of 11
iScience, Volume
26
Supplemental information
Immediate responses to ambient light
in vivo
reveal distinct subpopulations
of suprachiasmatic VIP neurons
Anat Kahan, Karan Mahe, Sayan Dutta, Pegah Kassraian, Alexander Wang, and Viviana
Gradinaru
Supplementary
information
Figure S1. Monosynapt
ic tracing from SCN
VIP
neurons showing the distributio
n of neuronal sources, related
to Figure 1.
(
a
) Validation of the injected population with
vip
HCR probes in the SCN (top), and a higher-
magnification view (bottom), showing co-localization. The image
s are maximum intensity projections (MIPs) of 66
m, taken in 22 steps. (
b
) An example of a control mouse injected with AAV9.CAG.FLEX.eGF
P (day 1) and EnvA-
G.Rabies.mCherry (day 21). EnvA virus is not expressed unless T
VA is present. (
c-e
) An example of an experimental
mouse injected with AAV1.Syn.FLEX.splitTVA.eGFP (day 1) and Env
A-
G.Rabies.mCherry (day 21). mCherry is
expressed in a variety of neurona
l nuclei, such as the arcuate
nucleus (ARC, c), ventromedial hypothalamus (VMH,
c), paraventricular nucleus (PVA
, d), and the medial preoptic a
rea (MPA, e).
Figure S2. FP response amplitudes depend on the distance betwee
n the fiber tip and the SCN, and differ from
VIP x Ai140 (GFP) controls, related to Figures 2-3.
(a)
The FP response intensity (width x prominence) of SCN
VIP
GCaMP6s to 15 seconds of room light (6 repeats, 1.5E14 photons/
cm
2
/sec) vs. the distance between the recording
fiber tip and the center of the SCN. Width and prominence are p
arameters of the response peak, identified with
‘findpeaks’ in Matlab. Exponential fit is an estimation based o
n
1
, (n=11).
(b)
Same as (a) for response to 15 seconds
of 650 nm red light at 5.9E14 photons/cm
2
/sec (n=4). Note the
different scales.
(c)
Two examples of responses to red
light (c1 and c2 are #1 and #2 in (b), respectively).
(d)
LiGS histology, showing 3D fiber localization (gray) based on
vacancy of fluorescence and SCN
VIP
neurons expressing GCaMP6s stained with GFP-ab. Two fibers wer
e implanted
in (d1), and the recording fiber is marked with a red asterisk
(based on the relative signal amplitude between the two
fibers measured in vivo). (
e
) Mean integrated FP responses to light of VIP-Cre x Ai140(GFP)
mice (control, n=4),
comparing the seven wavelengths. Light intensities were matched
to have an equal number of 3.4±1E14
photons/cm
2
/sec at the bottom of the cage. (
f
) Individual dF/F responses to red light from one SCN of a VIP-
Cre x
Ai140(GFP) mouse, demonstrating a lack of response (six repeats
, gray; mean, thick line).
(g)
A statsitical comparison
between the VIP x Ai162 (GCaMP6s) from Figure 3 and VIP x Ai140
(GFP) responses to red light, n=3 and n=4
respectively (t-test).
(h)
Individual dF/F responses to blue and red light, from one SCN
of a VIP-Cre x
Ai162(GCaMP6s) mouse (left and right, respectively, individual
repeats in gray, mean values in thick line).
(i-k)
A
comparison between blue and red light responses as recorded by
two optical fibers implanted in the same mouse. The
low amplitude response in (j) (right fiber) and the 3D LiGS his
tology in (k) emphasize that p
recise fiber localization
can explain a low amplitude response. This also demonstrates th
at the responses to red light are visible (i, bottom, left
fiber) when the fiber is positioned well relative to the cells,
but at the noise level when the fiber is mislocated, as in (j,
bottom, right fiber). Plots are not z-scored, to give an accura
te comparison of the amplitudes. Scale bar is 100 mm.
Figure S3. Vivarium red light causes increased GCaMP6s activity
in SCN
VIP
neurons, related to Figure 3
. (
a
)
Spectra of white and red lights in vivarium. (
b
) Experimental setup with 15 seconds of white or red vivarium l
ight
with intensity of 4.49E14 photons/cm
2
/sec or 2.63E13 photons/cm
2
/sec, respectively, and examples of dF/F responses.
(
c
) Averaged dF/F responses to 15
seconds of white and red light
(n=6, 4 females, 2 males).
Mean values (black) and
SEM values (gray and red). Dashed lines represent light
exposure on-off times. (
d
) Comparison of termporal response
profiles of normalized dF/F activity. (
e
) Integrated responses to white and red light (gray and red, re
spectively) vs.
dark baseline (black). Nonparametric Kruskal-Wallis test, follo
wed by Šidák correction (*p < 0.05, ***p < 0.005).
Figure S4.
Additional information about SCN
VIP
GCaMP responses to blue (438 nm) and red (650 nm) light in
the presence of AA92593, related to Figure 4.
(
a
) A comparison of the response to blue light, before and after
DMSO only. (
b
) Exponential fits determine the decay constant for each condit
ion. (
c
) Comparison of additional
parameters before and after AA92593 (Opn4 antagonist) applicati
on in response to red and blue light. The differences
are insignificant except for the
mean number of peaks, as indic
ated in the figure. Data is
shown in Matlab 'boxplot’
presentation, showing the sample median and the 25th (bottom) a
nd 75th (top) percentiles of the sample.
Nonparametric Kruskal-Wallis test
, followed by Bonferroni corre
ction (*p < 0.05, **p < 0.01, ***p < 0.005).
Figure S5. Extended k-means clus
tering information, related to
Figure 5
. (
a
) Main PCA components for the three
clusters in Figure 5. (
b
) Main PCA components for two clusters, for comparison. (
c-e
) An alternative clustering based
on Gaussian mixture models, for comparison. (
c
) Cell population distribution in three clusters, based on k-me
ans
clustering. (
d
) The averaged temporal patterns of the three clusters in (c).
(
e
) Main PCA components. (
f
) Parameters
of the response profiles of each
cluster. Nonparametric Kruskal
-Wallis test, followed by Bonferroni correction (*p <
0.05, all clusters, ***p < 0.005, #p<0.0005, ‘boxplot’ Matlab p
resentation next to full data d
isplayed as asterisks).
Figure S6. Single-cell SCN
VIP
neurons’ averaged responses to 15 seconds of light at differen
t intensities and
colors, related to Figure 6. (a)
Plots for each light condition are presented in the same cell o
rder, and are ordered
based on the response to 1.4E15 photons/cm
2
/sec of 438 nm (blue) light. n
mice
=4
,
n
cells
=56. (
b
) Response to 15 seconds
of 438 nm ambient light, six repeats. (
c
) Same as (b) for 650 nm light. Light intensity 5.65±0.15 E14 p
hotons/cm
2
/sec,
n
mice
=3, n
cells
=17. Seven and 17 cells responsive to 650 and 438 nm light, res
pectively. (
d
) Clustering shows that
Clusters #1 and #2 are separated in their response to red light
(Kruskal-Wallis Test).
Table S1. Light intensities measu
red at the bottom of the cage,
related to Figure 2
:
Light Source
peak
WL/width
(nm)
Lumencor
control Lux mW
mW/cm^2 photons/cm^2/s
log(photons/c
m^2/s)
dark
0.00004
0.0000
0
room red light 610
na
35
0.0076
0.0081
2.49E+13
13.40
room white
light
560
na
234
0.051
0.054
1.53E+14
14.18
violet
395/23
41
3
0.156
0.166
3.30E+14
14.52
blue high
438/24
12
39
0.132
0.140
3.09E+14
14.49
cyan
473/10
31
145
0.117
0.124
2.96E+14
14.47
teal
513/17
100
829
0.128
0.136
3.51E+14
14.55
green
560/25
8
661
0.106
0.113
3.18E+14
14.50
orange
586/20
18
948
0.118
0.125
3.70E+14
14.57
red
650/13
67
441
0.116
0.123
4.03E+14
14.61
Light in lux units measured with the Light Meter smartphone app
(‘My Mobile Tools Dev’). Light power (mW) was
measured with a Thorlabs S120C photodiode. The energy units wer
e divided by the sensor dimensions, 0.94 cm
2
. Unit
conversion from mW/cm
2
to photons/cm
2
/sec was done using:
௣௛௢௧௢௡௦
∙퐸ൌ푛
௣௛௢௧௢௡௦
௛௖
, where
is the Planck
constant,
c
is the speed of light, and
is the light wavelength.
Table S2. Animal ID and sex and distribution of clusters in res
ponse to room white li
ght, using k-means
clustering, related
to Figures 5, 6.
ID
sex # cells
# cells,
cluster 1
# cells,
cluster 2
% cells,
cluster 1
% cells,
cluster 2
134R F 11
7
4
63.64
36.36
148RR F 21
18
3
85.71
14.29
168RL F 7
3
4
42.86
57.14
177RR F 19
8
11
42.11
57.89
264L F 6
6
0
100.00
0.00
310L F 11
11
0
100.00
0.00
264L F 9
4
5
44.44
55.56
50L
M 43
17
26
39.53
60.47
304RL M 29
12
17
41.38
58.62
303L M 10
4
6
40.00
60.00
308R M 8
6
2
75.00
25.00
Responses were recorded from
84 cells/7 females, and 90 cells/4 males, data combined from Fi
gure 5 and white light
session in Figure 6. Each individual had responses from two typ
es, cluster 1 and cluster 2, except two females with
17 cells together, which were identified as cluster 1 only. Mal
e distribution is 43% and 56% clusters #1 and #2,
respectively. Female distribution is 67% and 33% clusters #1 an
d #2, respectively.
Table S3. mRNA probes.
Oligos/probes for
vip
transcripts mRNA detection,
related to Figure S1
and STAR Methods
.
Probe id
Sequence
vip_V1_164_pA CACATTTACAgACCTCAAtaAGGCTTGCTTCTGGCTTCCA
vip_V1_164_pB ATCAGGAATGCCAGGAACTGatCCTACCTCCAACTCTCAC
vip_V1_213_pA CACATTTACAgACCTCAAtaGCGACTGAGAGAACAGCACA
vip_V1_213_pB TCCAAAGAGAGGCCA
GGCCAatCCTACCTCCAACTCTCAC
vip_V1_255_pA CACATTTACAgACCTCAAtaGCCTACTCACTACAGAAGGT
vip_V1_255_pB AAACGGCATCCTGTCATCCAatCCTACCTCCAACTCTCAC
vip_V1_310_pA CACATTTACAgACCTCAAtaTTTAAAGAGACTTGGTCAGG
vip_V1_310_pB GCAAGATGTCAGAGTCTGCTatCCTACCTCCAACTCTCAC
vip_V1_400_pA CACATTTACAgACCTCAAtaACTCCATCAGCATGCCTGGC
vip_V1_400_pB TGCTGTAATCGCTGGTGAAAatCCTACCTCCAACTCTCAC
vip_V1_468_pA CACATTTACAgACCTCAAtaCAATGAGTGACTCAAGGTAT
vip_V1_468_pB GCTGCTGCTGATTCGTTTGCatCCTACCTCCAACTCTCAC
vip_V1_510_pA CACATTTACAgACCTCAAtaTTGGCACAGGATCTTCCGAG
vip_V1_510_pB GGCATCAGAGTGTCGTTTGAatCCTACCTCCAACTCTCAC
vip_V1_570_pA CACATTTACAgACCTCAAtaCCATTTGCTTTCTGAGGCGG
vip_V1_570_pB GTTCAGGTATTTCTTC
ACAGatCCTACCTCCAACTCTCAC
vip_V1_620_pA CACATTTACAgACCTCAAtaCTCACTGCTCCTCTTTCCAT
vip_V1_620_pB AGAAAGTCTGCAGAATCTCCatCCTACCTCCAACTCTCAC
Supplementary Reference:
S
1. Pisanello, M., Pisano, F., Hy
un, M., Maglie, E., Balena, A.,
De Vittorio, M., Sabatini, B.L., and
Pisanello, F. (2019). The Three‐Di
mensional Signal Collection F
ield for Fiber Photometry in Brain
Tissue. Frontiers in Neuroscience
13
. 10.3389/fnins.2019.00082.