Supporting Information
Conditional Guide RNAs:
Programmable Conditional Regulation of CRISPR/Cas Function
in Bacterial and Mammalian Cells via Dynamic RNA Nanotechnology
Mikhail H. Hanewich-Hollatz
†
,
§
, Zhewei Chen
†
,
§
, Lisa M. Hochrein
†
, Jining Huang
†
, and Niles A. Pierce
†
,
‡
,
¶
,
⇤
Contents
S1 Methods
S4
S1.1 Rational design of libraries of orthogonal cgRNAs using NUPACK . . . . . . . . . . . . . . . . . . . . . . .
S4
S1.1.1 Target test tube specification for terminator switch mechanism . . . . . . . . . . . . . . . . . . . . . .
S5
S1.1.2 Target test tube specification for splinted switch mechanism . . . . . . . . . . . . . . . . . . . . . . .
S7
S1.1.3 Target test tube specification for toehold switch mechanism . . . . . . . . . . . . . . . . . . . . . . .
S9
S1.2 Methods for bacterial studies in
E. coli
.......................................
S11
S1.2.1 Plasmid construction and molecular cloning for bacterial cgRNA studies . . . . . . . . . . . . . . . .
S11
S1.2.2 Bacterial culture and silencing assay for cgRNA studies . . . . . . . . . . . . . . . . . . . . . . . . .
S11
S1.2.3 Flow cytometry for bacterial cgRNA studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
S11
S1.3 Methods for mammalian studies in HEK 293T cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
S13
S1.3.1 Plasmid construction and molecular cloning for mammalian cgRNA studies . . . . . . . . . . . . . . .
S13
S1.3.2 Mammalian cell culture and induction assay for cgRNA studies . . . . . . . . . . . . . . . . . . . . .
S14
S1.3.3 Flow cytometry for mammalian cgRNA studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
S15
S1.4 Quantitative fluorescence analysis for
E. coli
....................................
S16
S1.4.1 Measuring signal in
E. coli
.........................................
S16
S1.4.2 Fold change for constitutively active cgRNAs (ON
!
OFF logic) with silencing dCas9 in
E. coli
....
S16
S1.4.3 Dynamic range for constitutively active cgRNAs (ON
!
OFF logic) with silencing dCas9 in
E. coli
...
S16
S1.4.4 Fold change for constitutively inactive cgRNAs (OFF
!
ON logic) with silencing dCas9 in
E. coli
...
S17
S1.4.5 Dynamic range for constitutively inactive cgRNAs (OFF
!
ON logic) with silencing dCas9 in
E. coli
..S17
S1.4.6 Fractional dynamic range for cgRNAs with silencing dCas9 in
E. coli
..................
S17
S1.4.7 Crosstalk for orthogonal cgRNAs in
E. coli
................................
S17
S1.5 Quantitative fluorescence analysis for HEK 293T cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
S18
S1.5.1MeasuringsignalinHEK293Tcells....................................
S18
S1.5.2 Fold change for constitutively active cgRNAs (ON
!
OFF logic) with inducing dCas9 in HEK 293T cells
S18
S1.5.3 Dynamic range for constitutively active cgRNAs (ON
!
OFF logic) with inducing dCas9 in HEK 293T
cells.....................................................
S19
S1.5.4 Fractional dynamic range for cgRNAs with inducing dCas9 in HEK 293T cells . . . . . . . . . . . . .
S19
S1.5.5 Crosstalk for orthogonal cgRNAs in HEK 293T cells . . . . . . . . . . . . . . . . . . . . . . . . . . .
S19
S2 Sequences
S21
S2.1 Sequences for cgRNAs, triggers, and control gRNAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
S21
S2.2 Transcriptional promoter and terminator sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
S23
S2.3 Genomically incorporated gene sequences (
E. coli
Ec001) ............................
S23
S3 Plasmids
S25
S3.1 Constitutively active terminator switch in
E. coli
.................................
S25
S3.2 Constitutively active splinted switch in
E. coli
...................................
S28
S3.3 Constitutively inactive toehold switch in
E. coli
..................................
S31
S3.4 pdCas9+lacI in
E
.coli ...............................................
S34
S3.5 Constitutively active terminator switch in HEK 293T cells . . . . . . . . . . . . . . . . . . . . . . . . . . . .
S36
†
Division of Biology & Biological Engineering, California Institute of Technology, Pasadena, CA 91125, USA.
‡
Division of Engineering & Applied
Science, California Institute of Technology, Pasadena, CA 91125, USA.
¶
Weatherall Institute of Molecular Medicine, University of Oxford, Oxford OX3 9DS,
UK.
§
Authors contributed equally.
⇤
Email:
niles@caltech.edu
S1
S4 Schematics of putative ON and OFF states
S43
S4.1 Constitutively active terminator switch cgRNA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
S43
S4.2ConstitutivelyactivesplintedswitchcgRNA....................................
S43
S4.3 Constitutively inactive toehold switch cgRNA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
S44
S5 Flow cytometry replicates
S45
S5.1 Constitutively active terminator switch in
E. coli
.................................
S45
S5.1.1 ON state, OFF state, and conditional response (cf. Figure 2c) . . . . . . . . . . . . . . . . . . . . . . .
S45
S5.1.2 Orthogonal library studies (cf. Figure 2d) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
S45
S5.2 Constitutively active splinted switch in
E. coli
...................................
S46
S5.2.1 ON state, OFF state, and conditional response (cf. Figure 3c) . . . . . . . . . . . . . . . . . . . . . . .
S46
S5.2.2 Orthogonal library studies (cf. Figure 3d) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
S46
S5.3 Constitutively inactive toehold switch in
E. coli
..................................
S47
S5.3.1 ON state, OFF state, and conditional response (cf. Figure 4c) . . . . . . . . . . . . . . . . . . . . . . .
S47
S5.3.2 Orthogonal library studies (cf. Figure 4d) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
S47
S5.4 Constitutively active terminator switch in HEK 293T cells . . . . . . . . . . . . . . . . . . . . . . . . . . . .
S48
S5.4.1 ON state, OFF state, and conditional response (cf. Figure 5b) . . . . . . . . . . . . . . . . . . . . . .
S48
S5.4.2 Orthogonal library studies (cf. Figure 5c) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
S49
S5.5 Quantifying ON state, OFF state, fold change, and dynamic range . . . . . . . . . . . . . . . . . . . . . . . .
S50
S5.6 Quantifying crosstalk for cognate and non-cognate cgRNA/trigger pairs . . . . . . . . . . . . . . . . . . . . .
S51
S6 Additional Studies
S52
S6.1 Constitutively active terminator switch using alternative plasmid layout and constitutive trigger expression in
E. coli
........................................................
S52
S6.1.1 Plasmids used for alternative terminator switch plasmid layout (cf. Section S3.1) . . . . . . . . . . . .
S54
S6.1.2 Flow cytometry replicates for ON state, OFF state, and conditional response (cf. Figure 2c and Sec-
tionS5.1.1).................................................
S57
S6.1.3 Flow cytometry replicates for orthogonal library studies (cf. Figure 2d and Section S5.1.2) . . . . . . .
S57
S6.2 Single and double sequence inserts for construction of allosteric cgRNAs in
E. coli
...............
S58
S6.2.1 Quantifying performance of candidate cgRNAs using 71
E. coli
strains .................
S58
S6.2.2 Candidate cgRNA and trigger sequences used for single and double insert studies . . . . . . . . . . . .
S61
S6.3 Characterization of splinted switch conditional response to lacI-regulated trigger expression at different time
points in
E. coli
...................................................
S64
List of Figures
S1 Target test tubes for sequence design of orthogonal terminator switch cgRNAs . . . . . . . . . . . . . . . . . .
S5
S2 Nucleotide defect weights for sequence design of terminator switch cgRNAs . . . . . . . . . . . . . . . . . . .
S6
S3 Target test tubes for sequence design of orthogonal splinted switch cgRNAs . . . . . . . . . . . . . . . . . . .
S7
S4 Nucleotide defect weights for sequence design of splinted switch cgRNAs . . . . . . . . . . . . . . . . . . . .
S8
S5 Target test tubes for sequence design of orthogonal toehold switch cgRNAs . . . . . . . . . . . . . . . . . . .
S10
S6 Nucleotide defect weights for sequence design of toehold switch cgRNAs . . . . . . . . . . . . . . . . . . . .
S10
S7 Illustration of gates used for flow cytometry analysis of
E. coli
...........................
S12
S8 Illustration of gates used for flow cytometry analysis of HEK 293T cells. . . . . . . . . . . . . . . . . . . . . .
S15
S9 Example plasmid map for terminator switch in
E. coli
...............................
S26
S10 Example annotated plasmid sequence for terminator switch in
E. coli
.......................
S27
S11 Example plasmid map for splinted switch in
E. coli
................................
S29
S12 Example annotated plasmid sequence for splinted switch in
E. coli
........................
S30
S13 Example plasmid map for toehold switch in
E. coli
................................
S32
S14 Example annotated plasmid sequence for toehold switch in
E. coli
........................
S33
S15 Plasmid map for pdCas9+lacI in
E. coli
......................................
S34
S16 Annotated plasmid sequence for expression of pdCas9+lacI in
E. coli
......................
S35
S17 Example plasmid map for terminator switch cgRNA in HEK 293T cells . . . . . . . . . . . . . . . . . . . . .
S37
S18 Example annotated plasmid sequence for terminator switch cgRNA in HEK 293T cells . . . . . . . . . . . . .
S38
S19 Example plasmid map for terminator switch trigger in HEK 293T cells . . . . . . . . . . . . . . . . . . . . . .
S39
S20 Example annotated plasmid sequence for terminator switch trigger in HEK 293T cells . . . . . . . . . . . . . .
S40
S21 Plasmid map for induction assay reporter in HEK 293T cells . . . . . . . . . . . . . . . . . . . . . . . . . . .
S41
S2
S22 Annotated plasmid sequence for induction assay reporter in HEK 293T cells . . . . . . . . . . . . . . . . . . .
S42
S23 Schematics of putative ON and OFF states for terminator switch mechanism . . . . . . . . . . . . . . . . . . .
S43
S24 Schematics of putative ON and OFF states for splinted switch mechanism . . . . . . . . . . . . . . . . . . . .
S43
S25 Schematics of putative OFF and ON states for toehold switch mechanism . . . . . . . . . . . . . . . . . . . .
S44
S26 Flow cytometry replicates for terminator switch ON state, OFF state, and conditional response in
E. coli
(cf.
Figure2c)......................................................
S45
S27 Flow cytometry replicates for terminator switch orthogonal response in
E. coli
(cf. Figure 2d) . . . . . . . . . .
S45
S28 Flow cytometry replicates for splinted switch ON state, OFF state, and conditional response in
E. coli
(cf.
Figure3c)......................................................
S46
S29 Flow cytometry replicates for splinted switch orthogonal response in
E. coli
(cf. Figure 3d) . . . . . . . . . . .
S46
S30 Flow cytometry replicates for toehold switch ON state, OFF state, and conditional response in
E. coli
(cf. Figure
4c)..........................................................
S47
S31 Flow cytometry replicates for toehold switch orthogonal response in
E. coli
(cf. Figure 4d) . . . . . . . . . . .
S47
S32 Flow cytometry replicates for terminator switch ON state, OFF state, and conditional response in HEK 293T
cells(cf.Figure5b).................................................
S48
S33 Flow cytometry replicates for terminator switch orthogonal response in HEK 293T cells (cf. Figure 5c) . . . .
S49
S34 Constitutively active terminator switch cgRNAs (ON
!
OFF logic) using alternative plasmid layout and consti-
tutive trigger expression with silencing dCas9 in
E. coli
..............................
S53
S35 Example plasmid map for alternative terminator switch plasmid layout with constitutive trigger expression in
E. coli
(cf.FigureS9)................................................
S55
S36 Example annotated plasmid sequence for alternative terminator switch plasmid layout with constitutive trigger
expression in
E. coli
(cf.FigureS10)........................................
S56
S37 Flow cytometry replicates for terminator switch ON state, OFF state, and conditional response in
E. coli
with
constitutive trigger expression (cf. Figure S26) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
S57
S38 Flow cytometry replicates for terminator switch orthogonal response in
E. coli
with constitutive trigger expres-
sion(cf.FigureS27) ................................................
S57
S39 Performance of single and double sequence inserts for construction of allosteric cgRNAs in
E. coli
.......
S59
S40 Characterization of splinted switch response to lacI-regulated trigger expression at different time points in
E. coli
S64
List of Tables
S1 Example duplex fragments for Golden Gate assembly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
S13
S2 Mammalianinductionassaycomponents......................................
S14
S3 Terminator switch sequences for studies in
E. coli
.................................
S21
S4 Splinted switch sequences for studies in
E. coli
..................................
S21
S5 Toehold switch sequences for studies in
E. coli
..................................
S22
S6 Terminator switch sequences for studies in HEK 293T cells . . . . . . . . . . . . . . . . . . . . . . . . . . . .
S22
S7 ControlgRNAsequences..............................................
S22
S8 Transcriptional promoter and terminator sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
S23
S9 Plasmids used with terminator switch cgRNAs in
E. coli
.............................
S25
S10 Plasmids used with splinted switch cgRNAs in
E. coli
...............................
S28
S11 Plasmids used with toehold switch cgRNAs in
E. coli
...............................
S31
S12 Plasmids used with terminator switch cgRNAs in HEK 293T cells . . . . . . . . . . . . . . . . . . . . . . . .
S36
S13 Quantifying ON state, OFF state, fold change, and dynamic range (cf. Figures 2d, 3d, 4d, 5c) . . . . . . . . . .
S50
S14 Quantifying crosstalk for cognate and non-cognate cgRNA/trigger pairs (cf. Figures 2d, 3d, 4d, 5c) . . . . . . .
S51
S15 Plasmids used for alternative terminator switch plasmid layout with constitutive trigger expression in
E. coli
(cf.TableS9) ....................................................
S54
S16 Quantifying ON state, OFF state, fold change, and dynamic range for candidate cgRNAs with designed single
and double inserts into the standard gRNA structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
S60
S17 Candidate cgRNA sequences used for single and double insert studies in
E. coli
.................
S62
S18 Trigger sequences used for single and double insert studies in
E. coli
.......................
S63
S3
S1 Methods
S1.1 Rational design of libraries of orthogonal cgRNAs using NUPACK
For each mechanism, orthogonal cgRNA/trigger pairs were designed using the reaction pathway engineering tools within
NUPACK (
nupack.org
; see the NUPACK 3.2 User Guide).
1,2
Target test tubes were specified using the general formulation
of Section S2.2.1 in the Supplementary Information of Wolfe et al.
2
using the definitions provided below (Section S1.1.1 for
the terminator switch, Section S1.1.2 for the splinted switch, and Section S1.1.3 for the toehold switch). Sequence designs
were performed for libraries of 4 orthogonal cgRNA/trigger pairs. For a given design trial, the sequences were optimized by
mutating the sequence set to reduce the multi-tube ensemble defect
2
subject to the diverse sequence constraints detailed below.
Within the ensemble defect, defect weights (see Section S1.6 in the Supplementary Information of Wolfe et al.
2
) were applied
to prioritize design effort as described below. Designs were performed using RNA parameters for 37
◦
C in 1M Na
+
.
3
After
performing several independent design trials for a given mechanism, a final sequence set was selected for experimental testing
based on inspection of the predicted structural defects (fraction of nucleotides in the incorrect base-pairing state within the
ensemble of an on-target complex) and concentration defects (fraction of nucleotides in the incorrect base-pairing state because
there is a deficiency in the concentration of an on-target complex) for species in the context of the target test tubes,
2,4
as well
as for each cgRNA in the presence of each non-cognate trigger (e.g., computational orthogonality study of Figure 6b (right)).
After preliminary experimental studies, 3 cgRNA/trigger pairs (termed A, B, C for bacterial studies and Q, R, S for mammalian
studies) were selected for full experimental characterization.
S4
S1.1.1 Target test tube specification for terminator switch mechanism
To design
N
orthogonal systems, the total number of target test tubes is
|
⌦
|
=
P
n
=1
,...,N
{
Step 0, Step 1
}
n
+ Crosstalk =
2
N
+1
; the target test tubes in the multi-tube ensemble,
⌦
, are indexed by
h
=1
,...,
|
⌦
|
.
L
max
=2
for all tubes (i.e., each
target test tube contains all off-target complexes of up to 2 strands). Final sequence designs for orthogonal cgRNA/trigger pairs
A, B, C are shown in Table S3 for bacterial studies and Q, R, S are shown in Table S6 for mammalian studies.
Reactants for system
n
•
cgRNAs: G
n
•
Triggers: X
n
Elementary step tubes for system
n
•
Step 0
n
tube:
products
0
n
⌘
{
G, X
}
n
;
reactants
0
n
⌘;
;
exclude
0
n
⌘
{
G
·
X
}
n
•
Step 1
n
tube:
products
1
n
⌘
{
G
·
X
}
n
;
reactants
1
n
⌘
{
G, X
}
n
;
exclude
1
n
⌘;
Global crosstalk tube
•
Crosstalk tube:
reactive
global
⌘[
n
=1
,...,N
{
reactive
n
}
;
crosstalk
global
⌘
L
L
max
global
−[
n
=1
,...,N
{
cognate
n
}
The reactive species and cognate products for system
n
are:
•
simple
n
⌘
{
G, X
}
n
•
ss-out
n
⌘
X
n
•
ss-in
n
⌘
G
ss
n
, the 30nt single stranded terminator loop insert domain
•
reactive
n
⌘
{
G, X, G
ss
}
n
•
cognate
n
⌘
{
G
·
X, G
ss
·
X
}
n
Step 0
n
Step 1
n
G
n
10nM
X
n
10nM
10nM
G
n
·X
n
G
n
X
n
10nM
10nM
10nM
Global Crosstalk Tube
Constrained by target gene
Constrained by dCas9
Constrained by synthetic terminator
Elementary Step Tubes
n=1,...,N
Sequence Constraints
G
ss
n
Complementarity between G
n
and X
n
Tube
On-targets (
on
h
)
Off-targets (
o
↵
h
)
Step 0
n
{
G, X
}
n
L
L
max
0
n
−
{
G
·
X
}
n
Step 1
n
{
G
·
X
}
n
{
G, X
}
n
[
L
L
max
1
n
Crosstalk
[
n
=1
,...,N
{
reactive
n
}
L
L
max
global
−[
n
=1
,...,N
{
cognate
n
}
Figure S1: Target test tubes for sequence design of orthogonal terminator switch cgRNAs.
Top: Target test tube schematics. Bottom:
Target test tube details. Each target test tube contains the depicted on-target complexes (each with the depicted target structure and a target
concentration of 10 nM) and the off-target complexes listed in the table (each with vanishing target concentration). The on-target structures
depicted above are used in the mechanism schematic of Figure 2b. To simultaneously design
N
orthogonal systems, the total number of target
test tubes is
|
⌦
|
=2
N
+1
.
L
max
=2
for all tubes. Domain shading reflects sequence constraints. Design conditions: RNA in 1 M Na
+
at
37
◦
C.
S5
Sequence constraints
•
Assignment constraints: portions of the cgRNA are constrained to match standard gRNA sequences for use with dCas9
(shaded gray in Figure 2b, Figure S1, and Tables S3 and S6), the synthetic terminators for the cgRNA and trigger are fully
constrained (shaded tan in Figure 2b, Figure S1, and Tables S3 and S6).
•
Watson–Crick constraints: cgRNA sequence domains “d-e-f” are constrained to be complementary to the trigger sequence
domains “f*-e*-d*” (shaded blue in Figure 2b, Figure S1, and Tables S3 and S6).
•
Assignment constraint: cgRNA domain “u” is constrained to be complementary to a subsequence of the target gene mRFP
(full template sequence in Section S2.3, constrained sequence shaded orange in Figure 2b, Figure S1, and Tables S3 and
S6).
•
Pattern prevention constraints: for the bacterial cgRNA mechanisms (cgRNA A, cgRNA B, cgRNA C) the following pat-
terns are prevented for cgRNA sequence domain “f”:
AAAA, CCCC, GGGG, UUUU
. For the mammalian cgRNA mech-
anisms (cgRNA Q, cgRNA R, cgRNA S) the following patterns are prevented for cgRNA sequence domains “d”, “e”, and
“f”:
AAAA, CCCC, GGGG, UUUU, KKKKKK, MMMMMM, RRRRRR, SSSSSS, WWWWWW, YYYYYY
.
Defect weights
•
Test tube weight for each elementary step tube: 1
•
Test tube weight for global crosstalk tube: 4
•
Nucleotide weights are depicted for each complex in Figure S2
G
n
X
n
G
ss
G
n
·
X
n
w = 1
w = 1.5
w = 3
w = 1
w = 1
w = 3
w = 1
w = 3
w = 0
w = 0
w = 3
w = 0
w = 0
w = 0
Constrained by target gene
Constrained by dCas9
Constrained by synthetic terminator
Complementarity between G
n
and X
n
n
Figure S2: Nucleotide defect weights for sequence design of terminator switch cgRNAs.
Within the target test tubes of Figure S1, the
nucleotides in a given sequence domain within a given complex are assigned a defect weight
w
as depicted.
S6
S1.1.2 Target test tube specification for splinted switch mechanism
To design
N
orthogonal systems, the total number of target test tubes is
|
⌦
|
=
P
n
=1
,...,N
{
Step 0, Step 1
}
n
+ Crosstalk =
2
N
+1
; the target test tubes in the multi-tube ensemble,
⌦
, are indexed by
h
=1
,...,
|
⌦
|
.
L
max
=2
for all tubes (i.e., each
target test tube contains all off-target complexes of up to 2 strands). Final sequence designs for orthogonal cgRNAs/trigger
pairs A, B, C are shown in Table S4.
Reactants for system
n
•
cgRNAs: G
n
•
Triggers: X
n
Elementary step tubes for system
n
•
Step 0
n
tube:
products
0
n
⌘
{
G, X
}
n
;
reactants
0
n
⌘;
;
exclude
0
n
⌘
{
G
·
X
}
n
•
Step 1
n
tube:
products
1
n
⌘
{
G
·
X
}
n
;
reactants
1
n
⌘
{
G, X
}
n
;
exclude
1
n
⌘;
Global crosstalk tube
•
Crosstalk tube:
reactive
global
⌘[
n
=1
,...,N
{
reactive
n
}
;
crosstalk
global
⌘
L
L
max
global
−[
n
=1
,...,N
{
cognate
n
}
The reactive species and cognate products for system
n
are:
•
simple
n
⌘
{
G, X
}
n
•
ss-out
n
⌘
X
n
•
ss-in
n
⌘
G
ss
n
, the 35nt single stranded handle and terminator loop insert domains with intervening gRNA sequence
•
reactive
n
⌘
{
G, X, G
ss
}
n
•
cognate
n
⌘
{
G
·
X, G
ss
·
X
}
n
Step 0
n
Global Crosstalk Tube
Step 1
n
G
n
10nM
X
n
10nM
10nM
G
n
·X
n
G
n
X
n
10nM
10nM
10nM
Constrained by target gene
Constrained by dCas9
Constrained by synthetic terminator
Complementarity between G
n
and X
n
G
ss
n
Elementary Step Tubes
n=1,...,N
Sequence Constraints
Tube
On-targets (
on
h
)
Off-targets (
o
↵
h
)
Step 0
n
{
G, X
}
n
L
L
max
0
n
−
{
G
·
X
}
n
Step 1
n
{
G
·
X
}
n
{
G, X
}
n
[
L
L
max
1
n
Crosstalk
[
n
=1
,...,N
{
reactive
n
}
L
L
max
global
−[
n
=1
,...,N
{
cognate
n
}
Figure S3: Target test tubes for sequence design of orthogonal splinted switch cgRNAs.
Top: Target test tube schematics. Bottom:
Target test tube details. Each target test tube contains the depicted on-target complexes (each with the depicted target structure and a target
concentration of 10 nM) and the off-target complexes listed in the table (each with vanishing target concentration). The on-target structures
depicted above are used in the mechanism schematic of Figure 3b. To simultaneously design
N
orthogonal systems, the total number of target
test tubes is
|
⌦
|
=2
N
+1
.
L
max
=2
for all tubes. Domain shading reflects sequence constraints. Design conditions: RNA in 1 M Na
+
at
37
◦
C.
S7
Sequence constraints
•
Assignment constraints: portions of the cgRNA are constrained to match standard gRNA sequences for use with dCas9
(shaded gray in Figure 3b, Figure S3, and Table S4), the synthetic terminator for the trigger is fully constrained (shaded tan
in Figure 3b, Figure S3, and Table S4).
•
Watson–Crick constraints: cgRNA sequence domains “d” and “e” are constrained to be complementary to the trigger
sequence domains “d*” and “e*” (shaded blue in Figure 3b, Figure S3, and Table S4).
•
Assignment constraint: cgRNA domain “u” is constrained to be complementary to a subsequence of the target gene sfGFP
(full template sequence in Section S2.3, constrained sequence shaded orange in Figure 3b, Figure S3, and Table S4).
•
Pattern prevention constraints: the following patterns are prevented for cgRNA sequence domains “d” and “e”:
AAAA,
CCCC, GGGG, UUUU
.
Defect weights
•
Test tube weight for each elementary step tube: 1
•
Test tube weight for global crosstalk tube: 4
•
Nucleotide weights are depicted for each complex in Figure S4
w = 1
G
n
X
n
G
ss
G
n
·
X
n
w = 1.5
w = 1.5
w = 3
w = 3
w = 1
w = 1
w = 1
w = 3
w = 3
w = 3
w = 0
w = 0
w = 0
w = 3
w = 3
w = 3
w = 3
w = 0
w = 0
w = 0
n
Constrained by target gene
Constrained by dCas9
Constrained by synthetic terminator
Complementarity between G
n
and X
n
Figure S4: Nucleotide defect weights for sequence design of splinted switch cgRNAs.
Within the target test tubes of Figure S3, the
nucleotides in a given sequence domain within a given complex are assigned a defect weight
w
as depicted.
S8
S1.1.3 Target test tube specification for toehold switch mechanism
Computational sequence design of the toehold switch mechanism was performed using a previous version of NUPACK that did
not yet support exclusion of a set of complexes from a target test tube ensemble. For example, to design
N
orthogonal systems,
the test tube specifications detailed in the bottom tables of Figures S1 and S3 use the “minus” operator to enable compact
specification of one Reactants tube for each system
n
=1
,...,N
and a single Global Crosstalk tube. Lacking the minus
operator at the time the toehold switch library was designed, we used a more verbose target test tube specification with separate
Reactants tubes for cgRNA and trigger for each system
n
=1
,...,N
, as well as one Crosstalk tube for each non-cognate
cgRNA/trigger pair. To design
N
orthogonal systems, the total number of target test tubes is:
|
⌦
|
=
X
n
=1
,...,N
{
Step 0
G
, Step 0
X
, Step 1
}
n
+
X
n
=1
,...,N
p
=1
,...,N
p
6
=
n
Crosstalk
p,n
=3
N
+
N
(
N
−
1) =
N
2
+2
N
The target test tubes in the multi-tube ensemble,
⌦
, are indexed by
h
=1
,...,
|
⌦
|
.
L
max
=2
for all tubes (i.e., each target test
tube contains all off-target complexes of up to 2 strands). Final sequence designs for orthogonal cgRNAs/trigger pairs A, B, C
are shown in Table S5.
Reactants for system
n
•
cgRNAs: G
n
•
Triggers: X
n
Elementary step tubes for system
n
•
Step 0
G
n
tube:
products
0
n
⌘
G
n
;
reactants
0
n
⌘;
•
Step 0
X
n
tube:
products
0
n
⌘
X
n
;
reactants
0
n
⌘;
•
Step 1
n
tube:
products
1
n
⌘
{
G
·
X
}
n
;
reactants
1
n
⌘
{
G, X
}
n
Crosstalk tubes for system
n
•
Crosstalk tubes:
products
crosstalk
p, n
⌘
{
G
n
,X
p
}
;
reactants
crosstalk
p, n
⌘;
, for each non-cognate cgRNA/trigger pair (
p
6
=
n
)
Sequence constraints
•
Assignment constraints: portions of the cgRNA are constrained to match standard gRNA sequences for use with dCas9
(shaded gray in Figure 4b, Figure S5 and Table S5), the synthetic terminator for the trigger is fully constrained (shaded tan
in Figure 4b, Figure S5 and Table S5).
•
Watson–Crick constraints: cgRNA sequence domain “d” is constrained to be complementary to the trigger sequence domain
“d*” (shaded blue in Figure 4b, Figure S5 and Table S5).
•
Assignment constraint: cgRNA domain “u” is constrained to be complementary to a subsequence of the target gene mRFP
(full template sequence in Section S2.3, constrained sequence shaded orange in Figure 4b, Figure S5 and Table S5).
•
Pattern prevention constraints: the following patterns are prevented for cgRNA sequence domain “d”:
AAAA, CCCC,
GGGG, UUUU, KKKKKK, MMMMMM, RRRRRR, SSSSSS, WWWWWW, YYYYYY
.
Defect weights
•
Test tube weight for each elementary step tube: 1
•
Test tube weight for crosstalk tubes: 1
•
Nucleotide weights are depicted for each complex in Figure S6
S9
Step 0
X
n
Crosstalk
p,n
Step 1
n
X
n
10nM
10nM
G
n
·X
n
Step 0
G
n
G
n
10nM
G
n
20nM
X
p
200nM
Crosstalk Tubes
Constrained by target gene
Constrained by dCas9
Constrained by synthetic terminator
Sequence Constraints
Elementary Step Tubes
Complementarity between G
n
and X
n
Tube
On-targets (
on
h
) Off-targets (
o
↵
h
)
Step 0
G
n
G
n
L
L
max
0
G
n
Step 0
X
n
X
n
L
L
max
0
X
n
Step 1
n
{
G
·
X
}
n
{
G, X
}
n
[
L
L
max
1
n
Crosstalk
p,n
{
G
n
,X
p
}{
G
n
·
X
p
}
[
L
L
max
crosstalk
p,n
Figure S5: Target test tubes for sequence design of orthogonal toehold switch cgRNAs.
Top: Target test tube schematics. Bottom:
Target test tube details. Each target test tube contains the depicted on-target complexes (each with the depicted target structure and target
concentration) and the off-target complexes listed in the table (each with vanishing target concentration). The on-target structures depicted
above are used in the mechanism schematic of Figure 4b. To simultaneously design
N
orthogonal systems, the total number of target test
tubes is
|
⌦
|
=
N
2
+2
N
.
L
max
=2
for all tubes. Domain shading reflects sequence constraints. Design conditions: RNA in 1 M Na
+
at 37
◦
C.
G
n
X
n
G
n
·
X
n
w = 200
w = 200
w = 200
w = 200
w = 200
w = 200
w = 1
w = 50 (4nt loop)
w = 100
w = 100
w = 300
w = 25
w = 200
w = 200
w = 1
w = 50
w = 300
w = 1
Constrained by target gene
Constrained by dCas9
Constrained by synthetic terminator
Complementarity between G
n
and X
n
Figure S6: Nucleotide defect weights for sequence design of toehold switch cgRNAs.
Within the target test tubes of Figure S5, the
nucleotides in a given sequence domain within a given complex are assigned a defect weight
w
as depicted.
S10
S1.2 Methods for bacterial studies in
E. coli
S1.2.1 Plasmid construction and molecular cloning for bacterial cgRNA studies
Sequences for parts used in bacterial studies are provided in Section S2. Plasmid layouts for each construct as well as example
plasmid maps and corresponding full sequences are provided in Sections S3.1-S3.4.
Control gRNA and cgRNA constructs were generated by inverse PCR, inserting sequence modifications into the previously
described pgRNA-bacteria vector
5
(Addgene plasmid #44251; gift from S. Qi). All PCR steps for the generation of exper-
imental constructs were performed using Q5 Hot Start High-Fidelity polymerase (NEB #M0494) according to manufacturer
instructions using primers designed using standard molecular cloning techniques and synthesized by Integrated DNA Technolo-
gies. Introduced sequences were verified by Sanger sequencing for single colony picks via colony PCR using GoTaq Green
polymerase (Promega #M7122).
For terminator switch cgRNAs (Figure 2), trigger-expressing constructs were generated by cloning lacI-regulated promoter
(BioBrick part number BBa
R0011), trigger template, and synthetic terminator (BBa
B1002) directly into the cgRNA vector
via inverse PCR. For splinted switch cgRNAs (Figure 3), trigger-expressing constructs were generated by first cloning syn-
thetic promoter, trigger template, and synthetic terminator (BBa
B1006) into a trigger-only cassette via inverse PCR, and then
cgRNA+trigger expressing constructs were cloned by inserting trigger cassette into the cgRNA vector using BioBrick assem-
bly.
6,7
For toehold switch cgRNAs (Figure 4), trigger-expressing constructs were generated by first cloning synthetic promoter,
trigger template, and synthetic terminator (BBa
B0050) into a trigger-only cassette via inverse PCR, and then cgRNA+trigger
expressing constructs were cloned by inserting trigger cassette into the cgRNA vector using DNA assembly according to man-
ufacturer instructions (NEBuilder HiFi DNA Assembly, NEB #E2621).
A lacI+dCas9 expression construct was generated by inserting a lacI template sequence with J23108 constitutive promoter
8
into the previously described pdCas9-bateria vector
5
(Addgene plasmid #44249; gift from S. Qi) between the dCas9 gene and
the p15A origin with a synthetic terminator (BBa
B0010) added upstream of lacI as a transcriptional terminator for dCas9 (see
Section S3.4), using DNA assembly according to manufacturer instructions (NEBuilder HiFi DNA Assembly, NEB #E2621).
S1.2.2 Bacterial culture and silencing assay for cgRNA studies
A previously described
E. coli
MG1655 strain with constitutively expressed mRFP and sfGFP inserted into the nfsA locus
5
(Ec001; gift from S. Qi) was used for all fluorescence assays. For experiments with constitutive expression of trigger, the
previously described pdCas9-bacteria vector
5
(Addgene plasmid #44251) was used for tetR-regulated dCas9 expression. For
experiments with lacI-regulated expression of trigger, the lacI+dCas9 vector was used for tetR-regulated dCas9 expression and
constitutive expression of lacI. Chemically competent chloramphenicol-resistant cells carrying either the dCas9 or lacI+dCas9
construct were transformed with gRNA, cgRNA, or cgRNA+trigger expression vectors and cultivated in EZ-RDM (Teknova
#M2105) containing 100
μ
g/mL carbenicillin and 34
μ
g/mL chloramphenicol (EZ-RDM+Carb+Cam).
Sequence-verified strains were grown overnight in EZ-RDM+Carb+Cam, then seeded at 100
⇥
dilution in 100
μ
L fresh
medium and grown at 37
◦
C with shaking in the Neo2 microplate reader (Biotek) to monitor absorbance at 600 nm. When cells
had reached mid-log phase (
⇡
4 h), cells were again diluted
⇡
100-fold in fresh medium with cell density normalized by A600
and, if applicable, dCas9 expression and trigger expression were induced with aTc and IPTG, respectively, in
N
=3
replicate
wells at 400
μ
L final volume in a 96-well high-volume glass bottom plate (Matriplate, Brooks #MGB096-1-2-LG-L). A final
working concentration of 200 nM aTc and 5 mM IPTG was used for terminator switch experiments (Figure 2). A final working
concentration of 2 nM aTc was used for splinted switch experiments (Figure 3). A final working concentration of 200 nM aTc
was used for toehold switch experiments (Figure 4). Induced cells were grown at 37
◦
C with continuous shaking for 12 h.
S1.2.3 Flow cytometry for bacterial cgRNA studies
Protein fluorescence was measured using the MACSQuant VYB flow cytometer (Miltenyi Biotec) using FSC/SSC to gate for
20,000 live cells per well (see example of Figure S7) at a flow rate of 25
μ
L/min. sfGFP fluorescence was measured using the
B1 channel (488 nm laser, 525/50 nm filter) and mRFP fluorescence was measured using the Y2 channel (561 nm laser, 615/20
nm filter).
S11
b
a
FSC-A
SSC-A
10
-1
10
1
10
3
10
5
10
-1
10
1
10
3
10
5
FSC-A
SSC-A
10
-1
10
1
10
3
10
5
10
-1
10
1
10
3
10
5
10
1
10
3
10
5
Fluorescence intensity (au)
0
1000
2000
Counts
No-target gRNA, ungated
10
1
10
3
10
5
Fluorescence intensity (au)
0
1000
2000
Counts
No-target gRNA, gated
Figure S7: Illustration of gates used for flow cytometry analysis of
E. coli
.
(a) Scatter plots for ungated sample (top) and gated sample
(bottom): side scatter area (SSC-A) vs. forward scatter area (FSC-A). (b) Fluorescence intensity histogram for ungated sample (top) and
gated sample (bottom). mRFP fluorescence for no-target gRNA control used for terminator switch characterization in Figure 2c.
S12