Who are we?
•
9 undergraduates advised by
– 4 graduate students
– 20 faculty from diverse departments
•
•
3 biologists, 2 engineers, 2 computer scientists,
1 physicist, and 1 biochemist.
•
Synthetic Biology Journal club
Bacterial Freeze Tag
AI-1
AHL
Unfreezer
“IT” Cell
Sender (“IT”) Cell
AHL Receiver
MotB Knock Out
LuxR
LacI
LuxR
Freeze Machine: Freezing
LacI
LacI
LacI
LacI
cI
cI
LacI
Freeze Machine: Unfreezing
LacI
AI-1
AI-1
cI
AI-1
AI-1
cI
aiiA
LasR
Dr. Unfreeze
AI-1
AI-1
AI-1
LasR
cI
cI
Freeze Machine: Unfreezing
LacI
LacI
LacI
AI-1
LacI
AI-1
cI
cI
cI
cI
AI-1
AI-1
The Tri-stable Toggle Switch
A Bi-stable Toggle Switch
A Tristable Toggle Switch
Let’s watch in practice…
Without interruption, this state is stable.
tetR
lacI
RFP
polymerase
pBad/Ara
tetR
lacI
CFP
?
pLac
araC
tetR
polymerase
YFP
pTet
lacI
araC
But we can induce a change with IPTG…
Without interruption, this state is stable.
We could have induced the pBad/Ara pathway with arabinose
or the pTet pathway with tetracycline.
RFP
pBad/Ara
tetR
lacI
araC
IPTG
polymerase
lacI
pLac
CFP
araC
tetR
tetR
pTet
YFP
lacI
araC
Modeling
Deterministic Model based on:
“Prediction and measurement of an autoregulatory genetic module.”
Isaacs et al. PNAS 2003
And
“A Bottom-Up Approach to Gene Regulation.”
Guido et al. Nature 2006
Model Derivation
Basic Idea
Fast reactions – “merizations” and operator binding events
•equilibrium equations
Slow reactions – transcription, translation, degradation
•differential equations
Combine these equations with an equation for total molecule
in the system based on plasmid copy number
Manipulate these equations to derive an expression for the
evolution of protein monomers/time
Model Derivation
Fast reactions – k’s on order of seconds
Ex:
k1 L V
L  L
 L2
Equilibrium equations
Ex:
l2 
k1L 2
l
k1LV
Trimer
formation
l3 
k2 L
l2 l
k 2 LV
Tetramer
formation
k3 L
l4 
l3
k3 LV
Operator
binding
k4 L
d1L 
l4 d 0 L
k 4 LV
Dimer
formation
k 1L
k2L V
L  L2 
 L3
k 2 L
k3L V
L  L3 
 L4
k 3 L
k4L V
D0 L  L4 
 D1L
k 4 L
Model Derivation
Slow reactions – k’s on order of minutes
 LT ktL
unbound
D0 L 
 D0 L  T
 L LT ktL
bound
D1L 
 D1L  T
Transcription and
translation of TetR
from unbound and
bound pLacI
promoters
Differential equations
Ex:
dZ T
  AT  A d 0 A  d1 A   A1d 2 A   A 2 d1aA   A3 d 2 aA    LT  L d 0 L   L d1L    T T
dt
TetR created by pBAD promoterTetR created by pLacI promoter
Prediction – no inducer
Fluorescence vs. Time - NO INDUCER
200
180
YFP
mCherry
ECFP
160
Fluorescence (A.U.)
140
KEY:
pBAD  mCherry
120
pLacI  ECFP
100
80
60
pTetR  YFP
All initial protein values =0
40
20
0
0
0.5
1
1.5
2
2.5
3
Time (Cell divisions)
3.5
4
4.5
5
System Reaches a natural steady state - pLacI promoter dominates!
Prediction - Tetracycline
Fluorescence vs. Time - Tetracycline
200
180
140
Fluorescence (A.U.)
KEY:
YFP
mCherry
ECFP
160
pBAD  mCherry
120
pLacI  ECFP
100
80
pTetR  YFP
All initial protein values =0
60
40
20
0
0
0.5
1
1.5
2
2.5
3
Time (cell divisions)
3.5
4
4.5
5
Model CONFIRMS our hypothesis in the presence of tetracylcine!
Test for stability
Will the tri-stable switch work…?
- Plug in final protein concentrations from
induced state into inducer-less system
- Test whether state remains stable
Test for stability
Fluorescence vs. Time - pTetR Stability test
200
180
YFP
mCherry
ECFP
160
KEY:
LacI0 = 5
pBAD  mCherry
TetR0 = 43
pLacI  ECFP
AraC0 = 41
pTetR  YFP
Fluorescence (A.U.)
140
120
100
80
60
40
20
0
0
0.5
1
1.5
2
2.5
3
Time (cell divisions)
3.5
4
4.5
5
System switches back
to natural steady state!
pTetR stability test
Natural steady state
Fluorescence vs. Time - NO INDUCER
Fluorescence vs. Time - pTetR Stability test
200
200
180
180
YFP
mCherry
ECFP
160
140
Fluorescence (A.U.)
Fluorescence (A.U.)
140
120
100
80
120
100
80
60
60
40
40
20
20
0
YFP
mCherry
ECFP
160
0
0.5
1
1.5
2
2.5
3
Time (cell divisions)
3.5
4
4.5
5
0
0
0.5
1
1.5
2
2.5
3
Time (Cell divisions)
3.5
4
4.5
Future
• Experimentally obtain more appropriate
model parameters
• Run parameter scans for tri-stability
– Potentially identify proper regulatory elements
and or tinkering required to for tri-stability
• Add stochastic variation to the model
Conclusions
• Tri-stable Toggle Switch
- Our model predicts that a tri-stable switch given our
experimental setup will be unable to reach a stable
state following transient induction
• Freeze-tag
– Theoretically shows an interesting biological circuit
which uses cell-cell communication with a bi-stable
circuit to create a novel observable interaction
– Circuit complexity may jeopardize the efficiency of the
system by causing a build-up of various proteins.
Modifications of the circuit may be required to
circumvent such effects.
Special thanks to:
•
•
•
•
•
•
•
•
•
Gary Wessel
Tayhas Palmore
Alex Brodskey
Nicola Neretti
Karen Haberstroh
David Targan
Ruth Simmons
Houseknecht lab @ Pfizer
Sir Josiah Carberry