Contact me in order to access the whole complete document. Email: smtb98@gmail.

com
WhatsApp: https://wa.me/message/2H3BV2L5TTSUF1 Telegram: https://t.me/solutionmanual
Exercises, chapter 2

2.1 A change in production rate. A gene Y with simple regulation is produced at a

constant rate pl. The production rate suddenly shifts to a different rate p2 .
(a)Calculate and plot Y(t). (b)What is the response time (time to reach halfivay
between the steady-states)?

Solution:
(a): Let's mark the time when the shift occurs as t=O. Before the shift, Y reaches
steady state at a level Y(t+)=YSt= P1 I a. After the shift,
(2.1.1) dYldt= p2 - a Y.
The solution of such an equation is generally Y = Cl + C2 exp(-a t), where the
constants Cl and C2 need to be determined so that Y(t=O)= 1 a, and Y at long times
reaches its new steady state P2 I a. This yields the following sum of an exponential
and a constant
(2.1.2) Y(t) = 1 a + (P2 l a -PI / a ) (l-e-"')= p2 l a+ (P11 a - P 2 l a) e-at
Take the derivative wath respect to time, dYldt, and verifjr that Eq. 2.1.1 is fulfilled,
(b) The response time. which is the time to reach half way between the two steady
states, is iog(2)lm.

2.2 mRNA dynamics. In the main text, we considered the activation of transcription of
a gene (mRNA production), and used a dynamical equation to describe the changes in
the concentration of the gene product, the concentration of protein Y: dYldt= 0 - a Y,
in which P describes the rate of protein production. In reality, mRNA needs to be
translated to form the protein, and mRNA itself is also degraded by specific enzymes.
(a) Derive dynamical equations for the rate of change of rnRNA and the rate of
change of the protein product, assuming that mRNA is produced at rate P, and
degraded at rate a,, and that each mRNA produces on average p protein molecules
per second. The protein product is degradeddiluted at rate a.
(b) Note that mRNA is often degraded much faster than the protein product am,,a (in
bacteria, mRNA lifetime is usually on the order of minutes (Bernstein et al., 2002)
whereas most proteins are stable for hours). Can this be used to form a quasi-steady-
state assumption that mRNA levels are at steady-state with respect to slower
processes? What is the effective protein production rate P in terms of P,, a, and p?
What would be the response time if the mRNA lifetime were much longer than the
protein lifetime?

Solution:
(a) The dynamic equation for the concentration of mRNA of gene Y, Ym, is
(2.2.1) dYm1 dt= Pm - a, Y,.
The dynamical equation for the protein product is due to production of p copies per
mRNA and degradatioddilution at rate a:
(2.2.2) dY/dt=pYm- a Y
(b) In the typical case that mRNA degradation is faster than the degradatioddilution
of the protein product, we can assume that Y, reaches steady-state quickly in
comparison to the protein levels. The reason is that the typical time for the mRNA to
reach steady state is the response time log(2)I a,, which is much shorter than the
protein response time log(2)l a because am>> a. The steady-state mRNA level is
found by setting dYm/ dt=O in Eq. 2.2.1 ,yielding
(2.2.3) Ym,st = Pm am
Using this for Ymin Eq 2.2.2 yields the following equation for the protein productior,
rate
(2.2.4) dY/dt=pP,/a,- aY
In other words, the effective protein production rate, which is the first term on the
right hand side of the equation, is equal to the steady state mRNA level times the
number of proteins translated from each mRNA
(2.2.5) P= ~ P m l a r n

In cases where am<<a, that is when mRNA is much more stable than the protein, the
response time is governed by the slower process, accumulation of mRNA. For each
level of mRNA, Ym(t), protein level 'instantly' reaches its momentary steady state
pYm(t)/a. In such a case the response time is log(2)Ia according to the same
reasoning as in the text for the protein response time. For advanced students, a &l1
solution of the dynamics is given at the end (not necessary for the solution required in
this exercise).

2.3 Time-dependentproductionand decay. A gene Y with simple regulation has a

time-dependent production rate P(t) and a time-dependent degradation rate a(t). Solve
for its concentration as a function of time.
Solution:
Verify by taking the time derivative that the following is correct:
(2.3.1) y(,) = e-ja't')dl[ y(0) + )ei~'t'Idf"
dt']

where all integrals are between 0 and t.

For example, for the case of problem 2.1, a(t) is constant over time so that
exp(-! a(t ') dt =exp (-a t), and l ( t 7 is constant after t=O and equai to P2 SO that:

and we obtain the desired result Y(t)=P2 / a + (Pi 1 a - P 2 / a) exp(-a t) .

2.4 Cascades. Consider a cascade of three activators, X+Y+Z. Protein X is initially

present in the cell in its inactive from. The input signal of X, S,, appears at time t=O.
As a result, X rapidly becomes active and binds the promoter of gene Y, so that
protein Y starts to be produced at rate P. When Y levels exceed a threshold K,, gene Z
begins to be transcribed. What is the concentration of gene product Z as a function of
time? What is its response-time with respect to addition of S,? What about a cascade
of three repressors? Compare your solution to the experiments shown in Fig 2.7.
Solution:
We will assume all proteins have the same dilutioddegredation rate a. After
induction, Y is produced at rate
pyand degradedldiluted at rate a :
(2.4.1) dY/dt=pY-aY
yielding the familiar exponential approach to steady-state:
Y(t)=bla (1 -exp(-at))
Assuming a step function for the activation of gene Z by Y (logic input function),
transcription of gene Z starts at time TYZ when Y(ryz)=Ky:

where Yst=pyla. Just for extra clarity, let's consider the limits of (2.4.2) to see if this
makes sense. When Ky<<Yst, Yst-Ky-Yst and rY=-+O. In this case the threshold for
Z activation is low, and Y levels cross it very fast. Conversely, if the activation
threshold Kyis very high, approaching Yst, Z is never activated because Yst-Ky+O
and r y z + m .
Production of Z starts after time t=ryz at a constant rate of@:
(2.4.3) dZ/dt= 0 (t<~yz)
pz-az (t'5z)
Solving this we get:
Z(t)= 0 (f<7yz)
PJa (l -exp(-a(t-TYZ)))(t>rYz)
Solving for the response time, the time to reach half of the steady state of Z:
(2.4.4) PJa (l - e ~ p ( - a ( t , , ~ - ~=~112
~ )PJa
) ) =,
t,,*=~~z+log(2)/G
Hence, there is an extra delay of sz in the response time of gene Z relative to simple
regulation with no cascade.
If Z activates a third gene W when it crosses a threshold KZ,this will occur at a time
TZW found from:

solving for TZW we obtain:

We can generalize this result: each step in a cascade, where a gene X activates a
downstream gene after crossing a threshold KXadds a delay of :

In the special case in which the activation threshold is half the steady-state level (this
can be shown to be in some cases an optimal value), the delay is 7delay=10g(2)/a. In
summary, since 11a is often on the scale of a cell generation, a transcriptional cascade
can be a slow process.

2.5 Fan-out: Transcription factor X regulates two genes YI and YZ.Draw the
resulting network, termed a fan-out with two target genes. The activation thresholds
for these genes are K1 and KZ.The activator X begins to be produced at time t=O at
rate p, and is degradedldiluted at rate a, and its signal S, is present throughout, What
are the times at which Y1 and Y2 reach halfway to their maximal expression? Design a
fan-out with three target genes in which the genes are activated with equal temporal
spacing.

Based on problem 2.4:

After the corresponding delays in gene activation, denoted 7 1 and 72, production of Yi
and Y2 starts at a constant rate reaching half the steady state after log(2)la. The time
to reach half maximum is therefore: tiiz=n+log(2)la (i=1,2), where i=1,2 for Yi and
Y2 respectively.
For three target genes, we require ~ ~ 1 (71+73).
1 2 This amounts to the following
requirements on the thresholds,

2.6 Pulse ofactivation: Consider the cascade of exercise 2.4. The input signal S,
appears at time t=O for a pulse of duration D, and then vanishes.
(a) What is the concentration Y(t)?
(b) What is the minimal pulse duration needed for activation of gene Z?
(c) Plot the maximal level reached by the gene product Z as a function of the pulse
duration D.

Solution:
a) Protein X*, the active conformation of protein X bound to its inducer, binds the
promoter of Y commencing its production according to (2.4.1) yielding the familiar
exponential approach to steady-state:
(2.6.1) Y(t)=PY/a( l -exp(-at))
b) Protein Z will be activated when Y levels cross the threshold KyZ:
(2.6.2) Y(t)+y/a ( l -exp(-aDmin))=Yst ( l -exp(-aDmin))=Kyz
Solving for D:
(2.6.3) / a log ( l /(l -Kyz/Yst))
Dmin=l
Let's consider the limits of (2.6.3). If KYZis very small D approaches 0, and a very
short duration of the input S, suffices to activate Z. When Kyz approaches YStD
approaches infinity.
c)

The schematic plot uses KyZ=0.5*Yst,and a = l . After the delay of DminZ

concentration follows the familiar exponential rise to steady state. In summary, the 3-
gene cascade results in a delay in the activation of the Z gene following an input to
the X gene.

Appendix
Full solution of the dynamics of problem 2.2
The full solution of this problem can be obtained by inserting the time-dependent
solution of equations 2.2.1 into equation 2.2.2 to obtain:
(2.A. 1) dYldt=p Ym-aY=p/3,1am (1-exp(-amt)) --ay
Using the method of problem 2.3 (below) the full solution for Y(t) is:

In the limit where mRNA degradation is much higher than protein degradation and for
11am<<&we can approximate (a-am)--amand exp(-amt)-0. Equation (2.A.2) then
reduces to:

In the limit where mRNA degradation is much smaller than protein degradation and
for l/a<<t, we can approximate (a-am)-a and exp(-at)-0. Equation (2.A.2) then
reduces to:

The stead) state solution for V is the same in both of these limiting cases : YSt= p P,: I
aa,, but the response time is governed by the slower process of the two degradation
processes - if mRNA degradation is much higher than protein degradation, the
response time is governed by protein degradation: tl/z= log(2)I a . If mRNA
degradation is much smaller than protein degradation, the response time is governed
by mRNA degradation t1/2=1og(2)/ am.

Problem 2.3 - integration factors

A general solution of a first order linear differential equation of the following form:
(2.A.5) dYldt+a(t)Y=,B(t)
can be obtained by first multiplying both sides by an integration factor y(t) defined as:
(2.A.6) p(t)= exp( la(tq)dt')

This gives us:

(2.A.7) p(dYldt)+ayY=/3y
The left side can be replaced by:
(2.A.8) d(pY)ldt=pp
Integrating this equation:
(2.A.9) PY= SPM~
which leads to
(2.A.10) Y(r)=1/p[Y(O)+-1fi)dt]

Finally, inserting p from 2.A.6 we get equation 2.3.1

(2.A.11) Y(r)=exp(-ja(t'fdt')[Y(O)+~~(t')ex~(ja(t")dt")dt']