All right thank you Russ for the
introduction and thank you everyone for

showing up early.

I know it's not that easy to get up early.

These days with
Daylight Saving Time just starting.

So thanks for

being here today I'm going to talk about
the two unknown's of nucleus zones.

Formation and removal.

I know the title sounds
kind of ambitious but

don't expect too much
biology out of this talk.

So first as you can see
from the title of this

talk nuclear is going to be at
the center of this talk so.

First I'm going to give you some
introduction to nuclear as arms so nuclear

is all means particles in the nucleus and
physicists would like to name it.

Nucleon.

Because physicists like to use O.-N.
to name particles.

Whereas biologists like to use or
Amy to name particles.

So it just means particle in the nucleus.

So when I think of a nucleus what comes
to my mind is this very delicious dish.

So many of you recognize
what this is is basically.

Bacon Bacon wrapping a scallop inside.

And this is very similar to nuclear
bombs in many aspects right.

So as we all know nuclear
arms are basically

feeds on a string type of structure.

You can find and you carry a tick genome.

So you start from this naked D.N.A. which
goes through you clues information and

then higher order structuring inside the
nucleus of the cell until you get this X.

shaped chromosomal D.N.A.

or the biggest difference between
the two is that this of course is.

Protein wrapped in fat or is this this
protein wrapped in nuclear gas that right.

Some people tend to think this.

Protein was wrapping the scallop but
I think it's just fat and salt.

So I lot of the talk.

I'm going to divide my talk
into two parts the first type.

First part is going to be about D.N.A.

looping which has to do new because
information in the second part has to do

with how gene expression can be achieved
from packaged D.N.A. and therefore we

are going to introduce a molecular
mechanism of new clues on removal.

OK So the first part D.N.A. looping.

So before we move on.

I like to introduce the physical
concept related to looping and

that is called the persistence
length which is the length scale

that is used to characterize the bending
stiffness of appall them or so.

Imagine if you have a flexible polymer
like this you can put two points

on the contour of this polymer.

And then you can also draw
factors that are tangent to

the Contrave polymer at those two points.

And there is going to be
an angle between the two vectors.

And the size of this angle is going
to depend on how far away these two

points are along the contre the polymer.

So if you take cosign of this angle
it's going to decay exponentially

with some characteristic length
scale called the Persis and slang.

OK So L. is a contra length
between the two points and

the LP is of course is a slant which
is the property of the polymer.

Just to give you some examples
a single strand of D.N.A.

is known to be very flexible so
that means it has a very short persistence

slant which is about two nanometers and
that corresponds to seven nucleotides.

All stranded D.N.A.

because it forms a sparing between
two single strand of D.N.A. It's.

Stiffer than a single strand of D.N.A.
therefore it's Persis link this longer and

it's about fifty nanometers which
corresponds one hundred fifty Euclid.

To shorter persistence
length means flexible.

Larger persistent playing means to.

Some macroscopic examples.

Since this is a breakfast club seminar.

So this is spoken of the noodles and
if there are.

Hooked and
they're very stiff as you know purses and

slant this about ten to
the sixteen meters so that means.

You can you know.

Imagine a very long spaghetti noodle which
is on the older attend to the sixteen

is going to you know stay pretty
straight even on that long and scale.

And you can think about ramen noodles and
they're very curly.

And if you measure OK how much they curve.

They curve on the order of a centimeter or
so.

So you can think of this Persis to slant
the ramen noodles I want to centimeter

OK for the Persis is length that is
explained here is a little different than

the concept that I'm going to focus
during this talk because this refers to

the static persistent intrinsic curvature
you can find in this structure whereas

a person has this link here is
the thermal property of the molecule.

Some numbers are going to be important.

Throughout this talk so I'm going to talk
about the dimension of the nuclear zone

and bending energy
stored in a nuclear own.

So it's known that about one hundred forty
base pair of double strand of D.N.A..

Wraps around this protein
about two full turns.

So one point seven turned so
this is going to be.

The circum friends right the contour of
the D.N.A. that surrounds the protein

molecule So if it divided by two PIII you
can recover the radius of this particle.

That's about four point one nanometers You

can also estimate the bending
energy stored in this D.N.A.

based on this person's slang and
the radius of curvature.

So you apply this less to see theory and
then.

You achieve about seventy K.B. T.

Where is where Katie is Boltzmann constant
and he is the room temperature right.

So as you can see from this number
this is a very energetically

costly process you need about
seventy times the thermal

energy to induce spending
in this nuclear zone.

So that's why we call this very show.

Our D.N.A. Ben it's not likely to
occur just under thermal excitation.

So that also implies that
the interaction between the D.N.A.

and the histone surface is
going to be very strong.

So I forgot to mention that there
are four different colors in this protein

complex and that's because there are four
different kinds of histones that make up

this protein complex inside the nucleus
and they are named H two A H T B H three

N H four and there are two of
each in this new chrism particle.

So what's interesting on the eclipse
on is that they position.

Neither randomly nor regularly.

So if you take a genomic D.N.A.
and map the positions of the nucleus and

we find that there are some regions that
don't like to form new clues on where some

regions like to form a creature as
a physicist then we tend to think

this is because of the mechanical
properties of the D.N.A..

Are not uniform as a function of sequence.

So for example some D.N.A. sequences
are going to be stiff and some skin D.N.A.

sequences are going to be flexible and
that's why you see this

non-uniform nucleus information
as a function of D.N.A. sequence.

So we like to really investigate
whether this is true.

So I'm showing here you a plot that
combines two different data sets but

actually it's one data set that is
compared with theoretical prediction.

So what's shown here is new
clues on the affinity as.

Versus persistence lank.

So all wisdom and

SIEGEL A few years ago did a very
important study using synthetic D.N.A.

molecules so they assembled about
ten thousand different D.N.A.

sequences and mix those D.N.A. sequences
with his tongue proteins and measured.

The affinity of all these molecules.

So what's shown here on the Y. axis is the
affinity measured from that experiment and

this is locked to based so
from minus four to four here.

It's about four hundred fold different and
then on the X.

axis is the person's length of that D.N.A.
sequence molecule predicted

from an independent study that measured
persistence lengths of all possible.

Di nucleotides So as you know there
are four different alphabets in D.N.A.

sequence H. E C T.

It turns out that you can form ten
unique Di nucleotide sequences.

So you can basically take you
know more than ten D.N.A.

sequences and measure their overall
persistence length which is going

to be some sort of combination of
the persistence length of die and

you could types that compose entire
molecule based on that you can extract

the persistence length of
those die nucleotides.

I'm not going to get into detail
of how you do that calculation.

But simply speaking if you know
the sequence of the D.N.A.

you can count all the Dynaco times and
us retrieve the entire persistence length.

So that's what's plotted on the X.
axis and as you can see here.

You don't see a very strong anti
correlation which you would expect.

If you cruise information really depends
on the mechanical properties of the D.N.A.

so and to correlation means that
if you have a very soft D.N.A.

molecule you're likely to form
a new clue some very well.

Right.
So you

would see points distributed
along that axis.

However you don't see that points
are just scattered all over the place.

I'll give you one extreme example.

Right.
So you can pick the D.N.A. sequence.

That has the highest new clues among
finity right so that's why we name it.

New clues on D.N.A. and
you can also pick another point in the.

Exact opposite.

So we call that anti-nuke ism D.N.A.

because it has the lowest
nucleus of affinity.

So between these two sequences you
actually see that the model predicts

that this D.N.A. sequence which doesn't
form new clues on very well it is actually

softer than the D.N.A. sequence
which forms and it was a very male.

So here you'll see that
the model prediction and

the experimental data they
don't agree with each other.

So and it's known that this study also
admits that this model prediction

doesn't work well for repeat sequences or
sequences that have intrinsic curvature.

So as a physicist we like to
think of nuclear muffin a D.

as having something to do with
the intrinsic stiffness of the D.N.A.

So if you see a four hundred
fold difference in new clues and

affinity you can apply the Balts
one factor to extract the bending

energy difference between these
two Di Maalik through D.N.A.

molecules and it comes to be about six
K.B. to six times of thermal energy.

So the question is is there an independent
method to measure this number.

The difference in bending energy
between the two D.N.A. molecules.

So the methodology we use is called single
molecule florescence resonance energy

transfer some of you might be
familiar with this technique.

Let me just explain it very briefly if
you have a molecule that's on the order

of five nanometers to be more accurate
if there's a molecule that moves

on the order of five nanometers or so
you can put two different I molecules

and the same molecule and
they are called Dorner and

accept her and the reason one is called
donor and the other is called accept or

is because the donor Canuck
set contrasts for some of.

It's energy to the accept or.

And Dorner is the molecule that is
directly excited by the laser or

your excitation and
then it's going to fluoresce

by itself when the two die molecules
are far apart from each other.

Further than five nanometers
Typically however when the distance

between the two die not to die molecules
becomes smaller than five nanometers and

like I said there is energy transfer
from the donor and accept are there for

the acceptor starts emitting
fluorescents again.

So by looking at the relative
fluorescence in testes you can extract

what's happening to the distance
between the two die molecules.

So this as you can see is
a very powerful method to

monitor conformational changes of
viral molecules at nanometer lengths.

So this is our experimental
design which is very simple.

So we're going to take about one hundred
and ninety based here along D.N.A.

molecules which is about
sixty nanometers and

you can compare that to the person's
languages about fifty nanometers right so

this you know molecule is pretty short and
it's pretty stiff.

And once in a while this D.N.A.
molecule is going to form this loop.

So a perfect technique to detect
this is obviously fluorescence or

isn't a stage of transfer.

So you put to die molecules
on both ends of this D.N.A.

molecule and then when the D.N.A.

is not looped meaning the distance
between the two by molecules as large and

read only see more since coming
from the donor molecule.

However when the D.N.A.
molecule loops then the distance between

the two is very short so we start seeing
fluorescents from the accept or molecule.

OK Now the trick here is
that even when the D.N.A.

molecule forms a loop it's
going to be very short lived.

So these two ends are going to
diffuse away and then a second.

So we cannot really detect
whether a loop formed or not.

So we enter.

Complimentary sequences here so that when
the two ends come close to each other.

There's a high probability that this
space pairing is going to form for

that's how we stabilize
this group state to detect.

And this is just to show
you that we have this.

Experimental protocol to
construct any arbitrary D.N.A.

sequences that have all
the necessary elements for

surface immobilisation and Fred detection
right so we start out with this D.N.A.

sequence that we are interested in and we
perform two separate P.C.R. reactions and

then we melt the D.N.A. and do a cross
strand exchange and then cool it.

And this is your final product and
it has a bio ten which is what you use for

surface immobilization and the accept or
die molecule at one end and

donor molecule at the other end with this
complimentary base pair in the region.

And the set up is very simple
we use a standard total internal

reflection my cross copy to detect

red signal coming from single molecules
immobilize on the surface and

this is the signal that we're going to
expect to see as a function of time.

Depending on whether the D.N.A.
salute or not.

So if the D.N.A. is looped.

We expect Hi Fred.

Signal in the D.N.A. is on looped we
expect a little Fritzy So this fret

fluctuation tells you the looping kinetics
of this D.N.A. molecule in real time.

So as I mentioned before we're going
to compare two different D.N.A.

molecules which have opposite new
chrism affinities one is called

anti-nuke wisdom D.N.A.
and the other is called nuclear D.N.A..

As you can see these are repeat sequences.

So these are synthetic D.N.A. molecules
that sequences are obtained from this

landmark paper from wisdom and
SIEGEL In two thousand and nine.

Now what's different.

Between these two sequences is.

In the anti-nuclear D.N.A.

You see these eighty richer regions
which are about five base pair along.

So they.

They're totally composed of A's and t's.

As you know they have very different
properties then G.'s and Cs So

overall we can say that
the eighty content in the D.N.A.

is much higher than that
in the nuclear D.N.A..

So this is our typical images as you
can see when donor intensity is high.

Except when to see low and
when donor it is low.

Except in Tennessee.

Hi.
Right.

So this anti correlation means
that the there is fret fluctuation

as a result of distance change between
the two ends of the D.N.A. molecule.

So from these fret

signal fluctuations we can extract
times spent in these different states.

So there is a loop state and there is on.

Loop state.

So you basically take all these times
of the events that are observed in

either the loops or the on loop state and
you can do sums of sort of analysis which

is very standard in single molecule by all
bio physics called all time distribution.

So you just take all these dull times and
pot a histogram OK so

how many events are and a given window.

And normally in the very For
example in a two state

system you would see a single exponential
decay from one state to the other.

However when you try to fit this to
a single exponential it doesn't fit very

well it doesn't fit even to a double
exponential we see that if it's well to

a stretch experimental and we have models
to explain why that is but I'm going to

I'm not going to get into the details
because it's a topic of another talk so.

What's important is that we can extract
the mean time spent in each state and

those are called lifetimes.

So what's plotted here is
the lifetime of the D.N.A.

molecule in the on loop state and
this is the lifetime of the D.N.A.

molecule in the loop state as you can see
the lifetimes of anti-nuclear the D.N.A.

and you clues them D.N.A.
are different in those two states so

this is a good sign because now we
can use this kinetics measurement

to say something about the differences
between these two D.N.A. molecules.

So I'm going to come back to these
figures later on during the talk but for

now let's just focus on you know
what sequence feature gives rise to

this difference as I mentioned and
tiny because of D.N.A.

has these eighty stretches
repeating every ten base pairs and

it's known from this
large scale study that

all among all five Mars among all
fibers that you can assemble it is

these fires that are very rich maize and
teas that don't like to form new chrism so

very well so we thought that maybe these
eighty structures are very important.

So what we did is we switched
the middle base of these eighty

stretches from tea to G..

So we don't have those
eighty stretch anymore.

And we did the same kinetics measurement.

So first of focus on these two

our original sequences anti-nuclear zone
menu cuisine and tiny Corazon in this.

OK So those were the sequences that
we tested a few slides ago and

then we are adding this new modified
sequence where we change the T.

in the eighty's stretch to a G. right so

this is labeled as anti-nuke wisdom teeth
to a cheek as you can see from this plot.

Now this modify.

Sequence behaves more similarly

to the new clues on sequence than
the anti-nuke wisdom sequence.

So the key point here is that
although overall sequence.

Is similar to the anti-nuclear D.N.A.
in terms of behavior.

It is now much similar to a new
quiz I'm doing so that means that

whatever they say teach stretch
does it's very important

in determining the kinetics of D.N.A.
looping

So now let me go back to
these lifetime parts and

we can compare the anti-nuke wisdom D.N.A.
and Euclid I'm doing and I'm going to I'm

going to conclude something about
the stiffness of these two miles.

OK so how do you relate stiffness
to the time it takes for a D.N.A.

molecule to loop right.

It's very simple in the elastic regime.

If you have a very stiff rod for example
it's going to be very difficult to form

a loop so that corresponds to a very
long life time in the on loop state

meaning it takes longer to form a loop and
that's what we can use to conclude

something about the stiffness difference
between these two molecules right.

So you compare the life time spent
in the open or the on loop state

the new clues in D.N.A. spends more
time than the anti-nuke assumed.

Right.

So what can we say about the flexibility
of these two molecules will we can say

that the anti-nuke ism D.N.A. is more
flexible and this is a complicated way of

saying that if you have a stiff molecule
you basically have a very very curved

harmonic potential right so
you have this steep slope

which raises the energy barrier that you
need to overcome to get into the loops.

And remember the loop status stabilized
by the bass player interactions so

that gives you another harmonic
potential energy well here.

So the difference between
a stiff molecule and

a flexible molecule is
basically the height

of this energy barrier in a flexible
molecule the energy barrier is lower.

Therefore it takes less time for
you to cross over the barrier.

So from that first plot we conclude
that anti's because I'm D.N.A.

is more flexible and makes sense.

However when I give you the second plot
which is the lifetime of the D.N.A.

molecule in the loop state.

OK First of all some of you.

My question.

OK Why is the lifetime of the loops eight
any different between the two D.N.A.

molecules because the loop state is
stabilized by this based pairing region

which is commonly shared between
these two D.N.A. molecules.

OK So the answer is because when
the D.N.A. molecule is looped there is

of restoring force stored in the D.N.A.
molecule due to bending right.

So this is like.

You know closing a very
stiff spring into a loop and

the loop is going to try to open up
because there's this restoring force.

So if you think about this
restoring force then you can

figure out that a stiffer D.N.A.

molecule will tend to open up faster
because the restoring force is larger.

So based on this plot.

Anti-nukes them D.N.A. spends less time
in the loop state which means that

it's actually the stiffer of the to write
and tell you some D.N.A. It moves faster.

And again you can introduce this energy
landscape argument a complicated way of

explaining this.

So you have this energy well rich
representing the base pairing interaction

and for a stiff D.N.A. molecule then
the slope is very stiff here which

basically corresponds to a lower
energy barrier to cross over.

OK so which leads to faster looping.

So based on that second
plot then we have this

result which is anti-nuke
with M D N A stiffer.

Now if you have an attention span
that is longer than three slides.

Then you remember that I just said
that antennae because of D.N.A.

is more flexible three size ago and

now I just said in the previous lie that
is stiffer than the new custom D.N.A.

So what's happening and tiny because
I'm D.N.A. can be a more flexible and

stiff at the same time than
the nucleus of D.N.A. right.

So we hypothesize that maybe
the initial starting point

in looping transition is not the same
between these two D.N.A. molecules.

OK so what that means is you
know we assume that the D.N.A.

in the state is straight all
the stars from the straight state but

maybe one of the two molecules
is intrinsically curved so

it starts from a different initial point.

Right.
So you start from

a point where the distance between
the two ends is closer and

maybe that affects
the looping time we measure.

So yes sumption or
the hypothesis is that the stiffer D.N.A.

molecule is also more curved and so even
though it's stiffer because this is B.

did two ends is closer you
can still move faster.

Contrary to what I just explained.

Few slides ago.

So we set out to test this
by using gel electrophoresis

the idea is that if you run jelly
periphery cysts at very low temperature.

Then you can say belies a D.N.A.
molecule in the open statement so

we're not really looking at the effect
of bending here we're just looking at

the intrinsic curvature or permanent
curvature of the D.N.A. molecule.

So although all these three D.N.A.
molecules have exactly the same length.

Re see very different to all mobility and
this is very well known.

That gel mobility represents
the intrinsic curvature of the D.N.A.

molecule so
the molecule that runs slower on the gel

is more curved than the molecule
that runs faster and

so this the curvatures we can derive
from this shell mobility shift as

so conclusion is basically anti-nuke
wisdom D.N.A. is more curved.

And although it's stiffer at
the same time you can loop faster.

And we do some basic calculation.

So starting from this
Focker Planck equation.

OK Before I go on to the details.

So what do we what do we learned so
far we learn two things we have two D.N.A.

molecules and we measure the looping
times the time it takes for

the molecule to live we
also measure curvatures

of these two molecules one is
more curve than the other.

So based on these two measurements
Cammy then extract the expected

bending stiffness of these two D.N.A.
molecules right.

So from looping time and curvature can
me pretty can we extract the bending

stiffness and there's a way to do this and
it's called this mean first passage time

calculation and it starts from
this Focker plot in question or

which is also known as small it shows
you some small chose to keep creation as

basically describing diffusion in
the presence of a potential And so here.

The time evolution of probability
density is going to be related to

the variation in the probability flux.

J..

OK so this equation is called the.

Constituent of equation and then you can
relate this problem of the flux of course

to diffusion term and
potential term in the absence

of potential This is just random walk and
then you have this potential because that.

D.N.A. molecule has bending energy as
a function of end to end distance between

the right so
using these two equations then you can

extract this mean first passage of
time as a function of all these years.

All of this is a three dimensional
diffusion if you use the Enter and

distance as your variable you could
transform it into a one dimensional

problem and this is of course classical
war published many years ago by Shelton.

What we don't know however is
this potential energy you X..

So potential energy of this D.N.A.
molecule as a function of Danton distance.

And the reason it's very complicated
is because although I'm.

I keep describing D.N.A.
looping as this right.

I mean this is not all that's happening
right you can get looping like this.

And so
there are all possible confirmations OK So

even with the same and in this sense you
can imagine that there are going to be

many states which means that there
is entropic effect coming into play.

However because the D.N.A.
molecule we are looking at is very short.

Well it's not very far but it's on
the same order as a persistent Slank so

we can assume that it's pretty stiff
meaning that it's not going to go through

many curves and with goals so we you know
assume that the molecule is going to be

pretty the curvature is going
to be uniformly distributed and

then this is the trajectory is this
going to take most of the time.

And if that's true then the potential
energy could be expressed as a function of

this angle theta.

Right.
So it's basically harmonic potential and

then because we need to introduce
intrinsic curvature right the starting

point is different.

So that's estate a not which
is a practice from the theta.

And then theta is related to the distance
between the two points which is X.

by some basic knowledge of.

So we can then then approximate

the potential energy as a function of
entering distance in this simple form.

And we plugged it in and
do some number crunching and

I can tell you that as a result of
this we find that entitles him D.N.A.

is one point four fold safer than
the nucleus and D.N.A. which makes sense

based on how we name these D.N.A.
molecules so anti-nuke was in D.N.A.

is the D.N.A. molecule that doesn't like
to form new clothes on there for it makes

total sense that it is the stiffer of
the two by a factor of one point four.

And we already know that based on
the average persistence length of D.N.A.

molecule the author of the energy stored
elastic energy stored in the new Quizon is

about seventy times the thermal energy and
then using these two numbers we can

estimate the bending energy difference
between the end because I'm Danny and

the new clothes I'm getting which is about
seven point four times the thermal energy.

And then you apply the Boltzmann factor
and you get at this number which is

pretty it's still larger than what's
measured right so the number four hundred

is the new nuclear some of the difference
between the two molecules.

But now we get this number seven
thousand eight hundred which is about

four times larger.

But again this approximation is not
very accurate so we expect that

this difference is not very surprising
from the viewpoint of an experimentalist.

But the key point here is that

all the new clues information is is
a pretty complicated interaction.

It has you know bending energy
electrostatic energy and

some cases there is might be hydrogen
bonds forming between the D.N.A. and

his own October because still narrow this
down to the elasticities problem right.

So even if you only.

Consider the less the energy
stored in the D.N.A.

molecule you can reach
the similar order of magnitude.

And so this is a summary of the first
part we measure D.N.A. looping K. X. Y.

single molecule fret and the length of the
D.N.A. molecule is about a hundred ninety

base pair law and that length scale
looping time is sensitive to curvature so

you cannot just ignore curvature fit
to take both into account when you

try to extract the looping time and
I also showed you that this is a stretch

is seems to be doing something
funny because it's very stiff and

curve than other normal sequences and we
would like to understand more about this.

And then I also did some approximation
to show that the nucleus

of affinity can be in part explained
by the estimated bending Stephanus.

A second part of my talk
is moving one level up.

So I'm going to talk about how gene
expression occurs from package D.N.A.

and how this is related to you.

Clues on formation.

Well how this is really
to new because a removal.

So what is gene expression and all of you
know central dogma Molik killer biology.

Which tells you that the information flows
from D.N.A. through messenger R.N.A.

to protein.

So if you have a gene on the D.N.A.
then through these two processes.

You can synthesize proteins.

So these two processes could be lumped up
as one process called gene expression.

So that's the term that I go going to
use throughout the rest of the talk.

You might question how is gene expression

initiated it's initiated by these
transcription factors which are other

proteins that are transported
into the nucleus under stress.

So genes are not always turned on.

Some genes are off and

they are only turned on in response
to some extracellular stress.

So for example if you start the cells for
phosphate and

there are transcription factors are called
Full for which enter the nucleus and

bind the D.N.A.
to activate gene expression.

So this is how we physicists like
to approximate gene expression.

Transmission factors coming in and
proteins going up and.

So you can define this and put up
a relationship or gene expression or

input is the transcription factors and

the output is the protein synthesize from
the downstream gene and you'll know that

there is a promoter region upstream of a
gene where transmission factors interact.

OK so this relationship.

Known as gene expression pattern or
gene regulation function is going to be

a specific function of this promoter
which means that if you swap this

problem here with some other provider then
you're going to see another relationship.

So you might ask what determines
the gene regulation function.

Well it's going to boil down to what
happens at this promoter level between

the transcription factors and the D.N.A.
how transcription factors interact with

the D.N.A. it's going to determine the
shape of this gene regulation function.

So what determines
the interaction between the two.

Well first the amount of
transmission factors right the cost

ration of transmission factors are going
to determine the occupancy of transmission

factors and also location of transmission
factor binding sites the ones

that are located closer to the gene will
have a stronger effect on gene expression.

And the affinity of binding sites
right not only the concentration of

transmission factors but
also the affinity.

How tightly the transmission factors
interact with those sites is going to

determine the gene regulation function
as you know the amount of transcription

factors is called trance factors because
it doesn't reside on the same D.N.A.

as a gene or as these two factors called
sis elements because they're on the same

D.N.A. as a gene itself but
now knowing that D.N.A.

doesn't exist naked in
the nucleus of the cell.

We have to consider the effect of
nucleus right because in the end.

New clues are going to compete
with transcription factors.

Therefore knowing the exact
positions of new clues ohms

is going to be very important in
understanding this quantitative

gene regulation function so accessibility
of binding sites are positions of new

clues omes which are often ignored
have to be taken into account.

And these three factors we call
it the promoter architecture.

So our interest in it.

Interests is an understanding how
promoter architecture contribution

to a quantitative gene expression
pattern or the gene regulation function.

Says her interests in how new clues
illness affect gene regulation function

we need to know something about what
happens during transcription activation.

So Joe says particle physicists
had their standard model of

particle physics where you have the
standard model of transitional activation.

So this is a model promoter called Full
five The name is not very important.

What's important is that there are three.

Well positioned new clues
omes in the promoter region.

OK.

And what's shown in red red and

green are the transcription factor
binding side so sometimes they

can have different affinities because
they have different sequences and

then some transcription factor binding
sides can be located outside new clue.

Songs like this one and some are located
inside of new prisms like this one and

also there is the tato box where
transcription machinery assemble for

transcription initiation and
it's often covered by a new crew zone for

a lot of stress induced genes.

So what happens under stress is that
transcription factor in this case called

Full for it's going to be
transported into the nucleus and

it's going to recruit chroma
to remodeling complex for

the promoter with the first
transcription factor binding

site as he says in the new clues on free
region so it's going to bind there and

the new clues and complex it's going
to bind to the neighboring Ukraine.

What happens after that it's not clear
what the end point is that the Tata

box is going to be exposed to nuclear
arms are cleared from the promoter so

that general transcription machinery
including the aria plumbers can assemble

and perform transcription.

OK so how do we measure
the gene regulation function.

So we have to somehow be able
to measure quantitatively

the amount of transcription factors and
the amount of proteins expressed so

we use force and proteins to do this.

So the transcription
factor which is full for

has labelled with yellow fluorescent
protein and the downstream gene which is

activated by full four is replaced by
another fluorescent protein C. of P. and

the transcription factor input level
can be controlled by this drug called

doxycycline In other words this is placed
under all drug induced of all promoter.

So what we can achieve
from this system is that

we have a true color strain yellow for

some protein representing transcription
factor and put sand for us in protein rep.

Scenting the output protein and

we can continuously dial up the
concentration of transcription factors.

What you expect a low
transcription factor concentration.

Not much gene expression but
as you increase the transcription factor

concentration urban accumulate more and
more see if you see.

I forgot to mention that our
transcription factor is also special

it has mutations that make it
reside in the nucleus all the time

independent of the drug concentration.

So that's why this is par for the system.

These are typical images that
we get from our experiment.

We have another force of protein
that tells us where the nucleus is

inside the cell and
using that nuclear marker we can quantify

since intensity is
coming from the nucleus.

So this is why of P. which represents
a transcription factor a signal inside

the nucleus and this is C.F.P. which of
course permeates throughout the cytoplasm

so we measure single selflessness in
testes and we can put them on this

scatter plot gene expression level
versus transcription factor and

put them in so different
differently color points represent

points coming from cells grown under
different doxycycline concentrations

right so at very low doxycycline
concentration all points are here and

as you increase the docs I can
conservation you start to accumulate

points with higher transmission factor
level and gene expression that will.

So this curve looks very noisy and most of
the noise is real meaning it's biological

and has both extrinsic and
intrinsic noise contributions.

But the mean values.

Are still meaningful so
you can try to fit this to a function.

Which is very often used by a biologist
called the hill equation and

this is useful because it gives
you two important promenaders So

one is called threshold which basically
tells you the amount of transcription

factors you need to activate gene
expression fifty percent and

then it also gives you
maximum expression level.

So this could tell you about
the intrinsic capacity of promoter.

How much gene expression
level it can achieve.

OK So we are interested in the role
of nucleus on in gene expression and

I'll give you a very
specific example here.

So we generated this
model promoter which has

new clues in pretty region between
these four new clues ohms.

So there are two new clues on the left
to me because I was on the right

of course there are more nuclear arms
here and more inclusion is here and

I'm showing you only before you
cruise on is in the promoter region.

So what happens upon transpersonal
activation as transfers factor is going

to bind here and

what's known from previous studies by
other people is that you can get removed.

Starting from the nearest
neighboring region.

So these two new clues of the removed and
then these two are removed and

then you have been exposed
how the box region.

The question that we got interested in is
the role of this particular because I'm

here and the reason is because
this nuclear has an obvious role.

It's covering the top box region so

it's down regulating gene expression
under no stress the situation

right over this new prism here is
not harboring anything inside.

It's not blocking the transmission factor
minus light nor the top of us region.

So we were questioning
the role of this nucleus.

So what what is this really doing
in terms of gene expression.

And generally people in the field
think that more new clues omes course.

Spawn to Lord gene expression level and
this makes sense right.

If D.N.A. is highly package then.

It's going to give hard time to
transcription factors accessing it.

Therefore you expect to get Gene
a lower gene expression level and.

So we like to investigate
this using this model system.

So how do you study a function
of something in biology.

Well you need to delete it and
see what the effect is right.

So that's what we are going to do.

There are different ways of
removing a nucleus on the right so

if you have this model promoter one way to
remove this nucleus zone is to basically

cut out the whole D.N.A. underlying the
nucleus and so you can do East genetics.

There basically remove the entire D.N.A.
which spans this new clues on so

as a result you're going to get this from
water which doesn't have a new cruise ome

intervening between the transcription
factor binding site and

the thought of oxygen.

So this we call excision.

The other way of removing any clues ohm
is obviously you can change the D.N.A.

sequence underneath this nucleus.

And if you remember the first part of my
talk I told you that if you have a lot of

A's and t's in the D.N.A. sequence.

It makes the D.N.A. stiff.

Therefore it doesn't form you
because it was very well.

So that's what we're going to do.

We're going to increase the eighty
percentage of this sequence which

will destabilize Euclid's information so

effectively you can get rid of the nucleus
on without getting rid of the D.N.A..

But these are two different ways
of removing the new clues on.

And of course the it comes from this fact
that I just told you if you increase

the G.C. content and increase new quiz I'm
occupancy and this is also shown by this

genome wide computational study
by Hughes group at Toronto.

OK So the first method is excision So

if you excites this D.N.A. They're for
moving the nucleus on what happens.

So what's shown here is gene expression
operate as a transmission factor input

which was defined as
gene regulation function.

So what you see is if you remove
this intervening nucleus and

you see that gene
expression level goes up.

And that sort of makes sense because
you cruise on in the end is something

that has to be removed from the promoter
for transcription only initiation.

So you can think of it as
an additional barrier for

Croma turmoil in complex to remove So
you know giving more challenges

to the current term on the complex will
slow it down make it more inefficient.

So that's why you see this change
special level increase and

we think that it makes sense.

However.

What's puzzling was that when you remove
the new quiz I'm using the other way

which is to destabilize
this new clue on by

increasing the eighty
percentage of the sequence.

We saw that the gene expression
level actually dropped in on.

And so this was a surprise
to us at the beginning but

I should be honest that I actually
expected this to happen and that's why we

designed this experiment the first way and
the reason this drops could be understood.

If you know what's going on in
terms of new cars in removal.

OK so this group of collaboration
between cloudy and Bartholomew.

There's some very nice in vitro studies.

So what's known about
current turmoil in complex.

It's known to be able to
remodel new clues omes but

remodeling doesn't mean
removing remodeling means that

you basically destabilize the his so
knocked him or from the surrounding D.N.A.

So you can now translocate the new
prison right so translocation activity.

Of commentary on the complex is known.

However how does that lead to
new clues a removal is not no.

So these people did this experiment
where they assembled a model and

you cruise on in vitro rice or.

You take a piece of D.N.A. you mix it with
his son October's a certain concentration

and you can form single
new clues on the stand.

And then they threw in chroma term
modeling complex which was purified

with A.T.P. because complex as
motor protein that burns A.T.P.

to achieve mechanical work.

The end result of this is you get two
different forms of new clues on D.N.A.

Sometimes you get new clues on
you find new clues on one end.

Sometimes you find it on the other end.

However what you do not see is
the removal of nucleus on you always see

that there are still new clues are present
on this D.N.A. again so they were puzzled

as to why pro-tem modern complex can
remove new clues omes and be evil

when they cannot achieve that in vitro so
they made the experiment more complicated.

So now they assembled two new
clues on the same D.N.A..

And they repeated the experiments.

So they add complex with A.T.P.
and they saw the same end result they saw.

New clues on one end or the other but
the net difference is different here.

Right because you start
with a different and

you clues on a template you have
two new clues omes to begin with

you have one you because I'm at
the end that means this reaction.

We don't know what happened but
it led to new clues or

removal micrometer modern complexes.

OK.

So the model is you need
more than one you could zone

to get removal of either of them.

And this is the model that it is
that is explained in more detail.

So if you have a promoter that has
three well positioned new clues And

this is a transcription factor binding
site and this is a Tarbox region and this.

Is the intervening You could own right.

Which we were questioning the role of.

So what happens is that in transcription
all activation transcription

factor binds here with the chroma
term remodeling complex and

this commentary modeling complex can
bind to the neighboring Ukraine and

Crimea and structure of the Carlton
from only complex is known and

the exact shape can be
simplified like this.

So it really has this cavity inside
which totally fits over the new

chrism OK So what happens next
is that there is translocate

activity of this complex which
leads to peeling of D.N.A.

from the histo not come inside
this comet or mind complex right.

So this extra length of D.N.A.
has to come from somewhere and

it will come from this
neighboring the NE so this D.N.A.

gets pulled in to the current
turmoil in complex.

Therefore this new quiz on which is
over the top of oxygen approaches

this complex between complex and
the nucleus and

that has to result in
some sort of disruption.

So what disrupt what gets disrupted.

We think is this new quiz on that is
covering the thought of oxygen not

the new clues on which is found by
the current turmoil in columns.

So the net result of this is that
this new quiz on March is Ford and

then the new chrism that originally
cover the top of oxygen gets removed.

OK so a ton of oxygen is now
covered by this new craze own and

then if you keep repeating this process
then at some point the promoter is going

to be cleared of this new put them and
you have a total box region exposed.

So I like to think of the mechanism across
her mind complex as this bulldozer.

So basically this new clues ohm is
like a wheel is so real where D.N.A.

can turn around like this and

the chrome tear mine complex it's like
a shovel so assists on this wheel and

then it's going to plow through
all the new prisms sitting ahead.

Which leads to new clues and removal.

And this similar model
has been proposed in fact

four years ago by Roger Kornberg lab.

So what I'm trying to conclude is

that new clues ohms can sometimes
promote gene expression.

On the contrary to what
people generally believe.

And the reason why we saw the drop

in gene expression level when we
destabilize the D.N.A. underneath.

This intervening Euclid zone is because
if you destabilize and you quiz them then

you create this empty D.N.A. region which
is about two hundred base pairs long.

OK And
you can think about the Persis length of

any which is about one hundred fifty base
for long and they're about the same order

of magnitude which means that this D.N.A.
is going to be pretty stiff.

So although the transcription factor
recruits the chrome undermining complex.

It's going to be located
pretty far away right.

With respect to the person strength of
the D.N.A. so that it will have a hard

time reaching this new crew zone which
is covering the part of us region.

So that's why we think this template
which doesn't have a nucleus on

it's not going to give rise to
very efficient gene expression.

Whereas if you have a nucleus.

That is formed in this region.

Then you efficiently bring these two
distant point close to each other so

new clothes on.

Although it's a barrier to be
removed it also functions.

As a wheel on which the current term Ali
complex can sit and perform its function.

So to summarize a second part of my talk.

I just introduce you how we measure
a quantitative gene expression

and I also explain to you
that the distance between

transcription factor binding site and
new clues on is very important and

that could be modulating by information so

we think that new clues on some times can
play an active role in June expression.

So with that I'd like to acknowledge
my group members who are very

essential for
me to carry out my research and especially

I gave you two talks today the first
part was done by grad student tone and

the second part by post at Ross is also
like my funding sources Georgia Tech and

Burroughs welcome fun and
thanks for attention.

Yes.

Yes very very good question.

So as you know all new clues omes
they have the same histone Octomom.

Right.
So it's the same identical his so

knock them are the only difference
is the D.N.A. sequence.

Now a D.N.A.
sequence is negatively charged all D.N.A.

molecules are nailers charge and
that's much much of the phosphate back but

not from individual basis.

So in terms of electrostatics it's going
to be the same in every nucleus and

not not much difference.

So that's why we think that this elastic
energy actually plays a very important

role in differentiating some affinity
between different sequence as.

Yeah yeah that's a very good question.

So to test this model
right one way is to OK we

shorten the distance between the two.

Yeah.
So if this is now positioned over here.

Then we should be able to recover the same
gene expression and there's going to be

some non-trivial distance dependence
here because this looping looping it

has this peak at some length some
distance between the two ends.

Yes So

that's an interesting experiment that
could further give support to this model.

Yeah

yeah.

So I mean yes I mean
that's a very good point.

Yeah.
So you could is all formation.

So and people have measured this.

Some call it something called J. factor
which takes into account both on rates and

offered And yeah so
the affinities going to depend not

only on the equilibrium ratio between
the two that's what you're saying right.

It's going to depend more on the on rate
or you think it depends on how it's felt

yes yes OK sure sure.

Yes yes.

Yeah.

Yes Yes Yes OK So

let me as I'm not going to be
able to answer a question.

But let me give you some thoughts so new
clues information what we study is D.N.A.

looking intrinsic looping in
the absence of his own proteins and

of course new clues information is
going to depend on both the on rate and

off rate because both rates depend on the
flexibility of the D.N.A. However we think

formation is mostly on rate driven
because you have this protein and

then the D.N.A. has to bend a little to
you know form additional context right.

You don't need to bend all
the way to form a loop.

You can just bend a little bit and

form new contacts along the surface
around the surface of the protein.

So once that contact is form which has to
be very strong we think that the off rate

for be negligible.

But yes

so.

So that's something that
I sort of explained so

often it is not something we measured you
chrism often in these measured by this

other people we measure the on rate and
off rate and

the on rate can be related to
bending stiffness of the D.N.A.

molecule by this first passage
of time calculation and

from that we extract the bending stiffness
rate from the extracted Bendix if this

we then calculate the new prism affinity
by applying the Boltzmann factor and

then we compare the two numbers and

it seems like the number we get is larger
than the number of those measured.

Yeah.

All right.

Her.