Well thanks for that nice introduction and
like to thank the Department and

the Institute for inviting me here today
to share my work our work on liver tissue

engineering and also using micro-manage
technology tools along those lines.

So I'd like to just start with this slide.

This is a slide from netters Anatomy it's
a complicated side of the liver meant to

emphasize the complexity of
Liver Functions and the row.

There are some five hundred
odd liver functions

that are divided into four main
categories detoxification.

Production plasma proteins
fat carbohydrate and

protein metabolism and
also production of bio for

gesturing So collectively these
are the functions of the liver.

It's not surprising then that
when the liver fails it's fatal.

And that's an increasing problem
due to have otitis B. and C.

worldwide it's estimated that about ten
percent of the worldwide population

carries one of these two viruses.

So the gold standard of therapy for liver
disease as whole organ transplantation.

And as is the case for other tissues there
are really not enough livers available for

to for demand that's required so
that has led to over the years.

A lot of work in liver
support technologies

because of the complexity of the liver.

Most of these these these
technologies are cell based so

I'm going to emphasize to you today.

So the first two just for completeness
are parasite transplantation so

this is typically intravenous or
intra portal injection of a pad of sites

that this is done for sort of single
gene defects in metabolic diseases.

People are making trans Gene
examiner grafts typically

humanizing the end of feeling them of
whole organs and then the two that I'll

focus on our show down here so the first
are these are extra corporeal devices.

You can think of them as dialysis like
machines they're typically hollow fiber

cartridges and they have had a site
like cells over the last few years there

have been a couple of very disappointing
very expensive very prominent failures

in this area primarily we think because
they have had a science of the past.

I like cells and these devices are not
functioning well and as I'll show you.

That's one of the main emphases then and
of my group is how to get a pad of sites

or parasite like cells to function in
the sort of synthetic environment and

the other thing that we and others are
working on which is a very lofty long term

goal is actually making an implantable
piece of liver tissue or

tissue engineering of the liver and
I'll touch on that a bit.

So this is an overview of our
what my group is working on.

So as I said we're working on making
functional engineer liver tissue and

there are a number of applications
that we're interested in.

So the therapeutic applications that
I just mentioned the other sort

of interesting thing that comes out of
having this stable liver tissue in the lab

is that it turns out to be a great model
system to study melty cellular Ophelia.

There's a lot of interesting
biology here that one can study now

in vitro instead of in vivo and
then the last application that we're

interested in that comes out of having
stable liver function in the laboratory is

using miniaturizing these
pieces of the rich issue and

thinking about how to interface
this with micro technology and

make in vitro platforms for drug screening
for pharmaceutical applications.

So these are the applications that we're
interested in our approach is twofold.

So there's a number of somewhat
unique problems to liver tissue.

Although I was talking to
Dr Service last night and

I think some of them are quite common.

Actually in the pancreas world as well so.

So the first is that how to sites mature a
pad a science a functional cell the liver

don't proliferate to any
measurable degree in vitro.

This is a huge.

Problem It's estimated that you need
about ten percent of the liver mass

sort of therapeutic benefit.

So we and others are looking at stem
cell sources than which our political

source then that could mature into
a differentiated a path of sight.

The two stem cell sources
that we're looking at our

liver specific stem cell
called the oval cell.

This is a by potential cell that
can make both the pad aside and

the clan to fight which
is a billiard ductile so.

The other stem cell source is an embryonic
stem cell source routing at mouse

embryonic stem cells so

I'm actually not going to talk at
all about our stem cell work today.

But I just wanted to let you know that
we're working on that what I will be

talking about is the second problem which
is relatively unique to have out of sight

it's a matter is that they rapidly lose
their phenotype in vitro so you take them

out of the animal you put them in this
device or in a construct or whatever

it is and they lose most of their liver
specific functions on the order of ours.

So what we've been interested in is
looking at the micro environment of

that data sites in vivo and

seeing whether we could learn what the
palace I see in B. But what's important.

Then for
stabilizing their function in vitro.

So this slide explains more what I mean so

this is the architecture of the liver in
vivo So for the chemical engineers in

the audience it's essentially a radial
flow by reactor blood comes near here in

the periphery drains to the central rain
here in this radial flow pattern and

along the way in quarters
encounters these a parasite cords.

So it's a very efficient
mass transport machine.

And these a pad of sites then have
very well defined interactions

with their microenvironment
they have very well defined So

Matrix interactions pretty well defined
interactions with the blood that streaming

through the Cybil factors that
are coming out of the gut.

And then interactions both with one
another we call those homo to pick and

interactions with five other cell types
in the liver we call those heretic so

if you look at the liver.

Biology literature to some extent and many
of these have been mimicked in vitro and

to some extent all of these
stabilized liver functions.

So what I'll be focusing on is two
particular ones header to pick Coke ulcer

and the role of oxygen in the blood.

So I'm going to start with the role
of this heretic Celso interactions

by describing an effect known as the coke
culture effect in the literature.

So this is real effect in the have outside
world first reported in the early one nine

hundred eighty S.

by a French group who can use a group and
I'll describe it here to you in data.

So the effect as I mentioned before
on how to sites when isolated.

So these are isolated from the rat liver
are plated in vitro they rapidly lose

their phenotype here we're using
albumin as a marker of one

axis of liver specific function that
the liver synthetic function but

this is true for all the other
functions of the liver as well.

So your plate is a pad of sites on
a college in a petri dish they lose

this liver specific function right away
if you look at them after ten days

have to trust me if you're not with
used to looking at a passage but

these look horrible these
are very fibro Blaster

they have indistinct nuclei
bright intercellular boundaries.

If on the first day of culture
you add another cell type.

So it turns out this works for with many
different cell types here we're showing

three to three fiber blasts
from the a mouse source.

If you add these on day one of culture
you see your dramatic regulation and

stabilization of this function for
many weeks.

OK So this is well known from the early
eighty's in the morphology of these

are past that cluster over here with nice
bright nuclei nice bright boundaries.

They also have all the other functions
that one can think about measuring.

So it's not clear how this works but if
you think about how one might use this as

a bioengineers you start to ask yourself
some sort of very simple question.

So if I was going to put these
in a hollow fiber diarrhea and

what would be the ratio of
the cells that I would use.

Do they need to touch or

could they be separated by a membrane is
it important to have a pad of sight may.

Some level of what's about contact or
is it OK to be alone.

So those are the sorts of questions
that we were thinking about asking and

we stress first started
working on this problem and

you can see in a conventional culture we
call this a random Co culture a pat of

sites here are forcefully labeled red and
they're surrounded by these fluorescent

labeled Green Fire fiberglass but you can
see is all of the variables are coupled so

for instance if you want to change from a
typical interaction you lower this L.C.D.

density you change the ratio and
so on and so forth.

So what we did was we developed a micro
fabrication based tool the micro

patterning tool to uncouple
some of these variables so

I won't go through the details of
the process but essentially what you

do is you use semiconductor technology
to do photo fog or fear photo patterning

of college in this case a glass
wafer OK So these are the exact same

techniques that one would use for making
my for electronic circuits on silicon.

So you pattern college and
in this case on a glass wafer

you see a pad of sites on that and they
selectively adhere to that college and.

Then you surround the cells by
seeding the fiber glass and

the presence of serum proteins.

Those serum proteins will
absorb on this bare glass and

then mediate the attachment of
those fiberglass in a pretty free.

OK So using this micro patterning tool
you can go from the random coacher

that I showed you before to highly
organized tissues in vitro OK So

again here these are the rat pack of
sites separated by a fiber glass.

So we set out then to ask some of
these questions that I loaded too so

the first was What is the role
of header to pick interaction.

So we set up these are four
different culture conditions and

in each dish we have the same number of
constituent cells same ratio of cells but

what's varying is the level of interaction
between the cell population so here.

These are single cell that had a slight
islands of forty my trying to mention

these are one hundred micron
colonies five hundred microns.

And this one is many millimeters.

OK.

So what you see is that in the absence
of the surrounding straw and

the absence of the surrounding.

Fiberglass if you look
at a couple of markers.

Again this is after ten days.

What you see is to both for
nitrogen metabolism and for

abdomen production in the absence of
the fiberglass you see very low function.

Goes with the data I showed you before.

In the presence of the fiberglass
if you look down here.

What you see is in a manner a dose
dependent increase in tissue

function such that the cultures
with the most interaction

with the smallest island are more highly
functional than for example these so

these cultures are also very stable but
they're stable at a lower level.

So we were curious as to
about what was going on here.

So we thought let's stay in
the cultures and see which cells

in the system are contributing to
these secreted bubbles of protein.

So we did just that.

That's shown here so this is immune.

It has to chemical stain for

albumin So Brown is album in using
horseradish Praxedis and what you

see on the left is a pad of sides alone
and on the right sides and Co culture.

So you see on day one.

This is a five hundred micron colony of
how to sites that are chock full of album

and they've just come out of the liver.

If you cultured them in
the absence of any neighbors.

You see that that staining declines as in
the secreted data case if you surround

them with fiberglass which you can't see
here because they're not standing for

albumin what you do see eventually is that
the cells in the centers after six days of

culture the cells in the center behave as
if they're seeing no fiberglass at all and

the cells in the prefer actually
are maintaining their album of production.

OK so this.

It follows then that the cultures
that have very large colonies

are relatively little interface
were stable a functioning.

But at a lower level.

OK it's a bit more
complicated than that and

I'm not going to go into the details here
but this this ring that was differentiated

ring a function if you will is actually
about two to three cells deep and

not just one cell layer deep and what we
think is happening here is there's a are.

Role for Homo to pick gap junction
mediated signal propagation.

So again I'm not going to I'm going
to skip the details of that but

I'm happy to discuss it later.

This is sort of our working model.

So what we think is happening is
that there is fiberglass out here.

They're signaling this a pad a slight
colony through some header to

pick mechanism and
then these a pad aside so

know one another in this
gap junction dependent way.

Our thinking is that this home
a typical signal at some point

falls below a threshold and

therefore the central part of sites lose
their phenotype So this is a working

model that we're actively testing with
Gap junction inhibitors and what not.

But again if you think
about this as engineers you

start to ask yourself what is the signal.

Do you need these surrounding cells and
could you get away without them.

OK so we were interested then and what the
MET what these specific mechanisms might

be and our approach then was to take
a gene expression profiling approach.

So what we did was we cope cultured
a pad of sites with multiple different

fiberglass cell types so
that data is here.

So we looked at two functions
again synthetic function in one

thousand metabolism and up here what
you'll see is we've used four different

fiber glass strains that are all
Mouse derived and highly related.

And we've cocoa trigger them with
a pad of sides and scored them for

their ability to induce
differentiated function.

So this plan was scored as high these
two as medium and these With low.

OK you can then gene expression profile So
look at all the genes that these cell

types are making and ask which genes
share this pattern of high medium and

low expression and then also the reverse
in case they're making an inhibitor.

OK so we did that using
after metrics gene chips and

again I'm not going to go
through the details but

we found that there are one hundred ninety
genes differentially expressed in these

systems twenty four of them have
this high medium and low profile.

You can see high medium low and
I'm not going to bore you with a gymnast.

But I have to show just one.

So this is the these this is the G.

list for the genes that are positively
correlated so there Heidi I'm alone

the same way that the function is and
what we did is we sorted them by.

Basically their role
in the fire blasts and

what you see is some
are in the cell surface.

Some are secreted and
some are in exercise your matrix.

So I guess engineers were
very interested in the ones

that are expressed on the cell surface
that are secreted or that are matrix

because we think those are candidates for
replacing the fiberglass completely.

So what I'd like to focus on is
this particular molecule Decker and

this is a pretty black and that's in the
extracellular matrix down to colleges and

what we want to do is then go back and
test in our system whether Jack Grant had

any ability to support the differentiated
function of the PAP sites.

So we did two experiments the first
experiment is shown here on the left

again synthetic function and
metabolism and

what we did was play the pad asides on and
absorb layer of college in or on and

it's our Blair of college and that was
Colin created with deferent OK and what

you see is a statistically significant
increase in function in both cases.

Unfortunately though if you
look at the axis over here.

The function of these cells
is actually still quite low.

So over time they go on to
lose their phenotype and die.

OK So we haven't rescued them completely
with this molecule the next experiment

that we did was to add background to
the low performing Co Coulters and

ask whether we could boost their
function to a higher level.

So you remember that we had CO cultures
with a low inducing fiberglass strain.

We asked How do they do in
the presence of background.

So here we've done a dose
dependent experiment.

These are these are have
had ascites alone.

These are the low
performing Co cultures and

these are the low performing in the
presence of increasing amounts of debt so

we see a dose dependent increase in
the function of these co called.

We get up to about forty
percent of the maximal level

by this does depend by
this exposure to Akron.

OK So again we don't have the whole
Co culture fact we've got part

of the coke ultra effect by adding this
one molecule So what we're doing now is

going back to our gene list and looking at
some of the other candidates of interest.

But we're encouraged by this data in
the sense that these are some of the first

molecules that are our
new in this field and

quite some time in terms
of what might be going on.

OK So this is just to summarize.

So the Colts are stabilizes a pad a site
phenotype we think that there's a Nanda

fusible signal that comes from
the fiber glass that signals might be

propagated via get junctions and the sort
of regardless of the mechanism the moral

of the story is we have very high
levels of function in our optimized

code cultures relative to other cultural
models that in the in the literature and

also relatives in the bubbles some
of the remaining questions that

we're working on are for
example is continuous signalling required

does a second cell type need to
keep signaling the palace lights or

is that enough to sort of kick them into a
differentiated state will they stay there.

So we're set up a collaboration with
me at the University of Chicago.

He works on dynamic surfaces and
he's got a chemistry with an electro

active tether on an R G D OK So
we see the fire blasts on R G D that

has a selector active linker you apply a
small potential then you can release these

fiberglass to after having performed the
CO called her for certain amount of time.

So this is a model system that
we've set up we've now just got our

first release and we're getting ready
to find out what the biology of

the second question as I mentioned
is Kemi replace the fiberglass So

we're looking at some of the other
candidates on that list and

the last question is a bit of
a teleological question but

it helps frame our thinking and
that is what actually are we seeing here.

What is this interaction that is so
robust return to Paris sites and

many different MS ankle cells is it
a physiological reaction interaction that

exists in the normal liver in
vivo is it something that mimics

embryology because in fact the end of term
the developing end of term of the liver

interacts with surrounding music or
is it something that happens in repair for

example in regeneration when again these
sorts of interactions are known to occur

and the reason we're interested in that
is we think that some of the other

gene candidates for example
the Cybill candidates might actually

have clinical utility as what are known in
the liberal world as a patio protectant

to boost regeneration for example.

OK so that's all I'm going to
say about salsa interaction and

I'd like to move on to another micro
environmental cue that's very well

defined in vivo that we think
might be useful in vitro and

that is the role of oxygen so
until now I've been

talking about a pad of sites where they
all have had a sites are created equal.

But in fact they're not so

if you look at the capillary of the liver
the liver sinusoid and you linearize it.

What you'll see is distinct populations
of gene expression a long list like.

OK that's referred to in the literature
as liver zone nation and so these cells

are referred to as per the portal or
near the peri and there is a portal vein.

These are Perry Venus or
any other centrally.

And what you'll say is that carbohydrate
metabolism and detoxification and

a number of other Liver Functions
are actually compartmentalise in

these different cell populations.

It's not clear what causes this so NATION.

This is a sort of a recurring question
in the literature what's the cause and

effect what we know is that these are
exposed to high levels of oxygen that is

progressively depleted and high levels of
hormones that are progressively depleted.

These are these are more
innervated these are more

exposed to the accumulated sell
products of the sinusoid So

they're essentially a series
of steady state grading.

Ants that exist in this environment we are
were particularly interested in oxygen for

a number of reasons one of them being
that as engineers we know we control it

very well in a reactor setting.

So this is some clinical significance and
also significant for

drug development in the following way.

This is a hostile audible section of
the liver this is a central vein here is

a normal liver what you've seen you'll
see in the presence of Tylenol or

acetaminophen which some of
you may know is a wonderful.

I Pad a toxin.

Is Perry central cell that.

And essentially what happens is
that the people are fifty enzymes

which are localized here in
the peri central region generate

a toxic metabolite locally and
that causes local cell.

OK So livers the nation also has to
zonal how to toxicity consequences.

So we were interested in
whether we could add some of

these features these onal features
of the liver back to our in vitro

models because they do seem to be so
important.

So what we did was set up a parallel
plate Vireo actor that works in

the following way and have a sites or
a monolayer on the bottom.

And we set up the flow conditions such
that basically oxygen is progressively

depleted as the solution traverses So
there's a simple

model of that which I won't go through but
essentially you can control the dimensions

of this reactor the convection the
diffusion and the reaction at the surface.

So this is what the schematic
of what it looks like

in the lab all show you
a picture in a minute.

It doesn't look quite so

pretty as though this is a virus actor
we have an oxygen probe at the outlet.

This is a computer controlled syringe
company gas exchanger on it so.

This is a picture of the system
in the lab this is a link

you better in the reactor sitting in here.

Here's a syringe pump.

This is a two D.

report of the model and essentially
what we first went outside out to do.

TESTER model predictions.

So we varied oxygen in
this reactor in two ways.

So the first was you could control the
inlet oxygen concentration in the you can

change the flow rate as a fairly
it gets slower than the outlet

oxygen concentration is lower and
lower as you would expect you can

model the oxygen uptake at the surface of
the cells either as a maximal uptake or

an I'm a callous man and Foreman we've
done both the other way to vary the oxygen

tension is just lower the inlet oxygen
concentration with which you elaborate.

And again you can see the model
matches the predictions quite well.

OK so we're pretty comfortable
that we were we knew

the oxygen concentrations in this reactor.

We set out to do the experiment that I
described what isolate a mixed population

of a pat of sites.

We played them this in this reactor we
expose them to this oxygen gradient for

twenty four hours and then we harvest
the cells from different zones and

in Western watts for particular proteins.

So this is a P four fifty
enzyme that's supposed to be up

in the peri central region.

This is unbearable carbohydrate
metabolism Pepsi K.

it's supposed to be up two to five fold
in vivo in the Parent Portal region.

So we start with a mixed
population of cells and

then added back some
zonal protein expression.

So the next experiment was then can we
capture the zonal have had a toxicity

feature that would occur because of this.

So what we did essentially was dose
our buyer reactor with acetaminophen.

OK So and then stained with empty teeth so
this is a viability stamp purple is

live this is all looking down on
the slide from the inlet to the outlet.

So this is a relatively low
concentration of Tylenol and

you should be able to see that most of
the cells are alive is a relatively high

concentration of Tylenol you should
see that all the cells are dead and

at this intermediate concentration
we should be able to see is that

the electrons in that cells are living and
the Alice cells are dying.

OK So again Perry central graph
that we think is mediated by

oxygen mediated local it's low.

Concentrations of P four fifty and
OK and that's connotation of the data.

OK so what I showed you so far is

the role of the micro environment in
stabilizing engineer tissue in the lab and

then adding some of the IN VIVO features
that are really missing today and

I think I've highlighted ways in which
we think this is a good model system for

exploring some of the structure
function relationships.

So what I'd like to now touch on is some
of our early ideas about how one might go

about building three dimensional tissues
that one could implant and again this

is a very lofty goal but we're sort of
making baby steps in this direction.

So liver architecture is highly three
dimensional if you look through

the literature and see what people have
done in the engineering space does or

some examples have sites
seated on micro carriers

pretty aggregated in severe oids
are seated on biodegradable scaffold.

And by and large what you see is that
a pad of sites being so highly metabolic.

Cannot essentially be passively
nourished in this environment so

you really need convection to carry oxygen
and nutrients to this large mass of cells.

So the two approaches that have been
taken is are to actually try and

encourage angiogenesis in the system.

Or to create some sort of profusion
system where you can add convection

through the scaffold.

So in the profusion world.

One approach that's been
taken is to build three

dimensional scaffolds that
then you would see with cells.

So these are a series of three dimensional
scaffold fabrication technologies

pulled from the literature and essentially
what you see in all the cases these

are layer by layer what we called additive
free form fabrication technologies.

So they're all layer by
layer technologies for

building three dimensional architectures
on which one would see the cells and

then one would have channels through
which to provide convection.

There's a few challenges with
using this approach for.

Or liver tissue engineering so
the first I've already mentioned but

I'll highlight again and
that is that a pair of sites don't grow.

So this is different than other
tissue engineering in the sense that

you can't just see the scaffold and
then grow up a whole bunch.

Hey basically you get used and with what
you start with the second problem is that

a pad of sites are notoriously non
migratory So essentially they stay where

they land they don't then go on to
populate the graft very effectively.

So our idea was to try and
combine this approach of something that

had convective channels with a scaffold
that had cells embedded within it.

So what we wanted to do was
an adaptive as a proof of principle.

Well known chemistry that was first
developed by Geoff hobbles group so

these are photo Plimer arising hydrogels
in which the peg back relate base so

polyethylene glycol hydrogels in which
you can mix cells photo initiator shine

light and hydrogen then cross links in
the presence of cells and traps the cells.

So we want to use that chemistry combined
with this idea which is a stereo

lithography idea.

So stereo a father fee is a rapid
prototyping layer by layer fabrication

technology where essentially a shine
light in a different pattern for

every layer of the part and
you build a three dimensional part.

So we wanted to do this layer by
layer building of a live to ship.

So this is our idea of what one might
make so this would be a first layer of

tissue that has these sort of hexagonal
structure with cells embedded within it.

This would be another layer within this
would be a third and then if you look

at it in this direction this is sort of
like a network with of branching channels.

So we built a bench top
device to do this essentially

then what you do is inject in this
device a solution with cells you

put a mask on top of it shine
light through the mask and

then wash away the I'm cross-linked
areas and then you can repeat.

This in a layer by layer fashion.

So we first did this with single layer
structures that's shown here on the left.

So these are hydrogels that contain
living cells on a glass surface.

This is again a single layer structure
that has two different domains.

So this has for people who are used to
working with hydrogen sort of the feel of

a let's say a soft contact lens but it
has different domains of cells within it.

So to our knowledge or it's very
difficult to make the structure and

in other ways so that all
the cells have a three dimensional

microenvironment that you can control
the architecture of the tissue.

These are cells that have
actually been a raid for

high throughput screening applications.

So the spatial resolution of
this technique we really haven't

pushed the limits were using from
the microfiber cation world what

would continue what one would consider
really low really poor technology reuse

transparency of those masks.

So the resolution of these things
is about seven to ten microns.

So we get about one hundred micron
resolution using the techniques that

we're using but we now think we can go
easily down to the single cell level which

is about twenty microns.

We first started with
very robust cell types so

these are all done with a habit
Thomas cell type have G two and

fiberglass we like to then move
on to doing primary Patta sites.

So these are some data with
primary parasite structures.

Against her mind what we're
trying to do these are viable for

us to label to pad a side so
show the slide my

grad student would kill me because
it took her a long time to get here.

We had to change photo initiators and you
know the whole thing the system is very

different from the previous slide but for
the purposes of this talk of bad faith.

We now have viable product sites and these
constructs here they've been fluorescent

labeled in these three structures here
we've made a multi-layer structure which

is actually quite difficult to visualize
you can try and get a sense for

it in this country will
image here we've rendered.

The first two layers in matlab for
you see a little bit better.

OK So we're starting to build these
three dimensional structures.

These are three dimensional
microstructure So these struts here

are hundred fifty microns this is a three
layer structure from the top down and here

we've labeled the first population read
the next one green and the next one blue.

So the reason I do that in
this in this slide is also to

highlight the fact that we think this
could be a tool that would be useful for

tissue engineering other tissue types and

the fact that we've labelled the cell
types different colors in different layers

is that meant to indicate that you
could have different cell types and

different layers you could have different
hydrogen chemistry is in different layers.

You could have different micro carriers
embedded within those hydrogels

secrete growth factors of different kinds
so we think it's a very versatile tool for

sort of starting to build
three dimensional tissues.

So what's next for us from a liver
perspective had a sites Artesian dependent

and these are completely inert hydrogels
So we have a collaboration with Jennifer

west at Rice to look at peptide
dramatizing these hydrogels

surreal looking at the interim
profiles of these the powder sides and

what peptides then to fit in with back
them to be degradable with M.M.P.

degradable sites matrix metallic
pertness dirigible sites.

We are actively conditioning
them in by actor units and

developing an animal model of
liver failure at U.C.L.A..

And then we'd like to study in
the lab some of the three dimensional

structure functional relationships
that we've shown in two D..

So again this is a very long way to go
before we have an implantable piece

of liver but
this is our sort of stepwise approach.

OK so what I've told you so
far is that we've we're trying to function

engineer these functional pieces of liver
tissue that they're good model systems for

some basic science that we're making some
progress towards therapeutic applications

and then what have I would like to talk
about is one how Mom might use these

as in vitro screening tools.

So let me just sort of.

Step back for a second this idea of
doing biology in a chip based platform

is part of a larger arena that is what
I'll call chip based bias systems.

So in this space.

There are.

People do molecular biology on a chip
essentially you could have chips of D.N.A.

R.N.A. protein or drugs.

It will do integrated laboratory
functions on a chip so

it's a whole sequencing reaction so
from a sample all the way to sequence.

We call those lab on a chip devices.

And then what I'll be if we're talking
about and what an For example in

a Griffith work is on are integrating
live cells with chip technologies.

OK so this would also be for example some
Lomax and some other companies of space.

So why are people interested
in doing things on a chip

there's lots of miniaturization arguments
in these essentially amount to you

get more for less.

And so I won't go through them all but

there's also some interesting microscale
phenomenon that motivate this work.

So it's been clear for a long time that
there is physics at the micro scale that

one can exploit for example surface
forces dominate lots of things and

you can use those to your advantage.

It turns out that we and others have shown
more recently that there's also some

interesting biology at the micro nano
scale that might be useful in these

platforms so that's been another driver
to go towards these lines scales.

OK So these are what these chips look
like first in terms of their electrical

engineering and years in the audience
they don't know what I mean so

these are these are plastic
chips with reservoirs and

microfluidic channels that's a kind
of chip that I'm talking about.

OK So this is a vision for
this type of work.

So this is a schematic of
an imaginary population of cells here

they're all fluorescing a different
color all different a reporting on all

different activities there and interface
with some synthetic platform they can

be interrogated with microfluidics could
maybe even have optical tweezers and

into manipulate particular cell.

It's of interest.

So what would it take to make
something like this happen.

You need to first and foremost control the
cell function in this kind of environment.

So I think one thing that's perhaps been
lost when this is been approached from

a straight engineering perspective is that
if you interface cells with a synthetic

surface and they are no they
no longer behave as if they'd

like they did in the body the data that
you get out is actually not useful.

OK so you need first and
foremost to have your cell reporting on

what it would be doing in vivo
to make this a useful idea.

And so in that regard.

We think lots of the tissue engineering
concepts can be actually applied

to this area of research.

I mean the next thing you need to do is
integrate your cells are synthetic in

living systems you think about how
one would interrogate them and

detect signals how to automate these
things and screen in parallel.

And then for some applications you'd
actually like to miniaturize these devices

and make them portable so
people have thought about this for

point of care diagnostics or for Cambaia
war fare detection that kind of thing.

OK so the applications as I
mentioned are in drug screening

also functional genomic said I'll show you
a little of our work in stem cell biology.

So what I will be talking at all about
is the automation and parallel screening

there are plenty of companies actually
in this space working on automated

image processing auto focus how to handle
large volumes of data that kind of thing.

And we're really not so much focused on
that aspect of automation because we.

Because we think that there's sort of
a downstream technology ready for that.

So what we're really interested in are the
first three issues controlling the cell

function at the interface integrating the
live cells with the synthetic surfaces and

then how one might start to detect
signals from these systems.

So the first one I really
already talked about so

I want to talk about it again here.

We think that we've come a long
way to stabilizing liver cells.

In a laboratory and that there
would be useful in this platform.

So what I will talk about those
what some of the handles one

can use to manipulate
cells in these settings.

Chemistry is very useful serves
the selected heat and as I've shown you

typology physical environments
fluidics treating the cells

like objects that are charged and
using electric fields to move them around.

Or objects that scatter light.

So you can also use optical forces.

So we've done many of these
all touch on a few examples.

So the first is a selective
He's an example and

this is a collaboration with
one of these companies just to

miniaturize our ass a in a way that you
can actually think about doing drugs

creating in any realistic way so this is
the miniaturization to a twenty four.

Well played assaye.

From the experiments I
described to you before.

So the micro patterning I described to
you previously was done on a two inch way

first to a piece sixty and
it was you had to do the fabrication on

every single wafer use it and
then you can't use it again.

OK so that's how I did my Ph D.

but no one's going to go around
screaming drugs that way.

So this is a an attempt to try and
paralyze some of that.

So what we've done here is switch micro
technology platforms from what I referred

to before is hardly far graphy to
this is called soft lithography So

this basically takes advantage
of methyl Psylocke saying so.

P.D. a mass which is an elastomer And

essentially you're looking
down here on a twenty four.

Well plate system on the bottom of each
of these wells is an array of holes or

stencils OK you can then put collagen or
whatever your protein of interest

is on top of this and it will absorb
to the exposed underlying surface and

then peel away the sensel
OK as shown here.

So we've done that with them and

created micro patterned arrays of a pad
of sides and Co cultures and gone out.

I should update this
now to eight weeks and

shown stable function in this platform.

So this is just a technological step to.

Waiting to miniaturizing and
paralyzing the system.

OK so that that was an example
of chemistry or selective in

Houston the next hour like to show you
is an example where Actually that was

not desired so we started a collaboration
with rusty Gage who's a neural stem cell

biologist he was interested in the
differences within a stem cell population.

So he actually did not want any kind
of micro patterning approach that pulls

down selective fell populations because of
that because of their immigrant profile

let's say he wanted something where we
could look at all the cells growing and

the differences between them.

So what we did with him is
set up the following system.

So this is actually
a modified Rose chamber for

those of you who do my cross could be so
this is essentially

a chamber in which you can grow cells
you can stick it on the incubator put it

back on the microscope stage and monitor
them every day and what we've done is add

basically ten thousand wells
in the bottom of this thing.

OK so.

This is a chamber is a micro fabricated
a way of wells in the bottom and.

What we do.

Is this is a syringe coming
in with cells suspended in it

is inject cells into the system.

Let them set a meant by gravity
This takes about five minutes.

Wash away the cells that have
not fallen into the wells.

OK And now this is on a cover slip so you
can do very high magnification if you want

to put this on the microscope stage which
is run on an X. Y. automatic platform and

you can raster through all the demands
all ten thousand wells every day and

come back to the same fields and
monitor exactly what's going on.

So if you see the set up these at a high
density these are fluorescent labeled

cells looking down at them but

again as I said he was really
interested in Comal cell growth.

So this is just one view of
again ten thousand wells.

So these are cells that started out as
single cells these are these neural

stem cells and we've come back to the same
field this is Day to day for day six.

You can watch.

I'm growing what we've done now
is differentiate these cells at

the end of the experiment and
ask how does the differentiation

potential of these cells correlate with
it's proliferated history and so this is

just an example of how one might use the
sort of technology in stem cell biology.

OK so the last example I'll
give of this sort of technology

is again treating cells as objects
that are either charged and

moving them around with electrophoresis or
treating them

as objects that are polarized the ball in
a different way than their outside medium.

OK And
that actually is called Dial-A to freeze.

So Dr Frederick force moves objects around
by the gradient of the field squared so

basically objects are polarized
the ball in a different way.

Relative to the outside outside media and
you can use that to move them around

in a ray them and interface them with
these sort of chip technology is so

this is a different set up than the one I
described previously so this is a micro

fabricated array of electrodes that
are made on a transparent semiconductor

called indium to not side and the reason
we use a transparent semiconductor is so

that we can watch things in
real time on the microscope.

OK so here these are the electrodes and

then they're surrounded by an insulator
here which is a what we call a fixed

photo resist and
the counter electrode is up top.

So basically what if you look
at the field in this thing.

You'll see that the field maxima
are on the electrodes and

the field minima are sort of in this
honeycomb pattern between them.

If you put cells in this
sort of configuration and

the frequencies that we use they undergo
what we call positive dielectric frissons

that is they go to the E.
field Maxima A shown here.

So if you do that.

These are conversely it's so
this is a time axis down here or

a over about a minute forming
clusters on the surface.

This is a ray of high.

Define clusters this
project actually is really

actually a tissue engineering project what
we're doing here is Arain Condor sites

into different cluster sizes to
look at the role of cell phone or

action Condor sites in a relatively
three dimensional way.

So we didn't want to array them on
a surface and have them spread against

the surface because that's not their
conversation microenvironment So

what we do is we cluster
them in these clusters and

then we embed them in the hydrogen
that I showed you previously.

So they have a relatively three
dimensional microenvironment

of a very well defined cluster size.

OK if you put beads into a system like
this what you'll see is that they

undergo for example a polystyrene
beat of about the same dimensions.

They undergo what's called
a negative dialect or freezes.

So they should go to the honeycomb
pattern between the electrodes.

So here we've done is put in both
what you should see is sorting so

the cells are going to be filled maxima
would be they're going to the field minima

OK this is a low Magritte in a vat and
from a sensing perspective

this actually becomes quite useful
because there are plenty of beads based

sensing technologies so there are oxygen
sensitive be coatings ph sensitive B.

coatings and so on and so forth.

So we started a coverage with David Walt
at Tufts University to look at

essentially where they're
exploring the idea of

whether we could have
a sort of family of sensors

around a cluster of live cells that
one could then it Terra gate remotely.

With for us since.

So this is a sort of sensing application.

OK so
we've looked at a lot of other handles and

I'm not going to go through
the details sort of as through them.

The moral of the story is depending
on your application various handles

are useful for interfacing cells with
these chip platforms these chemistry and

apology are very biocompatible in
the sense that you're really not

perturbing the biology very much.

These you start putting
energy in the system and

you have to work very hard then to look.

As viability and gene expression
changes as a result of perturbations.

OK so the last couple
minutes I'd like to touch on.

Some some work that we're
doing in detection.

So we were interested in applying
quantum dots to imaging in live cells so

quantum dots as many of you know are our
nano particles that have size tunable for

essence this is a slide taken
from seeming nice group.

So the contacts here for
example there are cadmium So an icon and

dots that have sync sulfide cap So

these are particles that are snapped other
two to five an anatomy hers and diameter.

They have size tunable for us.

And so these are smaller particles these
are bigger particles they're all being

exposed with one hand tell you the lamp
and they're emitting based on their size.

So people had previously used these for

tracking cells either via Anderson
labeling so they receptor mediated

the psychosis they get sequestered in the
end zone when you could then track those

cells as they are let's say migrate
around or tracking them in the embryo.

So sort of wholesale tracking people have
been very excited about using them for

biological applications because
they're relatively bright and

they're relatively photo stable that is
they don't photo bleach in the same way

that are getting dyes do.

They also have some potential for
being multicolor labels so

you could imagine labeling more
than one thing as one more thing.

One thing at once so we felt that
there their promise as subsidy or

labels had not really been fully realised
so we get anybody who works in sort of

the cell a molecular space things you know
the end zone is a nice place to be but

it's usually a place
you want to get out of.

And so.

So we took the following approach we

modified the surface of these
quantum dots with peptides and

our goal was to have the cells use
these peptides to traffic quantum dots

to various locations of interest and
then also we co ma.

By the quantum dots with polyethylene
glycol to prevent nonspecific binding.

So the first experiment that I'll
describe is in my projection experiment

where just the control Q. dots are
injected so that's done up here is a phase

micrograph of a cell you'll see that the
salesman my current actives with green Q.

dots and they're excluded from the nucleus
and as and quite well dispersed.

So the next thing we did was then put
a nuclear localization sequence taken from

the S B forty virus on
these quantum dots and

uses the important machinery
on the nuclear envelope and

to see whether we could traffic
these things now to the nucleus.

So again this tells when
Michael injected and

now the quantum dots have trans
located from the side of class.

I'm here to the nucleus and
they're excluded from the nucleus live but

they're in the nucleus.

OK And the last thing we did
was to put a mitochondrial

localization sequence on this.

OK So this is the mitochondrial
localization sequence these are the cue

dots that are have a punk
tape cytoplasmic staining and

weak coal like localize them
with a miter tracker die that's

commercially available and
this is the merge.

OK We've then shown that you can watch
these under continuous exposure for

eight minutes with relatively little photo
bleaching whereas the might attack or

bleach is out in about fifteen seconds.

So we think that this is at least a first
step towards trying to realise some of

the potential for
these things as subsoil ular dies.

OK And this is my last slide and
I just threw it in because I wanted to

share the data and it really doesn't fit
in my talk so I apologize but I hear it.

Is that we've used the same approach and
our idea was to actually use.

The same approach to target
tissues in vivo OK so

this was a study that we published
last year with a curious lobby

at the Burnham Institute and
he has a set of homing peptide.

It's that he's derived using
phage display technology

that can home to the end of feely
I'm tumors in various tissues.

So what we did was label red Q.
dots with a tumor homing peptide green Q.

dots with normal long and Ophelia peptide

Cole inject them in the animal and
then asked whether they would sort.

These are images where the regular
outs of sorted to the tumour and

agreed to the one.

So this is a part of a new three year
project that we we've just had funded

to build smarter nanoparticles But
don't just glow.

So OK So hopefully what I've shown you as

some of our work on engineering functional
tissue therapeutic scientific and

technological applications of these
things in general from a tissue

engineering perspective I think
it's clear in our system and

in many others that indeed can
enhance tissue function in vitro.

I show you examples of cell
cell interaction and oxygen.

I think it's also true that micro
nanotechnology tools that have been very

well developed in other industries can be
leveraged in cell and tissue biology and

violence nearing here we've shown you
examples about how one might use them to

study fundamental tissue biology how one
might go about building three dimensional

tissues how one might think about
building cell based bio chips and

how one might target tissues and Bebo.

So with that I'd like to thank you for
your attention and thank my collaborators

and my group and our generous funding
sources happy to take questions.

Thank you for the work.

You.

Talked about who will.

Well you look at those like.

So that's a good question and it's I think
a recurring one lately because there

are three dimensional model
systems even in vitro so

the functions that we measure
the cell Autonomy's functions.

So for
example albumin people at fifty metabolism

we see very high levels
of those functions.

It's very clear that what
we will never see is some

of the three dimensional functions of a
pad of sites and Bebo for example the bile

tree the biliary tree that emerges from
a three dimensional configuration.

So that for sure is missing and

I suspect other things that arise from
that which is plenty dependent function

very like your company we are your

world are you OK.

There's a low level

which OK.

So question number one.

So are we haven't extensively explored
that and the reason for that is there is

an expert in this area is a German fellow
by the name of younger men who spent quite

a bit of time in static model systems
over the years looking at the dynamics.

So his data suggests
about six hours is is is

that a time frame required for
protein expression

changes we were actually really
interested in hypoxia inducible factor

as a potential mediator of this and spent
a few years with a chance Unix system.

I'm trying to look at that and it's looks
like it's not if mediated but have might

be doing other things and that's actually
a faster time course than the six hours.

So there's a subset of another
oxygen responsive element.

We have not let the reversibility you'll

hold on to.

So I showed the bench data so

the viable cells a viable have had a sides
were early on those are twenty four hour.

V. I mention in the one side that word
perfuse ing these and buyer reactors.

So we're just now doing that.

So the three dimensional constructs the
good multi day experiments are ongoing.

Or he will grow.

He told

me so yes.

So we've started with this very simple
model system and actually another problem.

He didn't mention is that we've
got a very tight pore size.

So we're using a three thousand
molecular weight peg and

we need actually pretty large proteins
to come out of that patricide So

we actually want to loosen the polymer
network without compromising

mechanical integrity which is why we're
thinking about the M.M.P. system so

that we could be in putting down Matrix
and degrading the polymer system in

concert over time so tuning the transport
to have had a site surfaces

sort of a Paramount design feature of the
next step and we're just getting started.

Yes so I didn't show the data we
have done that actually so but

we've done it in the extreme.

So we did a pad of sides under hypoxia
versus normal the presence of absence of

him if and
in the presence of absence of in P. for

fifty inducers because we're interested
in the crosstalk between P four fifty and

poxy other than the literature and then
we've gene expression profile those and

got some candidate genes.

I haven't looked at the piece still
too specifically we see a lot of cell

cycle stuff.

So we're kind of interested in whether
this might be a trigger for regeneration.

There's a lot of literature on
a pup Whosis in the liver of

acids meeting a pop to us and it's a very
active field in their biology so you know

sort of the side of my brain is interested
in what might be going on there.

We haven't looked at a pub so supposably.

So

yes so the two things that
come to mind with Decker and

that we're looking at are T.G.F.
beta because T.G.F.

beta binds Dekker and there's a lot of
clear interactions and then Dr Hughes

group in Singapore is actually report of
the role for two D.S. beta and Co culture.

So we think that hangs together
with our story the other thing

is that background binds to a non E.G.F.
binding site on easier for

scepter and we've been thinking that there
might be an interaction with E.G.F..

And Decker and

in the same way that there's an F G F Have
himself a pretty big like interaction.

So those are experiments
that we're setting up to do.

Those words like riddles.

So yes Auto for us is a problem with most
photo resists systems in the stem cell

biology example is a big problem
with our last staining step.

So we don't have a side
of toxicity problem.

We used to in the ME started
leeching the surfaces.

We do an overnight leeching step and
super Roy has recently

published by materials paper looking at
the various federal system what you need

to do to make them back compatible so
it looks like there's so

solvent that is a leach the one that can
make them make it more biocompatible.

OK Yes so

Rico incubate it with college and
knowing that it binds college in one

way that we never even tried it alone but
I presume that it wouldn't work.

OK Well thank you thank you thank you.