Articles
https://doi.org/10.1038/s41592-020-0774-3
RecV recombinase system for in vivo targeted
optogenomic modifications of single cells or cell
populations
Shenqin Yao
1,8
, Peng Yuan
2,8
, Ben Ouellette
1
, Thomas Zhou
1
, Marty Mortrud
1
,
Pooja Balaram
1
, Soumya Chatterjee
1
, Yun Wang
1
, Tanya L. Daigle
1
, Bosiljka Tasic
1
, Xiuli Kuang
3
,
Hui Gong
4
, Qingming Luo
4
, Shaoqun Zeng
4
, Andrew Curtright
5
, Ajay Dhaka
5
, Anat Kahan
6
,
Viviana Gradinaru
6
, Radosław Chrapkiewicz
2
, Mark Schnitzer
2
, Hongkui Zeng
1
and
Ali Cetin
1,2,5,7
✉
1
Allen Institute for Brain Science, Seattle, WA, USA.
2
CNC Program, Stanford University, Palo Alto, CA, USA.
3
School of Optometry and Ophthalmology,
Wenzhou Medical College, Wenzhou, China.
4
Britton Chance Center for Biomedical Photonics, Wuhan National Laboratory for Optoelectronics, Huazhong
University of Science and Technology, Wuhan, China.
5
Department of Biological Structure, University of Washington, Seattle, WA, USA.
6
Division of
Biology and Biological Engineering, California Institute of Technology, Pasadena, CA, USA.
7
Present address: CNC Program, Stanford University, Palo Alto,
CA, USA.
8
These authors contributed equally: Shenqin Yao, Peng Yuan.
✉
e-mail: alic@stanford.edu
SUPPLEMENTARY INFORMATION
In the format provided by the authors and unedited.
NATuRe MeTHODS
|
www.nature.com/naturemethods
Supplementary Figure
1
Schematic representation of the DNA constructs generated for this study.
rAAV
constructs were used to
produce viruses expressing Rec
V
and the reporter
constructs
.
Supplementary Figure
2
In vitro
two
-
photon stimulation of DreV.
Mammalian cells were co
-
transfected with EF1a
-
EGFP, EF1a Dre red fluorescence reporter, EF1a
-
NDreV,
and EF1a
-
CDreV plasmids. Two
-
photon activation was conducted at 48 hrs after transfection at various
conditions, and reporter expression was observed 36 hr
s post stimulation. Two
-
photon activation conditions
were as follows:
λ = 900 nm, 90 mW, 1 ms/line (512 lines), 200 μm x 200 μm scan area, and duration of: 1) 3
mins (data not shown due to no signal), 2) 6 mins, 3) 9 mins, 4) 12 mins, 5) 12 mins (repeat),
6) 15 mins. In
condition 7 randomly selected single cells were scanned in 5 areas, 36 x 36 μm each, separated by 40 μm
roughly in a straight line across the plate with a duration of 1
-
10 seconds per area.
The experiments were
independently repeated twice w
ith similar results.
Top left scale bar: 100 μm.
Supplementary Figure
3
In vitro
efficiency comparisons of optogenomic modification constructs used in this study.
(a)
Relative fluorescence intensity of Cre
dependent and Dre dependent fluorescent reporters 48 hrs after 20
minutes of light induction for different co
-
expression constructs. The N
-
and the C
-
termini of the CreV
recombinase were combined using a variety of approaches. In the first construct, NCre
V and CCreV constructs
were mixed together. Constructs 2,3 and 4 contain NCre linked with VVD and CCre all within the same open
reading frame, with or without a 5 glycine linker. Constructs 5 and 6 contain both NCreV and CCreV linked by
the ribosome skippi
ng peptide PQR. Constructs 7 and 8 have NCreV and CCreV linked by IRES sequence
instead of PQR. Construct 11 is the DreV version of the most successful CreV co
-
expression construct.
Controls are represented by reporters alone.
(b)
Comparison of the light
-
i
nducible recombination mediated by
improved CreV and DreV. Cells were co
-
transfected with the appropriate Cre or Dre reporters, and
recombinase constructs 1. NCreV and CCreV, 2. Cre
-
Magnets 3. iCreV, 4. NDreV and CDreV, 5. Dre
-
Magnets
6. iDreV. Images were
taken 48 hours after 20 minutes of light induction.
(c)
Comparison of iC
reV with
improved CRY2 based light
-
inducible Cre recombination system. Cells were transiently transfected with
fluorescent reporter along with either iCreV or improved CRY2/CIB1 based
constructs. 48 hours after various
durations of light stimulation average fluorescence values were quantified with 4 replicas per condition.
Each
experiment is represented by 4 replicas. The line across the box represents the median, the lower and upper
h
inges correspond to the 25th and 75th percentiles, and the upper and lower whiskers extend from the hinge
to the largest or smallest values no further than 1.5 * inter
-
quartile range (IQR) from the hinge.
Supplementary Figure
4
In vivo
testing and comparison of iCreV with Cre
-
Magnets.
Ai14 tdTomato r
eporter
mice
received
RO injection of
PHP.eB
-
iCreV and Cre
-
Magnet viruses along with
EF1a
-
eGFP control virus
(n=2 per case)
.
For the indicated
groups l
ight stimulation
was
applied
on the left
hemisphere
t
wo
weeks after
virus injection
.
(a)
tdTomato expression in Ai14 mice injected with PHP.eB
-
Cre
-
Magnets.
4466
(left)
and 4244 (right) cells per section (CPS) were labeled with light stimulation,
and 43 (left)
and 60
(right)
CPS were labeled in the absence of light. Labeled cells in the absence of light are indicated by
white arrows. (b) tdTomato expression in Ai14 mice injected with PHP.eB
-
iCreV.
2858
(left)
and 4446
(right)
CPS
were labeled with
light stimulation, and no labeling was found in the absence of light.
Scale bars:
Overviews 1 mm, smaller brain segments 200
μm.
Supplementary Figure
5
Generation of a light
-
inducible Flp recombinase.
(
a
)
Schematic of iFlpV constructs and sequence of FlpO. Split sites are indicated with arrows and the construct
number.
(
b
)
Relative fluorescence intensity after transfection of iFlpV and reporter constructs, measured 48
hours after illumination for 20 minutes
.
The line across the box represents the median, the lower and upper
hinges correspond to the 25th and 75th percentiles, and the upper and lower whiskers extend from the hinge
to the largest or smallest values no further than 1.5 * inter
-
quartile range (IQ
R) from the hinge.
(
c
)
Heatmap of
fluorescence intensity after transfection of
iFlpV variants
and reporter constructs measured 48 hours after 20
minutes of light stimulation.
T
he 62 iFlpV variants, iFlpV and negative control were arranged in an 8x8 matrix
.
The last two brightest conditions are iFlpV2 (aa 27) with an additional linker sequence and iFlpV2 as a control.
Supplementary Figure
6
RecV viruses allow efficient light
-
mediated optogenomic
modifications at different loci.
Various reporter mice (n = 2 per case) received
RO
injection of
the indicated
PHP.eB rAAVs
. L
ight stimulation
for indicated groups
was conducted
on the left hemisphere two weeks post injection
. (
a
)
Light
-
induced n
uclear
-
localized tdTomato expression in
Ai75 mice (from the Rosa26 locus) injected with AAV
-
PHP.eB EF1a
-
Cre
virus
with
72669
(left)
and 172606
(right)
cells per slice
(
CPS
). (
b
)
Light
-
induced n
uclear
-
localized tdTomato
expression in
Ai75 mice injected with
AAV
-
PHP.eB EF1a
-
iCreV virus
with
9572
(left)
and 14211
(right)
CPS
.
(
c
)
Nuclear
-
localized tdTomato expression in
Ai75 mice injected with AAV
-
PHP.eB EF1a
-
iCreV virus
without
light stimulation
with
1
(left)
and 0
(right)
CPS
.
(
d
)
Light
-
induced
ChrimsonR
express
ion in
Ai167 mice (from
the TIGRE locus
33
) injected with AAV
-
PHP.eB EF1a
-
iCreV
with
161
(left)
and 2257
(right)
CPS
. Two coronal
planes are shown for each injection (top row) with enlarged views (lower two rows) for areas indicated by the
red boxe
s.
Scale bars: Overviews 1 mm, smaller brain segments 200
μm.
Supplementary Figure
7
Localized iCreV
-
mediated GCaMP6s expression and optical physiology within striatum.
(
a
) Virus application scheme and implant for one
-
photon illumination and GCaMP6s recordings. GCaMP6s
reporter mice were locally injected with 1:1 mixture of PHP.eB.iCreV and AAV5.CAG.tdTomato and implanted
with 400
μ
m optical fiber. After 7 days, fiber photo
metry signal was recorded as baseline activity, and 1P
illumination was performed by a 447nm laser using a 200 μm fiber, 5mW, 100ms pulses, 1Hz for 30 minutes.
At day 14 fiber photometry signal was measured again. (
b
) tdTomato expression and GCaMP6s expres
sion as
observed under the fiber tip after iCreV activation. (
c
) Fiber photometry activity in the striatum. ‘Before’
represents baseline activity before illumination
-
session 2, ‘after’ represents GCaMP6s activity a week after
illumination
-
session 3. (
d
)
Plots of ΔF/F over time and area, before and after illumination (n= 2 mice, 1 trial
each).
Supplementary Figure
8
Lower dose of CreV
viruses and shorter duration of light induction leads to sparse and strong labeling
of individual neurons.
Ai139 Cre
-
dependent EGFP
reporter
mice were injected with the 1:1 mixture of NCreV and CCreV rAAVs in
the visual or somatosensory cortex
, followed by indicated durations of light stimulation two
weeks after
injection. (
a
)
EGFP expression in Ai139 mice injected with
undiluted viruses,
and with
30 minutes of light
exposure; (
b
)
EGFP expression in Ai139 mice injected with
undiluted viruses,
a
nd with
5 minutes of light; (
c
)
EGFP expression in Ai139 mice injected with
undiluted viruses,
and with
3 minutes of light; (
d
)
EGFP
expression in Ai139 mice injected with
1:9 dilution of viral solution,
and with
5 minutes of light. Images are
maximum pro
jections of 100 consecutive fMOST images (each 1 μm
-
thick).
Each condition was repeated in
two mice, and fMOST images were obtained for one mouse per group.
Scale bars: 200 μm.
Supplementary Figure
9
In vivo
two
-
photon stimulation and sparse recombination using
i
CreV
.
(a)
Two
-
photon stimulated EGFP expression in
Ai139 mice
.
M
ice
received stereotaxic injections of a 1:5
mixture of EF1a
-
i
CreV:EF
1a
-
tdTomato into
VISp
, followed by
2P stimulation
t
wo to three
weeks
post
-
injection
.
D
iscrete 400
m x 400
m regions of layer 2/3, approximately 150
-
250
m below the
pial
surface (gray boxes),
were stimulated at
= 910nm for 15 minutes each. Two weeks later, animals were perfused. Scale bar is
1mm.
(b)
High magnificati
on images of
eGFP
-
filled neurons within approximate stimulated regions in (a). Scale
bar is 50
m. (
c
) Individual neuronal processes (right panels) and terminations (left panels)
in
multiple
subcortical regions. Scale bars are
10
m (left) and 30
m (right).
Experiments were repeated in two mice with
similar results.
Supplementary Figure
10
Additional examples of the
in vivo
single
-
cell 2P
-
induced labeling.
Additional two examples of the 2P induction experiment (15mW, 10min in
a
and 15mW, 3min in
b
), following
the same preparation and experiment protocol in
Figure 5b
.
tdTomato reporter Ai14 mice were locally injected
with
a mixture of
rAAV iCreV
and rAAV EGFP
viruses.
Arrows indicate the target cells. Asterisk indicates an
induced non
-
targeted cell. The background red signal observed in the pre
-
induction session is due to either
incomplete shielding of ambient light after surgery or high multiplicity of infect
ion related issues.
Experiments
were repeated in
five
mi
ce
with similar results
.
See
Supplementary Figure 11
for background expression rate.
Supplementary Figure
11
Background induction in the mouse cohort for the 2P
experiment.
(
a
) Example images of the first imaging session showed some cells already expressed tdTomato, indicating
some background induction prior to two
-
photon stimulation.
tdTomato reporter Ai14 mice were locally injected
with a mixture of
rAAV iCreV
and rAAV EGFP viruses.
Arrows indicate the induced cells. (
b
) Quantification of
this background induction rate showed an average of 3% (N = 8 mice). This induction was likely due to
incomplete shielding of the ambient light or high multiplicity of infecti
on. (
c
) Quantification of all the induction
plane depths for each field of view associated with the two
-
photon induction experiment (related to
Figure 5c
and
d
).
Supplementary Note
We chose VVD mainly because it is the smallest (450 bp) of all Light Oxygen or Voltage (LOV)
domain
-
containing proteins, and therefore suitable for fitting into viral vectors with limited genomic
capacity. In addition, the spectral properties of VVD
-
its
1P
excitation and emission drops sharply
to near 0 with light above ~520 nm
-
would allow us to
use other fluorescent proteins, activity
indicators or optogenetic molecules that work at longer wavelengths
1,2
. Upon light illumination,
VVD forms a homodimer
2
due to the conformational changes induced by Flavin Adenine
Dinucleotide (FAD) cofactor within the LOV domain. The FAD cofactor within the LOV domain
has a peak
one
-
photon (1P) excitation at 450 nm and a peak
two
-
photon (2P) activation at 900
nm wavelength
3
.
Supplementary Note References
1
Wang, X., Chen, X
. & Yang, Y. Spatiotemporal control of gene expression by a light
-
switchable transgene system.
Nat Methods
9
, 266
-
269, (2012).
2
Zoltowski, B. D.
et al.
Conformational switching in the fungal light sensor Vivid.
Science
316
, 1054
-
1057, (2007).
3
Huang, S., Heikal, A. A. & Webb, W. W. Two
-
photon fluorescence spectroscopy and
microscopy of NAD(P)H and flavoprotein.
Biophys J
82
, 2811
-
2825, (2002).