You are on page 1of 31

Contents

1 Introduction

2 The puzzle of DNA packaging


2.1 Worm-like chain model . . . . . . . . . . . . . . . . . . . . . .
2.1.1 Problems for you to solve: . . . . . . . . . . . . . . . .
2.2 Thermal energy vs Elastic energy . . . . . . . . . . . . . . . .

5
5
6
6

3 How proteins fold into a nearly unique structure in 3D?

4 Statistical mechanics (or statistical thermodynamics) in


ology: Prediction of number of proteins bound to DNA
4.1 If we know thermodynamics, we can predict many things:
4.2 Predicting number of proteins bound onto DNA . . . . . .
4.3 Entropy of the DNA-protein system . . . . . . . . . . . .
4.4 Free energy DNA-protein system . . . . . . . . . . . . . .

bi.
.
.
.

.
.
.
.

8
8
10
10
12

5 Diffusion, Einstein relation and Nernst equation


15
5.1 Diffusion: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
5.2 Einstein relation . . . . . . . . . . . . . . . . . . . . . . . . . 16
5.3 Diffusion of charged particles across a membrane channel can
create a potential difference: Nernst equation . . . . . . . . . 18
6 Life in salty water and at low Reynolds number
20
6.1 Inertia is negligible for objects in the cellular world . . . . . . 20
6.2 Electrostatic interaction gets screened due to the presence of
oppositely charged ions . . . . . . . . . . . . . . . . . . . . . . 21
7 Force generation and movement in biology
24
7.1 A simple polymer . . . . . . . . . . . . . . . . . . . . . . . . . 24
1

7.2
7.3
7.4

Critical concentration . . . . . . . . . . . . . . . . . . . . . . . 26
Force generation by a simple polymer . . . . . . . . . . . . . . 27
Actin: treadmilling . . . . . . . . . . . . . . . . . . . . . . . . 29

This is my quick attempt to prepare a soft copy of my lecture notes! I have


not proof read this, yet. So, read it carefully, as there could be many typographical errors. Dont trust any statement/equation until you understand
that the statement/equation is physically sensible!

Chapter 1
Introduction
Conventionally, biology is being thought (and taught?) as a descriptive subject. The aim of this part of the course is to demonstrate how one can
approach biological problems using physics and mathematics, in a quantitative way. The hope is that this will convince you that biology is equally
challenging and exciting as any other subject you have known, if not more
exciting and challenging!
One thing that puts off many of you, while learning biology, is the jargons
myriads of names that one has to remember. Here we will try to use minimal
number of jargons and names. Instead we will discuss a number of exciting
puzzles in biology. I start with assuming that you know one thing: All living
organisms are made of cells(well, assuming viruses are not living organisms!)

Chapter 2
The puzzle of DNA packaging
The size of a typical human cell is roughly 100m. A human being is made
of trillions of cells. Each and every cell in your body has a very long DNA
in fact every cell (with some exceptions) in your body contains DNA that
is two meters long ; yes, every cell contains two meters long DNA molecule.
What is amazing is that this 2-meter long DNA is packaged into a tiny space
a region called nucleus which is only (1/10)th of a cell, roughly (dimensions
10m). This leads to an immediate puzzle how is it possible to pack a
two meter long macromolecule
To understand how DNA is packaged, the first thing one needs is to realise
that, to pack in this small space, DNA need to bent and folded. So the first
quantitative thing one need to figure out is the energy needed to bend DNA.

2.1

Worm-like chain model

In a very simple model, one can imagine DNA as a long and thin filament (or,
simply, a curve) in 3D space. Any curve can be represented by a parameter
called arc length" (s) this is the length along the contour of the filament.
Let ~r(s) represent the position vector at any point s
~r(s) = xi + yj + z k
The local bending energy of the elastic filament, at any location s has to
be proportional to the local curvature. Given that curvature is
2
~ = d ~r(s)
C
ds2

~ 2 . You need to have the square of the curvature


the bending energy is Eb |C|
because, symmetry demands that whether you make positive curvature or
negative curvature, the energy to curve is the same. Then, the total bending
energy of the filament, having any configuration specified by a set of vectors
{~r(s)} is

2
kb Z L d2~r(s)
ds

Eb =
ds2
2 0
kb is the bending stiffness of the filament, and it reflects the material property
of the DNA in a given environment. For DNA, under physiological conditions
kb 200 1030 Jm.

2.1.1

Problems for you to solve:

~ of a circle of radius R?
What is the magnitude of curvature |C|
What is the bending energy needed bend a DNA of length L to a perfect
circle of radius R = L/2?
For what length L, will the above-mentioned energy be comparable to
thermal energy kB T ?

2.2

Thermal energy vs Elastic energy

(Please see the pdf file of the lecture uploaded)

Chapter 3
How proteins fold into a nearly
unique structure in 3D?
(Please see the pdf file of the lecture uploaded)

Chapter 4
Statistical mechanics (or
statistical thermodynamics) in
biology: Prediction of number of
proteins bound to DNA

4.1

If we know thermodynamics, we can predict many things:

In your thermodynamics course, you might have studied about Free energy,
entropy, enthalpy and so on. You might have also studied many different
formulas like F = E T S, G = F + P V and so on. What is the use of
studying all these ?
The use is that, if we know free energy of a system, we can predict
many properties of the system1 . Yes! sitting at your home, you can make
many predictions. Consider the following example:
Imagine a DNA that has N protein binding sites. We now add large number of proteins into the solution containing this DNA. How many proteins,
on an average, will be bound onto this DNA ? Will all the N binding sites be
occupied by proteins ? Can we predict the average number of proteins that
1

Of course, conditions apply! Knowing the free energy, we can predict only the equilibrium properties of the system

will be bound onto the DNA ? Yes, we can; if we know the free energy of the
DNA-protein system. How do we calculate the free energy ? Simple. If we
subtract the energy of thermal motion from the total energy of the system,
we get the free energy. As the formula says, free energy F is given by
F = E TS

(4.1)

where E is the total interaction energy of the system, T is the temperature,


and S is the entropy of the system. Product of temperature and entropy (T S)
gives the energy of thermal motion. Remember that at any nonzero temperature, all molecules are constantly jiggling in random directions (Brownian
motion). Entropy (S) is a measure of this random, disorderly, jiggling.
How do we calculate E ? If our system consists of charged objects, the
only interaction is electrostatic interaction, and hence E is the electrostatic
interaction energy. If we are dealing with a mechanical system (say, bending
a rod) E is the bending energy. If we are pulling a spring with a force f , E is
the sum of the spring energy ( 21 kx2 ) and the energy due to the force applied
(f x). Soon we will see how to get E for the protein DNA system.
Now, how do we calculate entropy, S ? Very easy. Just count the total
number of states that the system can have. For example, if the system is a
coin, it can have only two states head or tail (Let us denote the number of
states by . For coin, = 2). If the system is a protein, count all possible
shapes (3D configurations) that the protein can have. And if is the total
number of shapes (or states"), entropy is defined as
S = kB ln()

(4.2)

Why do we say that entropy of a protein is the logarithm of the number of


all possible shapes ? As we know, protein molecules are getting constantly
hit by the water molecules doing the Brownian motion. As a result of this
hit, protein monomers will start moving in random directions, and this can
lead to all possible shapes of a protein. And entropy is the measure of this.
Let us now get back to the example of proteins binding to DNA and try
to the prediction. As stated earlier, If you are given a DNA with N protein
binding sites, and if we add large number of proteins into the solution containing this DNA, how many proteins, on an average, will be bound onto
this DNA ? In other words, what will be the average protein density on the
DNA, when the system reaches equilibrium ?.

4.2

Predicting number of proteins bound onto


DNA

DNA is negatively charged. When positively charged proteins bind onto


DNA, the electrostatic interaction energy is at its minimum (You could convince yourself by taking an example: consider two oppositely charged particles; write down the Coulomb energy of these two particles. Convince yourself
that the energy is minimum when they are the closest.)
The only interaction between DNA and protein is the binding interaction.
Let - kB T be the binding energy of each protein. When all the N binding
sites are occupied, the energy is at its minimum (E= -N  kB T).
Now, how do we calculate the average number of sites that will be occupied by proteins ? For that, it is not enough to know the energy; but we
have to know the free energy, F = E T S; here S is the entropy of the
sytem.

4.3

Entropy of the DNA-protein system

: What is the entropy, when m binding sites, out of those N sites, are occupied by proteins ?
As we saw earlier, if we can count the total number of ways one can arrange
proteins on DNA, we can calculate entropy. Look at figure 4.1 to see an
example when there are two proteins (m=2), and three binding sites (N=3).
Now let us ask: how many different ways we can rearrange m proteins on
the N sites. The answers is:
N!
(4.3)
=
m!(N m)!
Entropy is given by,
N!
S = kB ln = kB ln
m!(N m)!

(4.4)

If we use the approximation that ln N ! N ln N N (this is called the


Stirling approximation), we can simplify this as
S = kB N [ ln + (1 ) ln (1 )]
10

(4.5)

Figure 4.1: DNA with 3 binding sites; this figure shows the three different
ways one can arrange 2 proteins on the 3 sites.

11

where = m/N is the density of bound proteins.

4.4

Free energy DNA-protein system

: Now that we know entropy, let us ask, what is the free energy when m
binding sites are occupied by proteins ?
When there are m proteins bound, the energy is given by
E = m  kB T = N  kB T

(4.6)

The free energy is given by


F = E TS
= N  kB T + kB T N [ ln + (1 ) ln (1 )]

(4.7)
(4.8)

Free energy per binding site, in units of kB T , is given by


F
=  + ln + (1 ) ln (1 )
N kB T

(4.9)

where  is the contribution from the energy, and ln + (1 ) ln (1 )


is the cotribution from the entropy. In figure 4.2 we have plotted these
contributions separately. Energy (red curve) is minum when = 1. But free
energy (blue curve) has a minimum for a different value of , which is less
than 1.
Let us calculate the minimum of the free energy; the free energy is minmum when
F
= 0.

 = ln
1
e
=
1 + e

(4.10)
(4.11)
(4.12)

This is the average equilibrium density. In other words, given a value of ,


we can predict a value of the average density.
For  = 2, the equilibrium density is 0.88; see Fig. 4.3. The energy is the
lowest when = 1. The entropy is maximum when = 0.. The interaction
12

S, E, F (in units of NkBT)

1
0.5

0
-0.5
E

-1
-1.5

F=E-TS

-2
-2.5
0

0.2

0.4
0.6
0.8
Protein density

Figure 4.2: The green curve is N kSB T = [ ln + (1 ) ln (1 )]; red curve


is N kEB T = , with  = 2. The blue curve is the free energy ( N kFB T ) given
by eq.4.9. For what density is the free energy minumum ?

13

S, E, F (in units of NkBT)

1
0.5

-0.5
-1
F=E-TS

-1.5
-2
-2.5
0

0.2

0.4
0.6
0.8
Protein density

Figure 4.3: Same as fig.4.2; but see the vertical lines : the pink line shows
the density at which entropy is maximum ( = 0.5); the black line shows the
density at which free energy is minum ( = 0.88). The energy is minmum
when = 1.
energy tries to take the system to = 1, while thermal fluctuations tries to
maximize the entropy a competition between energy and entropy. In this
competition, depending on the value of , the system reaches a density which
is minimum of the free energy.

14

Chapter 5
Diffusion, Einstein relation and
Nernst equation

5.1

Diffusion:

A small bead in water will undergo random Brownian movements. How far
does the bead move within a time t seconds? To answer this let us do the
following thought experiment. Leave the bead at position A in water and
measure the displacement (~r(t)) of the bead after a time of t seconds. Repeat
this experiment many times. Each measurement will give us a ~r(t), and one
would find that the average displacement
h~r(t)i = 0,

(5.1)

and the averaged displacement square is


h|~r(t)|2 i t.

(5.2)

Here the symbol h...i means average over many measurements. The proportionality constant is related to the diffusion coefficient. It turns out that, in
3 dimensions,
h|~r(t)|2 i = 6Dt
(5.3)
where D is the diffusion coefficient. (in d dimensions, the relation is h|~r(t)|2 i =
2dDt). The dimension of the diffusion coefficient is [D]=L2 /T. To quickly

15

show that the mean square displacement (h|~r(t)|2 i) is proportional to time


linearly, let us do the following calculation.
Now, let us consider a bunch of particles (say, protein molecules) diffusing.
Here by "protein molecules", I mean completely folded, globule-like proteins
they can be imagined as "spheres" for all practical purposes! Let C(x, t) be
the concentration of the protein molecules at location x at any time t. Here
concentration is defined as the number in unit volume (or length in 1D).
The protein molecules will diffuse from higher concentration region to lower
concentration region, and this flow of proteins (or the current, or flux) can
be written as
C
x
J~D = D
x
(Draw some concentration profiles, and convince yourself that this is true).
This is called the Ficks law of diffusion. Here the current, JD , is defined as
the number of molecules passing through unit area per unit time.
Concentration of molecules at any location will only change, with time,
if and only if the flow is space-dependent. If the flow is same everywhere
in space, at location x, the number of particles flowing in will be equal to
number of particles flowing out. Therefore, if you want C to change with
time, you should have JD chaining in space. This leads to
JD
C
=
t
x
~ J~D ]. The above two
[In 3D this equation can be written as C
=
t
equations can be combined to get the diffusion equation
2C
C
=D 2.
t
x
This is called the diffusion equation. Solving this equation one can get
C(x, t). Next question is, what exact is D? This question was answered
by Albert Einstein.

5.2

Einstein relation

In one of the most celebrated papers in physics, Einstein, in 1905, presented


the famous relation between Diffusion coefficient and viscous drag. And his
arguments were essentially the following.
16

Consider n particles suspended in a volume V containing a fluid having temperature T . Let there be an external force f~ = f ez acting on each
particle (e.g., in case of gravity f~ = mg
ez ; for simplicity we shall do the calculation in 1 dimension). In thermal equilibrium particles will be distributed
such that the concentration is given by the Maxwell-Boltzmann distribution
!

f z
.
C(z) exp
kB T

(5.4)

The diffusion of these particle will generate a flux, JD as defined above. On


the other hand, the external force f , will lead to movement of particles in
the direction of the force, with a velocity
~vf =

f~
.

(5.5)

where is the viscous drag. For sphere-like particles = 6a where is


the viscosity of the medium, and a is the radius of the sphere (or size of the
spherical molecule). The corresponding flux, is given by
f~C
J~f = C~vf =
.

(5.6)

The equilibrium is obtained when the fluxes are equal in magnitude but
opposite in direction. That is
Jf = JD ,

(5.7)

giving,
DCf
.
kB T
kB T
D =
.

fC

(5.8)

Problem 2 : Calculate the diffusion coefficient D of a spherical particle of


1 m diameter, in water.
D=

kB T
4.1 1021 J
=
6a
6 3.14 0.5 106 m 103 Jm3s
0.6 (m)2 /s.
17

(5.9)

5.3

Diffusion of charged particles across a membrane channel can create a potential difference: Nernst equation

Consider a box partitioned into two equal parts by a membrane. At t = 0,


the first part contains C10 concentration of K + and Cl ions (equal number of
positive and negative ions, so that the partition is charge neutral). The second partition contains C20 concentration of K + and Cl ions (equal number
of positive and negative ions, so that the partition is charge neutral). Assume
C10 > C20 . Then as the clock starts, the membrane is made permeable only
to K + ions. What will happen?
The + ions will from the first part will diffuse to part 2, because the
concentration in part 2 is less. The magnitude of this diffusive flow is given
by the equation
C
.
JD = D
x
However, since only + ions are flowing from part 1 to part 2, the number
of - ions in part 1 will be more than the + ions and will lead to a charge
imbalance. The part 1 will be more negatively charged, and part 2 will be
more positively charged. So the + charges in the part 2 will now be attracted
back to part 1 (remember only + charges can move through the channel).
This will lead to a flow of charges from 2 to 1. This flow is due to the
electrostatic created by the charge imbalance. This flow will be given by
Jf = Cvf =

fC

Noting that the electrostatic force is f = q d


, we get
dx
Jf = Cvf =

q d
dx

where is the electrostatic potential and is the frictional drag (see above).
As discussed above, the equilibrium is obtained when the fluxes are equal in
magnitude but opposite in direction. That is
Jf = JD ,

18

(5.10)

giving,
q d
C
dx
=D

x
Integrating on both sides with x, we have
Z x2
x1

Z x2
q d
C
dx
dx
dxD
=

x
x1

where x1 and x2 are two nearby points on either sides of the membrane (say,
x1 in part 1, and x2 in part 2). Substituting for D from the Einsteins relation
(D = (kB T )/), we have
(x2 ) (x1 ) =

kB T C1
ln
q
C2

(Check for errors in minus sign!). Here C1 and C2 are K + ion concentrations
at equilibrium in part 1 and part 2 respectively. This gives us an equation

19

Chapter 6
Life in salty water and at low
Reynolds number
6.1

Inertia is negligible for objects in the cellular world

Imagine a dead bacterium is put into water with an initial velocity v0 . How
far will it move? For simplicity, let us imagine a bacterium as a sphere of
radius a = 1m. The mass of the sphere essentially the mass of the water
inside (assume no gravity). The velocity will obey the equation
m

dv
= v
dt

This has the solution

v = v0 e m t
This tells us that the velocity will be nearly zero when t  m . Noting
that m =density.volume, and = 6a, we can substitute the density and
viscosity of water, to get m 107 s. When time is larger than the 100
nanosecond, the velocity will be nearing zero.
R
The total distance travelled by the bacterium is d = 0 vdt = v0 m Considering a reasonable number for initial velocity v0 = 1m/s, we find that
the total distance travelled by bacterium, d 1013 m. The total distance
travelled by the bacterium is less than the size of an atom!! This
means that, until you keep pushing, the bacterium will not move. To sustain
a nonzero velocity, you need to apply force!
20

Now, let us consider the bacterium under an external force fext (gravity
or any other kind of force). The corresponding Newtons equation is

fext
dv
+ v=
dt m
m

For typical parameters, since m very large, we will have situation where the
first term dv
negligible when compared to the second term m v. This would
dt
mean
fext
v=

This implies that, we have a situation where force is proportional to velocity!


In the cellular world, we are in such a regime where force is proportional to
velocity!
This regime, where inertial forces are negligible compared to the viscous
forces, is known as the low Reynolds number regime. Reynolds number is
defined as:
Inertial forces
.
Re =
Viscous forces
Life, in the cellular level, is typically at very low Reynolds number. If you
are enthusiastic in knowing more, google search for an article by E M Purcell
named "Life at low Reynolds number". Both the suggested books discuss
more details.

6.2

Electrostatic interaction gets screened due


to the presence of oppositely charged ions

When an electrically neutral big molecule is put in water, many small ions will
come out of the molecule and will wander in solution such that the entropy of
the system increases. For example, the chemical group COOH, in water, will
exist as COO and H+ , so that the H+ ion can wander around in solution
(reason: Brownian forces, entropy). Such small charges are typically called
counter ions. In physiological conditions, there is also salt: NaCl existing
as Na+ and Cl ; similarly K+ and Cl . But, remember, overall, the system
is charge neutral. That is,
X
qi = 0.
Imagine a big positively charged macro-ion (eg. a protein) given its charge,
many small negative ions (counter-ions) will be hanging around this big ion
21

the resulting system can be imagined as a big positive ion with cloud of
negative ion around it. How will this proteincounter-ion system will interact
with another such proteincounter-ion system?
Given that there are charges all around, what is the probability density of
finding a charge qi at any location? Given that this is a statistical mechanical
system at a finite temperature T , the probability density is given by the
Boltzmann equation
!
qi
Pi = A exp
kB T
where is the electrostatic potential of that location and A is a constant.
Now, to find out the potential at any location, one has to solve the Poisson
equation
C
2 =
0 r
where the charge density C is nothing but
C=

qi P i = A

X
i

qi
qi exp
kB T

where we substitute for the probability from the Boltzmann equation. Using
the expression for charge density in the Poisson equation, we get
qi
A X
qi exp
=
0 r i
kB T

The above equation is known as the Poisson Boltzmann equation. The solution of this equation will give you the potential due to the the charged
system, at any location, r distance apart. However, solving this equation is
very difficult.
 1, one can expand the exponential function as
When kq
BT
qi
exp
kB T

qi
kB T

This gives us
A X
qi
2 =
qi 1
0 r i
kB T

22

Noting that

i qi

= 0, we have
2 =

A X qi2
0 r i kB T

Rewriting the above equation, we get


!

2D

=
where
D =

v
uX
u
kB T 0 r
t

qi2 A

. You can immediately see that such an equation is likely to have a solution
r
exp
D


when one demands that the potential at r is zero. It turns out that,
this equation, when solved appropriately (taking spherical symmetry, etc),
one gets that the electrostatic potential due to the protein-counterion system,
at distance r apart is
 
exp rD
(r) = B
r
where B is a constant. This is called the Debye-Huckel potential. This tells
us that, unlike in the normal electrostatic system, the potential decreases
exponentially fast, and when r  D the potential is zero. This length (D )
beyond which electrostatic potential dies out is known as the Debye length
or the Debye screening length. In physiological conditions, the Debye length
is D 1nm. This implies that beyond one nanometer distance, in biology,
electrostatic interaction will die out. Because effect of any charge will be
screened" by oppositely charged
  ions that are wandering around! Therefore
exp

the potential (r) = B


potential.

r
D

is also known as the screened-Coulomb"

23

Chapter 7
Force generation and movement
in biology
In typical cells, force generation and cell movement are driven by certain
kind of polymers known as actin.

7.1

A simple polymer

First, let us consider a simple filament-like polymer (see Fig. 7.1). Assume
one end of the polymer is inert, and the other end can polymerise and depolymerise with a rate kon and koff . For example kon = 5s1 means the polymer
grows by 5 subunits every second. Similarly koff = 2s1 means the polymer
shrinks by 2 subunits every second. The net growth speed, on an average is
given by,
dhli
= kon koff
v=
dt
where hli is the average length of the filament (number of subunits on the
filament) and t is time. The unit of speed is subunits/second. In a simple model, we can imagine that the polymerisation rate will depend on the
availability of free monomers to polymerise such that
kon = k0 C
where k0 is the intrinsic rate constant and C is the free subunit (monomer)
concentration available in the solution (I will be using the term subunit and

24

Figure 7.1: A simple polymer, polymerising with rate kon and depolymerising
with rate koff at one end. The other end is assumed to be inert. The polymerisation rate could depend on the free monomer concentration C, floating
in the solution, such that kon = k0 C, where k0 is the intrinsic rate constant.

25

monomer interchangeably). This gives


v = k0 C koff

7.2

(7.1)

Critical concentration

Critical concentration is the concentration of free monomers below which


the filament would not grow. Just above the critical concentration, filament
grows with a nonzero speed. In other words, critical concentration is the
concentration at which the average velocity v = 0. From Eq. 7.1, we get
Ccrit =

koff
.
k0

Now, if we start with C > Ccrit , the filament will polymerise; as the filament
polymerises and grows, the free monomer concentration will reduce. Finally
the growth will happen until the free monomer concentration is equal to the
critical concentration.
Problems:
1. Free monomer concentration can be defined C = Nf /V where Nf is the
number of free monomers and V is the volume. As filament polymerises,
the free monomer concentration decreases such that
dNf (t)
= k0 C(t) + koff
dt
Taking V as a constant, using the relation C = Nf /V , solve the above
equation and obtain C(t). Substitute this result in the equation for
velocity and plot the velocity vs time. Note the velocity as t .
Can you integrate the velocity equation to obtain hli as a function of
time?
2. Now assume a more general case where both ends can polymerise and
+

depolymerise with rates kon


, koff
, kon
and koff
respectively. Here plus
signs represent rates for one end and the minus signs represent rates
for the other end. Since both ends are symmetric, thermodynamics
imposes a constraint that
+

koff
koff
=
.
k0+
k0

26

In other words Ccrit


= Ccrit
. If you think about it physically, you
will realise that, since these intrinsic rates depend only on the interaction between two monomers (how two monomers are bonded together),
these ratios have to be the same at both ends as long as the monomers
are of the same kind.

Now, write down the equation for the change in length (or average
velocity) of the whole filament
dhli
= v =?
dt
What is the critical concentration (Ccrit ) for the whole filament?

7.3

Force generation by a simple polymer

The polymerisation chemical energy of binding can be converted to mechanical energy force and work, by designing the system appropriately. For
example, consider the system shown in Fig. 7.2. A polymer is polymerising
and depolymerising against a movable rigid wall on the right side. At the
back (left side), it is supported by a fixed immovable rigid wall. Imagine
that you are applying an external force f on the movable wall on the right
(see figure) against the direction of growth. As the filament polymerises, the
movable wall will keep moving to the right. As the filament depolymerises,
the wall will keep moving to the left.
See the Fig. 7.2. To add a new subunit on the polymer, the wall needs to
be pushed by a distance d, where d is the size of a single subunit. In other
words, to insert a single new subunit, the work needed to be done is f d.
Since the force is against the growth, the polymerisation rate will decrease
as force increases. The polymerisation rate in the presence of the force can
be assumed to be
!
f d
.
kon = k0 C exp
kB T
When we put f = 0 we get back the old rate. For simplicity, let us assume
koff is independent of force. This gives us the growth velocity as
!

f d
v = k0 C exp
koff
kB T
27

(7.2)

External force
f

rigid immovable
wall support

Wall is free to move

Figure 7.2: A polymer is polymerising and depolymerising against a movable


rigid wall on the right side. At the back (left side), it is supported by a
fixed immovable rigid wall. An external force f is being applied on the
movable wall on the right, against the direction of growth. As the filament
polymerises, the movable wall will keep moving to the right. As the filament
depolymerises, the wall will keep moving to the left.

28

Maximum force that can be generated: As you increase the external


force, the velocity of the polymer will decrease. At some force value (when
f = fs ), the polymer will not grow anymore. This is the maximum force the
polymer can generate. This is also known as the stall force. This is given by
taking v = 0
!
fs d
= koff
(7.3)
k0 C exp
kB T
kB T
koff
ln
(7.4)
d
k0 C
fs is the maximum force such a polymer can generate. Cells use this kind of
force to polymerise and depolymerise.
fs =

Problems:
1. Plot force vs velocity according to equation 7.2
2. Equate v = 0 in eq. 7.2. Plot the resulting equation as f versus C.
Plot it with C in x-axis and and fs in the y-axis. The curve you get
divides the space into two. In one portion, for any value of f and C,
you will find that the polymer will grow. The in the other portion, for
any value of f and C, the polymer will shrink. This is like a phase
diagram with two phases. Shrinking phase and growing phase.
3. Assume that kon is independent of force, and the whole force is used to
remove (disassemble) the monomer. What would be the expression for
the force-dependent off-rate ? What would be the resulting stall force?
4. For the above case, plot force vs velocity.

7.4

Actin: treadmilling

Actin is a protein. Actin in solution typically binds to ATP (ATP is the


fuel for biological machines to work!) ATP-bound actin can polymerise and
become actin filament (often known as F-actin). Once polymerised onto the
filament, ATP-bound actin can change its chemical state (it will go undergo a
chemical reaction known as ATP hydrolysis) and become ADP-bound actin.
In this process, actin monomer also undergoes some structural change. This

29

ATP

ATP

ADP ADP ADP ATP ATP

ATP

ATP

Figure 7.3: At the right (plus) end: ATP-bound actin can polymerise with
T
T
rate kon
and depolymerise with rate koff
. At the left(minus) end, ADPD
be the rate of ADP-bound
bound actin can only depolymerise. Let koff
actin depolymerisation from left end (minus end). Assume that all other
possible rates are zero.
hydrolysis process happens only on the filament. Typically, it does not happen in solution. Therefore, you will have an asymmetric polymer where
one end has newly polymerised ATP-bound monomer and the other end has
ADP-bound monomer. It also turns out that ADP-bound actin is unstable
and prefers to depolymerise quickly. All these information can be summarised
T
and deas shown in Fig.7.3. ATP-bound actin can polymerise with rate kon
T
polymerise with rate koff from the plus end (right end) (Please note that plus
and minus are notations to distinguish both the ends. It has nothing to with
D
electric charges). Let koff
be the rate of ADP-bound actin depolymerisation
rate from left end (minus end). (see figure). Assume that all other possible
rates are zero.
Based on the what we did for a simple polymer, at the plus end, we can
write,
T
kon
= k0T C

30

where C is the concentration of the free ATP-bound actin. The net rate of
growth at the plus end is
T
v+ = k0T C koff
When
C=

T
koff
+
= Ccrit
k0T
kT

the growth velocity v+ will be zero. For any concentration C > koff
T , the right
0
end (plus end) will keep polymerising and make the polymer grow on an
average. However, since the left (minus) end has only depolymerisation, the
left end will always contribute to the polymer shrinking. The left (minus)
end will never grow. In other words, the velocity of the minus end v =
D
koff
(always negative). By varying the concentration, we can achieve the
condition where the plus end growth is equal to the minus end shrinking
D
T
k0T C koff
= koff

The phenomenon of plus end growth and minus end shrinkage (such that
filament length is constant on an average) is called treadmilling. The concentration at which this happens is known as the tread milling concentration.
Ctread =

D
T
koff
+ koff
k0T

At this concentration, filament, on an average will not grow. But note that
+
Ctread > Ccrit
. That means, at this concentration, the plus end will keep
growing, and the minus end will keep shrinking, on an average. (that will
give some idea on why we called it a plus end and a minus end!). However,
we have no net filament growth, but the centre of mass of the polymer keeps
moving. This is like a simple machine that can move!
Other parts we discussed during lecture 10 (that is not covered
here) will not be asked for the mid semester test

31

You might also like