recombinase-based circuits for environmental detection...
TRANSCRIPT
Recombinase-Based Circuits for Environmental Detection, Diagnostics, and Logging
Richard M. Murray California Institute of Technology
Victoria Hsiao (Amyris) Yutaka Hori (Keio U) Andrey Shur
Outline I. Event detection and introduction to recombinases II. Event diagnostics using population-level, stochastic response III. New directions: event logging, field-programmability IV. Summary and next steps
Richard M. Murray, Caltech CDS/BESynBio Control, 18 Jul 2017
Detection • Monitor the environment and look for a specified “trigger”
• Simple monitoring: signal detection; logic (AND, OR, …)
• More complex detection: temporal sequences of events Diagnostics • Extract quantitative properties: magnitude, duration, etc
Logging • Remember environmental conditions for later readout
Approach • Component technolo-
gies: signal detection, memory, species comp-arison, logic functions
• Event detectors: A > B, A followed by B, A > thresh
• Interconnection frame-work: modular techni-ques for interconnecting components & detectors
Environmental Detection, Diagnostics, and Logging
2
Interconnection of modulesto detect more complex events
GFP if A is always less than B RFP if A is greater than B
Applications: environmental monitoring, diagnostics for health, circuit debugging, …
Richard M. Murray, Caltech CDS/BESynBio Control, 18 Jul 2017
Integrases and Excisionases(Serine) Integrases • Mechanism by which phage insert DNA into the
chromosome of a bacterial host • attP = phage recognition site
• attB = bacterial recognition site
• Integrase action: insert phage DNA into bacterial chromosome, leaving changing recognition sites
Excisionases • Reverse reaction requires second phage-coded
protein, excisonase
Other recombinases • Cre recombinase - tyrosine recombinase
• Cre-Lox & Flp-FRT recombinases - insertion, excision, inversion, translocation
• CRISPR/Cas9 - guide RNA-directed excision, insertion
3
A. Campbell, “General Aspects of Lysogeny”. In Bacteriophages, R. Calendar (ed), 2006.
Richard M. Murray, Caltech CDS/BESynBio Control, 18 Jul 2017
Repurposing Recombinases for Synthetic BiologyBasic trick: put attB/attP on same piece of DNA • Integrase activity causes DNA between sites to flip
• Excisionase (+ integrases) causes reverse flip
Effects depend on orientation of attachment sides • Attachment sites pointing toward each other: flip
• Attachment sites in same direction: excise
4
attB attP
attB attP
attB
attP
Cutting dimer
Rotating dimer
{
{
attL attR
Original DNA
( ‘B’ and ‘P’ stand for ‘bacteria’ and ‘phage’)
Integrase monomers bind
Integrase tetramer forms
DNA digestion
DNA rotation and ligation
Final DNA state
( ‘L’ and ‘R’ stand for ‘left’ and ‘right’)
J. Bonnet, P. Yin, M. E. Ortiz, P. Subsoontorn, and D. Endy. Amplifying genetic logic gates. Science, 340(6132):
599–603, 2013.
N. Roquet, A. P. Soleimany, A. C. Ferris, S. Aaronson, and T. K. Lu, “Synthetic recombinase-based state machines in
living cells”. Science, 353(6297):aad8559, 2016.
Richard M. Murray, Caltech CDS/BESynBio Control, 18 Jul 2017
Several examples of recombinase based circuits in the literature
Recombinase-Based Circuit Examples
5
P. Siuti, J. Yazbek, T. K. Lu, “Synthetic circuits integrating logic and memory in living cells”.
Nature Biotechnology, 1–6, 2013
A. E. Friedland, T. K. Lu, X. Wang, D. Shi, G. Church and J. J. Collins, “Synthetic gene networks that count”. Science,
324(5931), 1199–1202, 2009.
J. Bonnet, P. Subsoontorn, and D. Endy, “Rewritable digital data storage in live cells via
engineered control of recombination directionality”.
PNAS, 1–56, 2012.
Richard M. Murray, Caltech CDS/BECCC/SICE, Jul 2015
0 2 4 6 8 10
0
100
200
300
400
500
600
700
800
900
1000
∆TB−A
Fin
al n
um
be
r o
f ce
lls in
sta
te (
N)
Original state
A onlyB first
A then B
Event Ordering Detection (A then B)
6
Arabinose (“A”) aTc (“B”)
GFPRFP α2
α1 α3
GFPRFP GFPRFP
GFPRFPGFPRFP
X
Z
Y
WTP-901 Bxb1
A 1 0B 0 0
A then B 0 1B then A 0 0Integrases:
Terminator PromoterReporter Reporter
GFPRFP
0 5 10 15 200
100
200
300
400
500
600
700
800
900
1000
Fina
l num
ber o
f cel
ls in
sta
te (N
)
Time (h)
∆T = 2 hr
0 5 10 15 200
100
200
300
400
500
600
700
800
900
1000
Fina
l num
ber o
f cel
ls in
sta
te (N
)
Time (h)
∆T = 1 hr
0 5 10 15 200
100
200
300
400
500
600
700
800
900
1000
Fina
l num
ber o
f cel
ls in
sta
te (N
)
Time (h)
X, original DNA stateY, A onlyZ, B firstW, A then B
X, original DNA state
Y, A onlyZ, B first
W, A then B
∆T = 0 hr
X, original DNA state
Y, A only
Z, B first
W, A then B
A and B simultaneously
0 5 10 15 200
100
200
300
400
500
600
700
800
900
1000
Fina
l num
ber o
f cel
ls in
sta
te (N
)
Time (h)
∆T = 2 hr
0 5 10 15 200
100
200
300
400
500
600
700
800
900
1000
Fina
l num
ber o
f cel
ls in
sta
te (N
)
Time (h)
∆T = 1 hr
0 5 10 15 200
100
200
300
400
500
600
700
800
900
1000
Fina
l num
ber o
f cel
ls in
sta
te (N
)
Time (h)
X, original DNA stateY, A onlyZ, B firstW, A then B
X, original DNA state
Y, A onlyZ, B first
W, A then B
∆T = 0 hr
X, original DNA state
Y, A only
Z, B first
W, A then B
A at t = 0 hr B at t = 1 hr
Number of cellsthat switch dependson interval between
the two inputs
Experiments match
simulations
# ce
lls in
eac
h st
ate
# ce
lls in
eac
h st
ate
Frac
tion
of m
ax G
FP
DNA layout
Steady state responseMarkov process model
for DNA state in each cell
∆ Tind1-ind2 (h)0 2 4 6 8 10
Frac
tion
of m
ax G
FP0
0.2
0.4
0.6
0.8
1
1.2
A only
B only No inducer
A then B
B then A
Hsiao, Hori and M., MSB, 2016
Richard M. Murray, Caltech CDS/BESynBio Control, 18 Jul 2017
Additional Event “Diagnostics”Q: can we keep track of other things in addition to ΔT? • Have two measurements: #red, #green (versus total concentration)
• Idea: GFP population depends on the duration of pulse B => can also measure PWb
Use calibration phase (or models -:) to create lookup table and determine properties of inputs
7
Simulation Experiment (flow cytometry)
Hsiao, Hori and M., MSB, 2016
Richard M. Murray, Caltech CDS/BESynBio Control, 18 Jul 2017
Event Logging CircuitObjectives • Implement a genome ‘recording
site’ with a chronological order of inserted DNA fragments
• Utilize plasmids as the source of recording material, and use integrases as the means for DNA insertion
Status • Built event logger design
consisting of four modules: event plasmid (ID sequence), input selector (not shown), controller, logging site
• One event detector circuit tested and working
• Event plasmid selector using Cas9 gRNA to block integrases working in TX-TL
8
Event logging circuit using DNA integrases
Proof-of-concept experimental validation
Event Plasmid
A. Shur (unpublished)
Richard M. Murray, Caltech CDS/BESynBio Control, 18 Jul 2017
Field Programmable Circuits
9
Shur and M., bioRxiv, 2017
Idea: use dCas9 to block integrases • Use gRNA to guide dCas9 to
recognition site => no integration • Can create different circuits by
controlling insertion of elements
Similar to field programmable gate arrays (FPGA) technology in circuits • Use expression of different integrases
to interconnect circuits
Preliminary experiments: it works! • Cell-free assays show repression of
integrates activity
A. Shur, R. M. Murray, “Repressing integrase attachment site opera-tion with CRISPR-Cas9 in E. coli”. bioRxiv, 2017. DOI 10.1101/110254
Richard M. Murray, Caltech CDS/BESynBio Control, 18 Jul 2017
Recombinase-Based Circuits for Environmental Detection, Diagnostics, and Logging
Recombinase-based circuits compliment capabilities of genetic networks • Ability to reconfigure DNA in cells can be
used for logic and memory (detection logic)
• Stochastic response across populations of cells provides diagnostic capability
• Use DNA as a “recording tape” (logging) [see also recent paper by Shipman et al.]
Other uses to be explored • Integrases as a means of “programming”
circuits (FPGA-style)
• Use integrase/excisionase pairing as feedback mechanism (see Folliard poster)
Some open problems • Compilers: specs → (recombinase) circuits • Leaks are still a problem (leaky integrate
expression => noisy flipping) • Better exploiting stochastic dynamics
10
Richard M. Murray, Caltech CDS/BEEngSci Syn Bio workshop, Sep 2014
Some Challenges and Research Directions (BFS)Better understanding of uncertainty • How do we capture observed behavior using
structured models for (dynamic) uncertainty
Stochastic specifications and design tools • How do we describe stochastic behavior in a
systematic and useful way? • How do we design stochastic behavior? • What are the right design “knobs”?
Higher level design abstractions • What are the right “device-level” design
abstractions (and corresponding diagrams)?
Redundant design strategies • Start implementing non-minimal designs
• Analogy: stochastic memory storage
Scaling up: components → devices → systems • How can we use in vitro “breadboarding” to
design and implement complex systems
11