of 8
advances.sciencemag.org/cgi/content/full/3/12/eaao5520/DC1
Supplementary Materials for
Deep tissue optical focusing
and optogenetic modulation with
time
-
reversed ultrasonically encoded light
Haowen
Ruan, Joshua
Brake, J.
Elliott
Robinson, Yan
Liu, M
ooseok
Jang, Cheng
Xiao,
Chunyi
Zhou,
Viviana
Gradinaru, Changhuei
Yang
Published 8 December 2017,
Sci. Adv.
3
, eaao5520 (2017)
DOI: 10.1126/sciadv.aao5520
Th
is
PDF file includes:
method S1. Calculation of the focal spot size of
TRUE
and conventional f
ocusing.
method S2. Viral injection surgery.
method S3. Electrophysiological recordings.
method S4. Daily alignment procedure.
fig. S1. Setup.
fig. S2. Electrical signal flow diagram.
fig. S3. Ultrasound pulse
-
echo image of the tip of the glass pipette ele
ctrode.
fig. S4. Electrophysiological photocurrent traces from neurons in 500
-
and 300
-
μm
-
thick acute brain slices.
fig. S5. Electrophysiological photocurrent and membrane voltage traces
comparing ultrasound on and off conditions.
Supplementary Materials
method S1. Calculation of the focal spot size of
TRUE
and conventional focusing.
To calculate the focal spot size of TRUE focusing from the images captured with the observation camera
(Camera2 in
f
ig.
S
1), the image was first cropped to a region (82 × 82 pixels) surrounding the focus
profile, and was low
-
pass filtered by a 3 × 3 averag
ing filter to smooth the speckle. Then, the focus
profile was fit using a 2D Gaussian function
2
22
00
( , )
exp
(
)
2
fxyAxxyyB

   

, where
A
,
B
,
x
0
,
y
0
,
are the fitting parameters. The full width at half maximum (FWHM) focal spot size was
obtained by FWHM =
2 2 ln 2
. The focal spot size of the conventional focusing was calculated in a
similar way. However, since the focal spot size was m
uch larger than that of TRUE focusing, the fitting
was applied to a region composed of 1040 × 1392 pixels.
method S2. Viral injection surgery.
Mice were anesthetized with an isoflurane gas/carbogen mixture and placed to a stereotaxic frame (David
Kopf Inst
ruments, CA, USA). After sterilizing the skin with chlorohexidine, a midline incision was made
with a sterile scalpel, the skull surface was cleaned with autoclaved cotton swabs, and the bregma and
lambda were identified and leveled to be on the same z
-
axi
s. AAV
-
DJ
-
CaMKII
-
bReaChES
-
TS
-
YFP (500
nL) was injected into the prelimbic cortex (AP + 1.65 mm, ML ± 0.2 mm, DV − 2.25 mm) using a blunt
35
-
gauge microinjection needle within a 10 μL microsyringe (NanoFil, World Precision Instruments, FL,
USA), which was c
ontrolled by a microsyringe pump (UMP3, World Precision Instruments) connected to
a controller (Micro4, World Precision Instruments). Following injection, the needle/syringe was held in
the same location for an additional 10 min to allow further diffusion.
To prevent potential backflow, the
needle/syringe was slowly withdrawn over approximately 10 min. After infusion, the scalp was sutured
and the mouse was returned to the vivarium for post
-
operative recovery. Animal husbandry and all
experimental procedure
s involving animal subjects were approved by the Institutional Animal Care and
Use Committee (IACUC) and by the Office of Laboratory Animal Resources at the California Institute of
Technology under IACUC protocol 1650
.
method S3. Electrophysiological recor
dings.
Mice expressing CaMKII
-
bReaChES
-
TS
-
YFP were euthanized with carbon dioxide and transcardially
perfused with ice
-
cold sucrose
-
based cutting solution saturated with 95% O
2
/5% CO
2
(carbogen) that
contained (mM) 85 NaCl, 75 sucrose, 2.5 KCl, 1.25 NaH
2
PO
4
, 4.0 MgCl
2
, 0.5 CaCl
2
, 24 NaHCO
3
and 25
glucose. The brain was removed and 300
800 μm thick coronal slices that contained the mPFC were
prepared in oxygenated cutting solution using a vibratome (VT
-
1200, Leica Biosystems, IL, USA). Slices
were recovered
at 32°C for one hour prior to recording in carbogenated ACSF containing (mM): 125
NaCl, 2.5 KCl, 1.2 NaH
2
PO
4
, 1.2 MgCl
2
, 2.4 CaCl
2
, 26 NaHCO
3
, and 11 glucose. Slices were transferred
to the recording chamber and perfused (1.5
2.0 mL/min) with carbogen
-
sa
turated ACSF at 32 ± 0.5°C.
Whole
-
cell patch clamp recordings were obtained using a Heka EPC 10 USB amplifier/digitizer (Heka
Electronik, Germany) with data sampled at 10 kHz and filtered at 2 kHz. Patch electrodes had resistance
of 4
8 MΩ and were fille
d with a potassium gluconate intrapipette solution (mM): 135 K gluconate, 5
KCl, 5 EGTA, 0.5 CaCl2, 10 HEPES, 2 Mg
-
ATP, and 0.1 GTP.
method S4. Daily alignment procedure.
Before operation each day, the system was tuned to maintain optimal performance of
the digital optical
phase conjugation (DOPC) system by ensuring fine pixel
-
to
-
pixel alignment between the camera and the
SLM. This procedure was conducted using the quality assurance (QA) arm of the system by opening
shutters SH1 and SH2 as shown in
f
ig.
S
1. The following steps were performed for the daily alignment.
1.
Playback beam normal to SLM.
To ensure the SLM was perpendicular to the sample and
playback beam (S/PB)
, we enabled the flip mirror FM
and tilted the SLM to maximize the
intensity measured by t
he QA photodiode (PD2).
2.
Record and playback a scattering field from QA path for DOPC loop B.
We disabled flip
mirror FM
and optimized the alignment of DOPC loop B, which was enabled by setting the
shutters BSS2 and BSS3 (
f
ig.
S
1
B
). This allows the QA ligh
t beam to interfere with the reference
beam (R), so the phase map of the scattered light could be measured using four
-
step phase
-
shifting holography. After measuring the phase map, the conjugate phase map was displayed on
the SLM, so the playback beam woul
d propagate through the scattering medium (SM) and the
single mode fiber (SF2) to reach PD2. The single mode fiber enables higher quality correction
than previous quality assurance systems which use a camera to monitor the phase conjugation
quality.
3.
Digitally search for tilt and shift misalignment (loop B).
To assess the respective contributions
of the shift and tilt misalignment between the camera and the SLM, the phase conjugate solution
was digitally shifted and tilted while monitoring the intensit
y of PD2. This allows the
misalignment to be mapped out before physically correcting the misalignment. Although we can
digitally correct for the shift and tilt, we prefer to correct these parameters physically to reduce
the computational time and improve t
he system performance during TRUE focusing operation.
4.
Physically correct for the tilt and shift misalignment (loop B).
The tilt misalignment was
corrected by tilting mirror M10 to maximize the signal captured by PD2. Similarly, the shift
misalignment was c
orrected by translating the SLM in plane to maximize the detected signal on
PD2.
5.
Aberration correction (loop B).
The aberration of the SLM, the playback (PB) beam, and the
associated optics in the light
-
path were digitally corrected by adding a compensati
on phase map
determined using a series of Walsh
-
Hadamard patterns
(
76
)
. The compensation map was saved
for later use during the TRUE focusing procedure.
6.
Aberration correction for the CW laser.
Since loop B w
as also used for the continuous
-
wave
playback beam, a compensation phase map for this beam was measured as well. In this case, we
switched the playback beam from
the
PW laser to
the
CW laser by switching the shutter BSS1.
7.
Alignment for DOPC loop A.
The PW
laser was enabled again by flipping shutter BSS1. DOPC
loop A was also enabled by flipping shutters BSS2 and BSS3. Steps 2 to 5 were repeated to align
DOPC loop A. Since loop A travels a different optical path, the physical tilt and shift correction is
dis
tinct from that of loop B. For loop A we corrected the tilt misalignment by tilting beamsplitter
BS4 and the shift misalignment by simultaneously tilting beamsplitters BS8 and BS4.
fig. S1. Setup.
(
A
and
B
)
S
chematic of the setup for DOPC Loop A
a
nd
Loop
B.
(
C
)
The physiological
setup
(the A
-
A section in
A
and
B
)
running in the neuron observation and patching mode.
(
D
)
The
physiological setup running in the neuron stimulation mode
.
Abbreviations: ACSF, artificial
cerebrospinal fluid; AOM, acousto
-
opti
c modulator; BB, beam block; BS, beamsplitter;
BSS, beam
selecting shutter;
CL, camera lens; CP, chopper; CW laser, continuous
-
wave laser; DL, delay line; FM,
flip mirror; HWP, half
-
wave plate; L, lens; LS, light source; M, mirror;
MF, multi
-
mode fiber;
ND
,
neutral
-
density filter; NP, Nomarski prism; OI, optical isolator; P, polarizer;
PB, playback beam;
PBS,
polarizing beamsplitter; PD, photodiode; PH, pinhole; PN, plane; PP, pipette; QA, quality assurance; R,
reference beam; S, sample beam; SC, sample cha
mber; SF, single
-
mode fiber;
SH, optical shutter;
SLM,
spatial light modulator; SM, scattering medium; TS, tissue slice; UST, ultrasonic transducer; ZB, zero
th
-
order block (a black 100
μ
m
-
diameter disk printed on a transparency).
fig. S2. Electrical
signal flow diagram.
The experiment can be divided into two phases. In the first phase,
the TRUE focusing system searches for the wavefront solution using
a pulsed
-
wave (PW) laser.
After that,
we
switch to
a continuous
-
wave (CW) laser by controlling the sh
utters
using software trigger
s
.
To obtain
a correct wavefront solution, we implement
nine
iteration
s
of
the
TRUE
focusing
process
by switching
the shutters that select between
DOPC loop A and B. In each DOPC loop,
a
four
-
step phase
-
shifting
holography approach
is
used to measure the
optical
field of
the
ultrasound modulated light. In this case,
an arbitrary
function generator (AFG1)
is
used to synchronize the camera and
the
phase
of the reference
beam
.
T
he frequency of
the reference beam
is
shifted
by 50.020 MHz by AFG2. The reference beam
shares
a
clock with
the
ultrasound pulses and lasers pulses, both of which are triggered at 40 kHz. A 7.46
μ
s delay
is
used to compensate for the propagation time of the ultrasound
wave
traveling to the ult
rasound
focus.
Abbreviations: AFG
, a
rbitrary
function
generator; AMP
, a
mplifier
; AOM
, a
cousto
-
o
ptic
m
odulator
;
DG
, d
elay
g
enerator;
UST
, u
ltrasound
t
ransducer.
fig. S3. Ultrasound pulse
-
echo image of the tip of the glass pipette electrode.
T
o
focus ultrasound to
the targeted neuron that is patch
-
clamped by the glass pipette, the ultrasound transducer is scanned to
image the pipette
tip
in
pulse
-
echo mode. The echo is recorded at each
scanning position
to form an
image of the tip where the neuro
n
is
locate
d
. The transducer is then moved to
th
at
location
so that
the
ultrasound focus overlap
s
with the
neuron. Scale bar
,
100
μ
m
.
fig. S4. Electrophysiological photocurrent traces from neurons in 500
-
and 300
-
μm
-
thick acute
brain slices.
TRUE focusing enables
photocurrent enhancement at depths of 500
μm
(
A
)
and 300 μm
(
B
)
in addition to the 800 μm thickness shown in the main text
.
fig. S5. Electrophysiological photocurrent and membrane voltage traces comparing ultrasound on
and off
conditions.
(
A
) Photocurrent traces when performing the TRUE focusing procedure with the
ultrasound
on and off. When the ultrasound is
on
,
a phase
map
is
measured which enables light to be
focused
to the location of the ultrasound focus
and the
neuron
cell
body
.
In contrast
, when the ultrasound
is
disabled
by turning off
its driving
amplifier,
an incorrect
phase map
is measured so that the playback
beam
is not
focus
ed
onto the
neuron
.
As a result, we see a decrease in the amplitude of the photocurrent.
(
B
)
Membrane voltage traces when performing the TRUE focusing procedure with the ultrasound on and
off.
W
hen the ultrasound is on
,
the
playback beam
is
focused onto the neuron and elicit action potentials.
In comparison, when the ultrasound is off, light is no
t focused onto the neuron and therefore no action
potentials are elicited. It should be noted that the photocurrent and membrane voltage are lower with
ultrasound off than those with ultrasound on followed by a shifted TRUE phase map displayed on the
SLM (
Fig. 3) which leads to an artificially increased photocurrent enhancement since the overall energy
in the “US OFF” condition is lower than in the “No Shaping” condition.
This is because the random phase
map measured with ultrasound off has a higher spatial
frequency (1/pixel size) than that of a normal phase
map, and these high frequency components are filtered out by the limited numerical aperture of the
optical system. In contrast, the shifted phase pattern in the “No Shaping” condition does not change th
e
spatial frequency distribution of the phase map measured with ultrasound on and therefore maintains the
same background intensity level as the background of the TRUE focus
.