscle, there’s a very
coordinated pattern that allows this heart
to beat effectively. And the question was, can you
extract from this information something that tells you
there’s a little blip over here? And the answer is, you can. I wouldn’t show the example
if it weren’t that true. And at a microvolt level, at a
level that you can’t even see here, it turns out that
there’s something called T-Wave alternans that can show up– a microvolt change if your
heart rate is reasonably high. And this has been tested in
a number of clinical trials, and this is one. And in this group
of patients, those that had a negative T-Wave
alternans test in the 18 months had no significant
cardiac event, whereas a quarter of
those with a positive test had a significant cardiac event. And this is the only
statistically significant risk predictor of a risk of
sudden cardiac death. And this technology has
been commercialized by MIT to a company called
Cambridge Heart. It’s been recently approved
for Medicare funding. And I would point out that
this is one early example of the idea of being
able to predict who you should treat when
the population is very large. It’s the personalized
medicine problem. How do you know how to give
a very effective treatment? Which people should receive it? The third fundamental
therapeutic concept– if blood flow isn’t the
problem, or electrical isn’t the problem, or you
can’t do anything about it, restoring muscle function. And that’s the third
example I want to give, and it’s work of Gordana
Vunak-Novakovic and Lisa Freed. And they’re in the
tissue engineering area, and they’re focused on principle
design of tissue engineering– along the lines that Linda
was talking about earlier. And here’s the
conceptual concept. There’s an area of the
heart that’s damaged and not functioning properly. Could you actually put a
patch on there that works and restore function that way? And the idea was to
take a polymer, seed it in the experiment. I’m going to show
you rat heart cells. They penetrate the polymer. You let them seed and
take hold for a few days, and then put in a
dish and stimulate it with a pacemaker,
actually– a real pacemaker. And so that during the
time it’s developing, it’s contracting in a way
that’s analogous to a heart. And after eight days in
culture, you actually get a tissue that’s beating. These are silicon spacers. Here is the electrode. You can’t see the other
electrode very well. You can get individual
muscle cells. They’ll contract spontaneously. This is the first
time anybody’s been able to create a tissue they
can beat in an organized way. And so this– there’s millions
of cells cutting across there. And if you look at the
[? altered ?] structure, the engineered tissue both
in electron microscope level and a light
microscope level, look very similar
to native tissue. So these have really been
transforming advances in technology– improved
therapy for blood flow that came about from
understanding cell tissue interaction, trying
to predict the risk for electrocardiac events,
and at least as a potential, restoring muscle function. And this was on the MIT
website this February. So if I saw George now
and I had his address, I’d send him a Valentine’s
Day card with this on it and try to describe what
I’ve just described to you. Now what I have worked
on is not the heart, but on problems
related to arthritis. And the work I’ll
describe for you is work I’ve done with another alum,
Professor Deb Burstein. Arthritis is, according
to the Arthritis Foundation, the leading cause
of disability in the United States. One in three of you
probably have it. So very, very common. And the treatment is
really one of two things. One is pain relief,
either by over-the-counter or by prescription
kinds of drugs. And then finally, the
last resort treatment is replacing the joint. Now, it’s easy to say oh,
what a crude treatment. But this was invented
only in the late ’60s, and this has kept many, many
people out of wheelchairs. Now, they last about 20 years. So this has been an
extraordinarily high impact innovation. But again, as Linda
noted, the ideal would be to try to prevent
this from happening so you didn’t even get to that stage. And there are several new
therapies on the horizon. Some are nutraceuticals. Virtually every drug
company I’m aware of has a program in trying to
develop arthritic therapies. There’s new surgical
strategies coming down the pike and that have been
suggested and used. And the challenge
is, so you’ve got this thing that works
in an animal or works in a test tube or a Petri dish. How do you really know if they
work, particularly in humans? And the current way you can
look at either development of disease or progression or
the efficacy of a treatment is x-ray. So this is an x-ray of one of
my student’s patient’s knee joints. And this is the thigh
bone, the femur, the tibia, the calf bone. And you can see the
space between the joints. And if you’re used
to looking at x-rays, you’d say this space
is narrower than usual. But you can’t actually
see cartilage by x-ray, so the fact that
this is narrow means that the cartilage that would
normally be between the joints must be missing. And in fact, this is the
knee of this patient. And this yellow stuff should
be covering the whole joint. You can see it’s been completely
stripped of its cartilage. And I think you can
appreciate how insensitive this kind of approach is. Magnetic resonance imaging has
been a dramatic improvement in the first instance,
because you can actually now see the cartilage tissue. You’re not just looking
at space between bone. And what’s been emerging is
actually molecular imaging of cartilage, looking at,
as I’ll describe briefly in a second, the sugars that
are really functionally very important inside the cartilage. And it gives you information
that you can’t visualize on an anatomic measure. And so if you can actually
see very early stage changes, then you have the
potential to ask questions, like does the treatment work? Who gets OA? How fast does it happen? And so forth. And I wanted to
give an example– one example. Autologous chondrocyte
transplantation popularized in this
country by Genzyme. And the idea is that if
you have a local defect in your cartilage, you go
in surgically, clean it out, remove it, fill it
with cartilage cells, and then cover it with
a piece of tissue. Now, how do you know
if this idea works? What did Genzyme have
to do for the FDA? Well, they had to
ask subjects to agree to what’s called a
second-look biopsy, second-look arthroscopy,
which means they make an incision
in your knee joint, put a light tube in, and
actually put a device in and remove a small
piece of your cartilage. Then they stain
at histologically. You’ve taken it out of the
body, you slice it very thin– six microns thin–
and you stain it. And this purple color,
the darker the purple, the better the cartilage. This is much bigger
than a biopsy. It would be much smaller. Here’s bone. Here’s cartilage. The biopsy wouldn’t go
down to the bone normally. It would just be a
small piece here. This has at least two problems. One is, if you’re doing well as
a patient, the last thing you want is somebody go in and
take a piece of cartilage out that you’ve just tried to fix. And the second thing is
is the sampling problem. If, let’s say, this were it,
where exactly do you sample? And how do you interpret
that single sample from something that’s a
very heterogeneous piece? So it has all kinds of
problems, but that’s the best that had been available. And what we can do now for
the very first time is take– now using the same example. Using imaging, you can
actually get a surrogate for this histology, where the
darker blue is like dark blue here. Light purple turns
red in this diagram. But you can see there’s good
correspondence between this. Yet this did not require
that you take it and slice it into six microns thick sections. This block is a few centimeters
by a few centimeters. And this is a thin
imaging MR slice. So you can do this in vitro. It turns out you can
also do this in vivo, and this is one of the
first images showing that you can make these
measurements in vivo. This is from somebody who had
the autologous chondrocyte transplantation in this
region of the tissue. And two months after she
had had that operation, you can see that there’s tissue
there, but it’s colored red. And here’s the scale. Red means there’s not much
of that important molecule. Yellow and blue
means there is more. And so this region is filled,
but presumably not functional yet. And the x-ray and the
other MRs for this, you could not distinguish
those patterns. For a different patient, one and
1/2 years after the operation, this was the region
of her transplant. And you can’t distinguish
the transplant region from the rest of the tissue. Now, this isn’t to say
that autologous chondrocyte transplantation works. The study wasn’t designed
in a randomized way to follow people over time. That’s ongoing. But what it does say is that
this measurement is actually sensitive to the
kinds of changes you’d want to be able to detect. And a number of drug
companies have now begun to adopt this for their
arthritis clinical trials. And a number of groups
have begun to adopt it, to begin to ask some of the
more fundamental physiology and biology questions,
like who gets arthritis? How does it develop? And I wanted to share
one of those with you from Leif Dahlberg’s
group in Sweden, who adopted this technique
about six years ago. And he looked at subjects who
had complaints of knee pain, but had normal x-rays– normal in every other regard. So each of these bars is a–
think of them as a person, and they’re just
ranked according to the molecular index to that
color that I was showing you. And you can see there’s
a broad distribution. It’s now six years later,
and he’s gone back to them, and they’ve had repeat x-rays. And seven of them
actually have evidence on the x-ray of osteoarthritis. And one of them has already
had a total joint replacement, and they tend to be bunched
towards the low end. So this is early
indication that this may have some predictive
value about what might happen in arthritis. So one of the things that
was eye opening, if you will, about the imaging things
I just showed you– it just allows you to see
things you couldn’t see before, which allows you to
ask questions you couldn’t ask before. And I wanted to highlight
one last set of stories, the Athinoula A. Martinos Center
for Biomedical Imaging that was established about
five years ago in 1999, and their work in brain imaging. And Bruce Rosen is the director
of the Center, Greg Sorensen the Associate Director,
and George Bush is one of the faculty there. Now, about 20 years ago, if you
were to look at– or, 25 years ago– brain images. Again, this was
transformative in its time. You could see the
outline of the brain. This is a cross-section
taken, if you will, that way. You can get a sense
of structures, but I’m sure you can
immediately discern how much more detail
in the kinds of images that we can get today. And just from that
kind of images and the kinds of
computational technologies around a sequence of images
taken through the brain, you can get a three-dimensional
representation. And what our pioneering
advances is you can take this very
convoluted structure, blow it up like a raisin
turning into a grape, and begin to see what
anatomic structures might be next to one another. And again, so you can begin to
ask physiological questions. Or where do the different
pathways in the brain go? And I’m not going to talk
about any use of these for any medical application
or neuroscience application, but I think you can appreciate
the kinds of questions you could ask that you
simply couldn’t ask before. That was focusing on measuring
structures in anatomy. What you’d really love to
be able to do in addition is to be able to tell where
the brain is activated. And this is one of the
pioneering images called functional magnetic resonance
imaging that Bruce Rosen was one of the first inventors of. And this little dot
here shows that there is a difference, after
a particular task, in blood oxygenation. And blood oxygenation
is used as a surrogate for electrical function. So presumably, when you think
you use an energy source and it reduces the
amount of oxygen in the blood–
that’s the concept. So I want to show you how a
functional magnetic resonance, fMRI, study is done, in
running an experiment with you. On the next series of slides,
you’re going to see a word. It’s going to be written in a
font, and that font is colored. And what I want you to do is
yell out as fast as you can– and your time to
lunch depends on this. [LAUGHTER] As fast as you can,
what color the font is. You got it? Okay. AUDIENCE: Green. AUDIENCE: Blue. AUDIENCE: Red. [LAUGHTER] MARTHA GRAY: With
practice, you get better. And I didn’t give you
a chance to practice. That’s called a
Stroop test, and it’s one of the tests
that’s done in looking at people with attention
deficit hyperactivity disorder. [LAUGHTER] I know. My friends tell me that
to be prevalent here, and I didn’t want
to use that joke. But you obviously
thought of it yourself. [LAUGHS] So as you can– quite obviously
from this picture, if you say, which regions of
the brain light up? It’s very, very different
in kids with ADHD. And so again, we’re opening
up a whole new vista. Who knows what that means
biologically or diagnostically? This is early data. But you can see how
it’s just opening up the door to all kinds
of different questions. Now, I said that
measured brain activity. You can do it through
measuring blood oxygenation. Now, brain activity really
happens on a millisecond time scale, not on a
second time scale of MR. Magneto-encepholography
is a technique that allows you to measure electrical
activity of the brain. The problem is, you
can’t localize it. You can’t tell when you measure
its surface of the brain exactly where it came from. And the folks at
the Martinos Center said, well, what if we combine
the best of both worlds? Use the MRI to constrain
the inverse problem. And then, if you will,
make a movie of activation. And this is one example of that. Again, in this case they’re
reading novel words. And when you see
the thing restart, you’ll see it starts at
the posterior of the brain and it moves itself up to the
frontal lobe, back and forth. So you can see this
fleeting thought is very complicated
in terms of going back to front, to back to front. And it gives you a
whole new appreciation for the complexities of
what systems neuroscience is likely to mean when you
think about what does it mean to read a word and remember it. So I’ve given you a
number of vignettes– heart, mind, arthritis,
and of things that have really over the past
20 years gone, if you will, from bench to bedside. They’ve all been
things that I think every single person on that
list has gotten a grant proposal back saying, it’s impossible. It will never work. And I’ve showed you that these
things, in fact, have worked and had tremendous impact. And each of these advances
was enabled by a true advance in technology. Another important
common theme is, though I told you the
head investigator– in fact, if you were to go
into any of their laboratories or talk to them
about their work, they won’t say oh, it was just
my engineering background that did it. Actually, they had
teams of people within the labs that
include physical scientists and engineers, biologists,
physicians, and so forth. So that’s the case for all
of the stories I told you, and they really
attribute the advances I’ve shown you to
the ability to create that integration
and the ratcheting that you get by having the
multiple disciplines talking to one another. The theme you’ve heard, I
think, throughout this morning. Which brings me to Health
Sciences and Technology, that a number of
people has mentioned. I just wanted to give you a
sense of the landscape in 1970 and to fill out what’s
here at MIT now. Health Sciences and Technology
was established in 1970. It’s, if you will, a
joint department that’s owned by Harvard and MIT. It’s done interdisciplinary
education and research programs, and it really
focuses on the integration, the translation from bench
to bedside, and of course, as everybody at MIT, innovation. And we use this to talk
about equal footing of the different
disciplines, and really integrating the
different professions so we can both identify the
critical unmet medical needs and bring the solutions
to the bedside. And industry is an important
player in all of that. We’ve just passed
1,000 alums, and we have 450 students enrolled,
including PhD students, MD students, MS MBA
students, and so forth. And we have 50 primary
and joint faculty. Like Doug mentioned in the
Biological Engineering, this group has grown
in recent years. I put this up here to say
I’ve selected a very small set of what’s happening here. I said, okay,
well, for Tech Day, why don’t I pick
the alums who are faculty who are doing
something that’s already reached the clinic? So that means that
I haven’t been able to talk to you
about the genomics, or the bioinformatics,
or the stuff that’s dealing with predicting
outcome in space through our active work
with NASA about quantum dots to bring molecules–
target molecules for cancer, and on and on. So hopefully, this
will intrigue you and you can come
visit us sometime. Doug talked about the map. And I think one
of the things HST does is bring under
the MIT umbrella the hospitals and
Harvard Medical School, so that the landscape for Bio
and Medical Engineering at MIT actually brings in the local,
very vibrant teaching hospitals and medical school and industry. Which leads me to my
closing point, which is, we are extraordinarily
well-positioned, I think, at MIT, to
have just a huge impact over the next
decades on advancing human health around the world. And when you think about
the number of disciplines, such as biological engineering
that’s starting in the fall, through biology,
which has been here for a long time, the number
of centers that have formed, plus the fact that with
HST, we’ve got Harvard under the same umbrella,
there is no other institution that I’m aware of
in the world that has the strength
and the collection that we have here at MIT. So we’re delighted that Susan’s
excited about this as an area, and I think we
hopefully come back in 35 years for
the next Tech Day and have equally
transformative advances. Thank you. [APPLAUSE] DOUG VINCENT: That was great. MARTHA GRAY: [INAUDIBLE] DOUG VINCENT: That was perfect. Perfect. Okay, I think we’re going to
still have some questions, but a little bit
abbreviated time. So if Professor Hockfield
and Professors, if you can join up here on stage. There are two microphones
in the aisles, folks. Downstairs in Little Kresge,
you’ll need to come up. We’ll try to–
again, I apologize. We don’t have the
full half hour. If we can line up questions
at those microphones, and we’ll get our
faculty ready to go. One thing while we’re
getting set here– hopefully all of you
collected blank surveys as you came in the door. If you would be so kind
to complete those surveys and leave them as you exit,
they’re very important for the Tech Day committee
that’s– if you’ve enjoyed what’s happened here today,
very much so, it’s a result of feedback that we’ve
gotten on those surveys. So please, please in
fact do take the moment to complete those. It’s very helpful to us. And thanks, too, for
all the patience as we run a little bit over. We’ll get you to lunch
in time, we promise. They told me that nobody’s
going to eat the food until we get there, so– okay. SUSAN HOCKFIELD: Ready to go? DOUG VINCENT: Yes. I think we’re good. Thank you. SUSAN HOCKFIELD: Go ahead
and start the questions. AUDIENCE: Since sugars have
such an important influence on the operation of the
body, what sort of risks are you taking when
you eat a bar of candy? [LAUGHTER] RAM SASISEKHARAN: I
think the short answer is they’re two unrelated
events, but I think we’re trying to really get
to the heart of how much sugar we take eventually is used in
the synthesis of these more complex sugars. But I think the short
answer is, I think you shouldn’t worry about it. Except your diet. AUDIENCE: Hi. With the brain imaging
and all, what impact does that have on mental
health, like schizophrenia? MARTHA GRAY: There’s– CREW: Mic one. MARTHA GRAY: There’s
a number of emerging– AUDIENCE: Can’t hear. AUDIENCE: Can’t hear you. RAM SASISEKHARAN: Now. Six? CREW: [INAUDIBLE] MARTHA GRAY: I don’t
know what number I am. Can you hear me yet? AUDIENCE: No. MARTHA GRAY: No. SUSAN HOCKFIELD: You
can get a handheld mic. ANGELA BELCHER: Here. MARTHA GRAY: All right. Well, let me talk
through Angela’s. So there are a number of
emerging analysis techniques. The schizophrenia is an
extraordinarily difficult problem. There does look like there
is some anatomic differences that you can see
between schizophrenics and non-schizophrenics. I’m not aware of
yet anything that’s showing the kind of functional
differences, though. There are groups
looking at that. But if you’re
interested offline, I could give you the names
of some people to talk to. AUDIENCE: Oh, thank
you. yes, please. SUSAN HOCKFIELD: Actually– MARTHA GRAY: And
yeah, you would know. SUSAN HOCKFIELD:
Can I just jump in? AUDIENCE: Can’t hear you. AUDIENCE: Can’t hear you. SUSAN HOCKFIELD: My mic– am I on? Okay. CREW: [INAUDIBLE] SUSAN HOCKFIELD: What? Can you hear me? AUDIENCE: Yes. SUSAN HOCKFIELD: Yes. Terrific. Many of you have
seen the new building going up just across the street
from the new Stata Center– the Brain and Cognitive
Science Project, which is going to be
the new home for all of our neuroscientists. There is tremendous
interest among the neuroscientists to make the
kinds of connections that we believe can be made almost
nowhere except at MIT between the basic biology,
the great engineering, and the implications for not
just neurological disease, but also psychiatric disease. So there are a lot of
people interested in it. And working with
the Martinos Center, we really think
we’re going to be able to open up this area we
have never opened up before. AUDIENCE: I used to do
cardiac tissue culture, and you showed something that
I thought was impossible. You showed cells that
had differentiated into mature cardiac myocytes,
with their myofilaments all lined up parallel
from cell to cell. And they seem to have
grown that way in vitro. How did you do that? Was that just the
electrical stimulation, or did you add some
magic growth factors? [LAUGHTER] MARTHA GRAY: Well–
[LAUGHS] first of all, I want to make it
clear that I didn’t do it. Lisa Freed and Gordana
Vunak-Novakovic did that. But I can’t speak to any magic
formulation growth factors, but they explored in
some empirical way and some strategic ways
what population of cells. It turns out just using myocytes
alone didn’t work very well. You had to have a
spectrum of myocytes. And they had to wait three days
before stimulating, otherwise the stimulation itself would
prevent the right connections from happening between
the different cell that you’d need to create that
kind of organized structure that you see. I don’t know how many days out– what it would look
like if you had looked after two days of
stimulation versus the eight days that I showed you. But I believe they
explored that, and they also looked at the kind
of stimulation they were doing. Not an exhaustive
search, but they were wanting to find something
that would work that you could imagine could work in vivo. So using– that’s one of the
reasons they used a pacemaker. AUDIENCE: You talked about
autologous chondrocyte transplants for putting
cartilage in knees. Can that also be used for
cartilage in vertebrae for people with back problems? MARTHA GRAY: No, probably not. At this point in a back problem,
there’s a number of issues. And if you’re referring
to disk disease, there are people that have
artificial disks that are– I don’t know if anything’s
been FDA approved, but– LINDA GRIFFITH:
It’s FDA approved. SUSAN HOCKFIELD: Sure? LINDA GRIFFITH: Yes. MARTHA GRAY: It is FDA approved. So Linda may have more
information on that. LINDA GRIFFITH: No. I just read The
Wall Street Journal. MARTHA GRAY: Oh. [LAUGHTER] And remembers what she reads. But the problem with
using that technique and putting it between a
load-bearing joint like you might have there is,
usually in the disk thing, you’d have to remove
most of the disk. In this case, they take
advantage of the fact that most of the cartilage is
there, and replace the defect. But the properties
of that defect could never support
your body weight. And so trying to imagine
a little hole in a disk is not really the way
the disk degenerates. AUDIENCE: Okay. Thank you. AUDIENCE: We saw that looking
at things from the engineering perspective, all of a sudden we
start realizing, wait a minute. We need to do glycomics. And as I understand it, if
we start at the cell level, then the glycomics are getting
immediate interconnection between the cells. But then there’s still quite
a distance between there, even to the organ level. And I’m wondering, is
there something similar that appears to be interesting
as a possible additional discipline that’s just above
the level of glycomics, at the next level
of interconnection? SUSAN HOCKFIELD: Who
wants to pick that one up? Doug? MARTHA GRAY: Doug. DOUG LAUFFENBURGER: Yeah,
I’ll say something about that. I can’t imagine another
discipline there. Because to form a new
engineering discipline, you have to have, I think, a
new science, quite honestly. And I think what you’re
talking about with that is just an extension of the sciences,
as the things Ram was working on integrate things from
the molecular and cell level and are able to
then explain what happens when they come together
and operate as a tissue. So I think it’s a
natural extension of a discipline based on
the molecular and cellular sciences. AUDIENCE: Dr. Griffith
and Dr. Belcher both talked a lot about
structure creating function and how to really
create something new. I’d like to have you just
have a chance to expand, having heard some of the
more practical things of the second half,
where you think some of the future
design products will actually evolve into? What you think will
actually hit the marketplace to help people themselves. LINDA GRIFFITH: Well,
I think probably you’re directing that more to Angie. But I can say, for
what we’re doing, we have defined some approaches
that would potentially greatly speed the drug
development process and make it much safer. We’re in the process
of translating those into pharmaceutical
companies right now. And so we just got
interviewed on NPR last week, and they wanted to
know the same thing. When is it going to
be in the market? Not by Christmas, so you can’t– [LAUGHTER] –get this. But we’re really trying to
get it in, in a research way, into some pharmaceutical
company laboratories. And I think this idea of having
the human body on a chip, where you have these things that
can screen against how humans really respond,
will be trickling into pharmaceutical research
labs over the next five years. And more pervasively,
maybe, in 10 years. SUSAN HOCKFIELD: Angie? ANGELA BELCHER: So for– we haven’t been focusing that
much on medical applications, but more for electronics
applications. And the applications that
we’re most interested in are making efficient solar
cells– large area solar cells. We’re very interested in energy. We’re interested in
batteries and different kinds of alternative energy sources. Some of the first things that– I have a company based
on this technology. They’re looking, or
are processing things like large area displays,
where you wouldn’t make the whole
display biologically, but you would make
certain components of it in a self-assembling way. So I think this technology
has the possibility to reach many different
kinds of applications. We’re talking to people about
medical applications as well. But we’re really–
we’re interested in disruptive
technology, so we’re interested in trying
to figure out ways that you may not be able to
make a material any other way. How to make
inexpensive large area self-assembling
solar cells and such. AUDIENCE: As a
follow up question, I was just wondering, even
though you might be developing this in one particular area,
like in health or in energy, I could see it being
translated into clothing, where you want to
change your outfit, you just change the
charge on your clothing and you’ve got a whole
different display going on. ANGELA BELCHER: Yes. AUDIENCE: Or make
it invisible, if you wanted to do something else. [LAUGHTER] ANGELA BELCHER: Well, we’re– LINDA GRIFFITH: [INAUDIBLE] ANGELA BELCHER: We’re looking
into textiles, but mostly for army applications. But also interesting
to have paints that are reversible
paints, as well, so you could change
the color of your car. [LAUGHTER] SUSAN HOCKFIELD: Over here. AUDIENCE: With regard
to glycosolation issues that we just heard
about earlier, does measured immunoactivity,
where you’re actually just measuring
levels of peptide, ever give you an exact
correlation with bioactivity? If you’re really with
absolute peptide levels, if you’re not necessarily– for example, within
endocrinology, TSH-producing tumors with relatively
low levels of TSH, you get a tremendous stimulation
of the thyroid, apparently in association with very
low levels of glycosolation of thyrotropin. So the question is,
what’s the purpose of measuring immunoactivity
if bioactivity is so– RAM SASISEKHARAN: And
with that specific point, I think one thing that
we’ve begun to understand is that by removing
glycosolation, you actually affect
protein folding that leads to immunogenic response. One of the interesting
thing is carbohydrates, by and large, generally have not
been immunogenic in the sense you can’t really raise
antibodies and things like that that effectively. So there’s some sort of
a biological mechanism by which they’re selected out. AUDIENCE: Yes. I wanted to ask
Professor Belcher– inorganic and traditional
organic chemistry. The same ports tend
to work all the time. Biological organisms,
viruses, and bacteria, as wastewater treatment
plants and brewers know, are not really very stable
and can reject their diet after a while. Do you have– I wouldn’t like to spray an
airplane and then wonder, did the yeast go bad today? Do you have some experience
working with that? ANGELA BELCHER: That’s
really a good point. For the most part, we’re
looking at display. And so these are short
peptides that are actually very stable, that we’ve shown– like our virus films
that I showed you, I carried those around in my
briefcase for about two years. And I could redissolve it, and
it still had greater than 90% of its function. That was partly
because of the way it self-assembled
itself, in a way that made it just really,
really stable, so you could store
it as a solid. But it’s an important
point, and I think that we’ve never had
any problems with mutations that no longer worked anymore. But it’s something
you’d have to test. Fortunately, it’s a
relatively easy test, and then you can keep
amplifying them back up. So we don’t have any data
that extends probably beyond two years, but I
think it’s an important thing to think about. AUDIENCE: A political minefield
at the intersection of science, religion, and politics
is, of course, stem cell research,
which is getting tremendous play in the press. A lot of interesting
things happening. I was wondering what MIT
is doing and thinking along these lines. And in particular, beyond
that, what other options, such as use of somatic stem
cells and growth factors for achieving some
of the same promise. What’s happening? LINDA GRIFFITH: So I’ll
start with that one. A lot of the work
I showed you, we use adult somatic stem cells
to achieve tissue regeneration. Somatic stem cells are used
every day in the operating room in the form
of marrow in bone transplants and other things. There’s an awful lot
of activity at MIT in the area of stem cells in
my lab and several labs here. A lot of us work with
adult stem cells, simply because they probably
have the greatest promise to be used for a lot of
things, both clinically as well as technologies like the
ones I talked about tomorrow. There are also faculty who
work on embryonic stem cells, and I think a lot of the
effort goes into understanding how humans develop. So embryonic stem
cells, in addition to what all gets in the
press, are fantastic tools for understanding
developmental biology. And so I think we have a very
pragmatic approach at MIT, as you might imagine, that
there is work across the board. And we certainly hew
to the regulations that are imposed by the government. But we also don’t eschew
the practical implications of a source of potentially new
technologies or new therapies. SUSAN HOCKFIELD:
Yeah, and I would only add that a lot of stem cell
research going on outside of biologic engineering. The Whitehead Institute has
a new stem cell initiative. We have some of, I would
think, the world leaders in stem cell biology
who are currently working on our campus. And walking this narrow line
between federally approved and federally non-approved
lines is tricky. And we are talking about
establishing a place at MIT at the Whitehead where it would
be funded not by federal funds so that some of this
work could go forward. I think, as Linda
points out, if you had a choice of getting stem
cells for your own therapy that came from someone else
or some other organism or cells of your own that had
been modified to do what you want, obviously we’d prefer
to get our own cells, because then you avoid
the complications of the immune response. But understanding what a cell
actually has to put in place to be a totipotent cell
is a kind of understanding that is likely only to come out
of research on embryonic stem cells. So even to realize the full
potential of adult stem cells, we’re going to have to
have a better knowledge of these embryonic stem cells. On this side, please. AUDIENCE: You’re touching on
the question I was going to ask. But put aside stem cell
research as a way of reproducing tissue and organs. How about– I’ve
been making parts and maybe many of my colleagues
here have made mechanical parts for all of our lifetime. How about taking the individual
structure and the elements that you’re learning
so much about and making tissue from that? And organs, eventually. Is that being actively done,
to create manufactured parts, if you will? SUSAN HOCKFIELD:
Go ahead, Linda. LINDA GRIFFITH: Okay. So I think both Martha and
I spoke to aspects of that. What I tried to
convey in my talk is certainly, that
is being done. And you saw examples of where
we’re trying to do that. I think there’s also a
very forward-looking vision of the future that yes, we will
continue to get in accidents. I drive a motorcycle– trauma. Might need to have
a knee replaced. But you’d love to
get away from having to have replacement
parts put in, because you’ve got a disease. It’d be much better– you’d be much better off getting
cured the disease very early in the disease process
so that you never have to go in the
hospital and have surgery. AUDIENCE: I’ll buy that. LINDA GRIFFITH: Anyone
who’s had surgery knows it’s not a
pleasant experience to go in and get sedated
and have things done to you. AUDIENCE: Thank you. DOUG VINCENT: I’m sorry
[INAUDIBLE] we’re out of time. If we’d like to eat
some more lunch, so– or eat lunch, so– thank you. SUSAN HOCKFIELD: Thank you. DOUG VINCENT: Thank you. SUSAN HOCKFIELD: You’re welcome. [APPLAUSE]

Leave a Reply

Your email address will not be published. Required fields are marked *