iGEM in Tsinghua
The Department of Biological Sciences and
Biotechnology,
Tsinghua University,
China
2007.11.3
Why do we come for iGEM?
We have a common dream:
To build BioMachines as powerful
as electronic and mechanical
machines.
Even the simplest machine has two
parts:
• Power generator
-The original force to
drive the system.
• Wire
-information propagation
What did we make?
• An oscillator: RAP (AC power)
Recoverable Auto-Pulse generator (RAP)
by using a negative feedback loop.
• A communicator: CelCuit (wire)
Cell-cell communication system by using
conjugating DNA as protein carrier.
Part 1. RAP
• Oscillator is a basic element in engineering.
- It’s importance in electronic engineering is
well known.
- It’s also widely required for genetic
machine and natural organism (circadian
rhythm).
• Only one successful genetic oscillator
published up to now.
(Nature 403:335-338, 2000)
We use another strategy
• Negative feedback loop consisting of two
components.
Principle so easy!
Never constructed…
So, why not try it?
• T7 RNA polymerase (T7 RNAP) is used as
the driver and Lac operon repressor (LacI)
is used as the brake.
• Both protein are tagged with Umd to
implement fast degradation.
Implementation
Amplifier:
Repressor:
Modeling
• We model the mRNA levels and protein
levels respectively.
(In the following description, we use lower-case
of the first letter to indicate mRNA levels
whereas up-case to indicate protein levels. )
• mRNA level is governed by:
1. spontaneous degradation (minus)
2. transcription (plus)
d(lacI )
ln 2

lacI  rTranscribeT 7 round[R(T 7) nlac _ plasmid ]
dt
 mRNA
d(t 7)
ln 2

t 7  rTranscribeT 7 activatedNum  rTranscribeP freeNum
dt
 mRNA
First order reaction
Binding of proteins to DNA
according to M-C equation
• M-C equation
Occupation
ratio
Protein
concentration
h
[ P]
R([ P]) 
h
K d  [ P]
Dissociation
constant
Hill
coefficient
• Protein levels are governed by:
1. spontaneous degradation (minus);
2. translation (plus)
d( LacI )
ln 2

LacI  rTranslate lacI
dt
 LacI
d(T 7)
ln 2

T 7  rTranslate t 7
dt
T 7
First order reaction
Stochastic Model
• To make our model more realistic, we
modify our model to a stochastic one. In
this model, all the parameters, except
copy numbers of plasmids, are assumed
to obey to a norm distribution and a
‘sigma’ value (which indicates the peak
width) is set for each parameter.
Parameter Estimation
• Copy number of the repressor plasmid:
n_lac_plasmid=30 copy
• Copy number of the amplifier plasmid:
n_t7_plasmid=5 copy
• Transcription rate of T7 polymerase:
r_transcribe_T7=300 mRNA/min
• Transcription rate of host polymerase:
r_transcribe_P=30 mRNA/min
Transcription rate of host polymerase is
adapted form (Michael B. Elowitz, Nature,
2000 )
Transcription rate of T7 polymerase is about
10-20 folds of host polymerase.
• Translation rate:
r_translate=5.0 proteins/min
Adapted from (Michael B. Elowitz Nature,
2000 )
• Half life time of mRNA:
τ(mRNA)=3min
• Half life time of LacI:
τ(LacI)=10min
• Half life time of T7:
• Half life time of EGFP:
τ(EGFP)=40min
Modified from (Michael B.
Elowitz, Nature, 2000 )
τ(T7)=10min
• Dissociation constant of LacI:
K(LacI)=20
dimers per cell
• Dissociation constant of T7 polymerase:
K(T7)=20 monomers per cell
K(LacI) is adapted from (Michael B. Elowitz
Nature, 2000 )
According to (William R. McClure, PNAS, 1980),
the Kd of T7 binding to double strand DNA is
comparable to that of LacI.
Simulation Results
• Numerically simulated in MATLAB by
using ODE toolbox.
When Translation Rate Fluctuates
What makes our RAP
different?
• RAP is more stable than the repressilator.
It should be emphasized here that it is unnecessary for
the driver to degrade faster than the brake
τ(T7)=40min
• In a synchronized oscillator model
proposed by coupling oscillator to
quorum sensing system, RAP, which
is of an relaxation type, converges
more rapidly than repressilator,
which is an phase oscillator. (PNAS
99(2):679-84, 2002 )
Proceedings
• We finished almost all the molecular
cloning.
• We confirmed that T7 RNAP we cloned is
functional by luciferase test.
• We have no time to put repressor and
amplifier together before we leave for the
Jamboree… 
Parts Registered for RAP
pACYC-Lux
pEBAclANcoIpEB-SC101
pEB-T7X
T7BLUE
pT7BlueXisT7X
pEB-EGFPL
pET15b-lacI()
AAV-110
LLA-110
LVA-110
AAV-111
LLA-111
LVA-111
AAV-112
LLA-112
LVA-112
Summary of RAP Q&A
• What is RAP proposed to?
-automatic oscillation
• How is RAP constructed?
-negative feedback loop by T7 RNAP and
LacI
• What are the advantages of RAP over
existing oscillator?
-stability
-converge rapid when synchronized
Part 2. CelCuit
• Cell-cell communication is a very
important issue in synthetic biology.
• It is widely required for synchronized
oscillator, populational computation,
control of population size, artificial
organogenesis and so on.
• Drawbacks of current cell-cell
communication system-quorum sensing
system (small diffusion molecule as
messenger)
1. unrecoverable
2. difficult to achieve multiple independent
signals
Why are they so important?
Focus shifts from small chemical
to CONJUGATION
• Two strategies have been tried to implement
information transference by using conjugation.
Recovery Muti-signal New problem
problem
problem
“relaxasefusion”
Solved
“plasmidencoding” Unsolved
Solved
Too weak
Solved
Plasmid
compatibility
CelCuit: Using DNA to Carry
Proteins
Step 1
FLP
recombinase
SSB
EGFP
Step 2
SSB
binding site
Terminator
Step 3
movie
Which Executor?
• The pool of executor is very large:
transcription factors, two-hybrid system, …
• In this case, we chose FLP recombinase for
its high efficiency.
Which Carrier (SSB)?
species
High
amount
Functions
in monomer
Specific
binding site
Known to be
transferred during
conjugation
SSB1
(E. Coli)
+
+
-
-
TraD
(E. Coli)
?
-
?
+
M.
jannaschii
-
+
-
-
S.
cerevisiae
-
-
-
-
POT1
(mammals)
depends
+
+
-
• We also tried other modes:
- Lamda cI protein as the carrier
- “relaxase-fusion” strategy: Mob-FLP
fusion protein
Implementation
Proceedings
• We finish all the molecular cloning of
CelCuit except SSB part.
• We confirm that the FLP recombinase we
cloned is functional by using LacZ
reporter.
• Because the cloning of SSB is not finished
yet, we are still not sure whether the
protein transference during conjugation
does occur.
Parts Registered for CelCuit
pLZHinDIII2
pLZCI3
FLP-p15a
pEASY-p15a
pBL-Blunt
pKD-pBL
pKD-pBL-Lambda
pEB-FLP
pEB-Lambda
pEB-Lambda AmppBL-FLP
pEB-C1
pEB-Mob
pBL-Linker
pBL-FLPMob
pGEM-FLP
• We are all exciting for what we are doing.
We think the CelCuit, once finished, will
certainly revise the methodology of gene
machine in population in many cases!
Three Application Examples
• Pattern Formation (Nature. 2005 Apr
28;434(7037):1130-4 )
Pattern formation by
reciprocal interaction
between nearby cells.
• Bacterial neural network (U.C Berkley’s
dream several years ago)
Neurons in one layer
innervates those in their
downstream layer in an
all-connection mode.
• Cell Circuit (ambitious concept, isn’t it?)
If we could set a cell
array…
Insulate cells which are
near from each other.
Summary of CelCuit Q&A
• What is CelCuit proposed to?
-transfer proteins between cells
• How is CelCuit constructed?
-fuse signal proteins (FLP, T7 RNAP…) to SSB
• What are the advantages of CelCuit over
existing communicator?
-recoverable
-large signal pool
These two projects are done by:
Prof. Chen
Guoqiang
Dr. Xu Feng
Jiang Ming, Li Keyu, Dong Peng, Zhao Xinyu, Chen Chen, Shi
Zhenyu, Liu Yexing, Zhou Lijun, Yu Zhou, Liu Zhe, Song Xindong
Acknowledgement
• We thank Dr. Xie Liping for her
encouragement.
• We thank the students who took practical
synthetic biology summer course for their
technique help.
• This work is supported by the International
Competition Funds of Tsinghua University.
Thank iGEM for giving us such an
unforgettable experience!
Thank you for your
attention!