Supplementary
Materials
for
Single
-
cell nuclear architecture across cell types in the mouse brain
Yodai
Takei
et al.
Corresponding author
:
Long
Cai
,
lcai@caltech.edu
DOI:
10.1126/sci
ence
.
abj1966
The
PDF
file
includes:
Materials and Methods
Figs. S1 to S25
References
Other
Supplementary
Material
for
this
manuscript
include
s
the
following:
MDAR Reproducibility Checklist
Tables S1 to S4
1
Materials and Methods
Primary and readout probe design and synthesis
RNA seqFISH
primary probes for marker genes were designed similarly to our previous studies
(
17
,
39
,
40
)
. In brief, individual RNA species were encoded in fluorescent channel 1 (635 nm)
and 2 (561 nm) at each hybridization round and sequentially called. To implement this, 35
-
nt
RNA target binding sequences, four 15
-
nt unique readout probe binding sites encoded for each
RNA target, and a pair of 20
-
nt primer binding sites at the 5′ and 3′ ends of the probe for probe
generation were concatenated. Marker genes were selected based on previous studies with mouse
brain sam
ples
(
20
,
35
,
40
–
45
)
.
DNA seqFISH+ encoding strategy and primary probe design with GRCm38/mm10 mouse
genome are described in detail in our previous study
(
17
)
. In brief, a total of 80 serial rounds as a
16
-
base coding scheme with 5 rounds of barcoding was encoded in fluorescent channel 1 (635
nm) and 2
(561 nm) for the 1
-
Mb resolution data using a subset of barcodes in the codebook
(table S1). In addition, a combined strategy of diffraction limited spot imaging (60 serial rounds)
and chromosome painting (20 serial rounds) was encoded in fluorescent chan
nel 3 (488 nm) to
resolve 20 distinct regions (1.5
-
2.4
-
Mb in size) with 25
-
kb resolution. In total, 3,660 loci were
targeted including 2,460 loci at 1
-
Mb resolution for 20 chromosomes and 1,200 loci at 25
-
kb
resolution for the specific 20 distinct regions.
To implement this, 35
-
nt genomic DNA target
binding sequences, five 15
-
nt unique readout probe binding sites encoded for each barcoding
round, and a pair of 20
-
nt primer binding sites at 5′ and 3′ end of the probe for probe generation
were concatenated.
The repetitive element DNA FISH probes (LINE1, SINEB1, Telomere, MinSat) were designed
as described before
(
17
)
. Similarly to Telomere and MinSat probes, the MajSat probe (Integrated
DNA Technologies) was designed as a dye
-
conjugated 15
-
nt probe using the following sequence
(5′
-
TGTCCACTGTAGGAC), which directly targeted genomic DNA. In addition,
rDNA primary
probes (a total of 101 probes) targeting 45S pre rRNA gene (GenBank: X82564.1) were designed
similarly to DNA seqFISH+ probes except for containing four of 15
-
nt unique readout probe
binding sites specific to rDNA primary probes.
RNA seqFISH
and DNA seqFISH+ primary probes were generated from oligoarray pools (Twist
Bioscience) with enzymatic amplifications as previously described
(
17
)
based on Oligopaint
technologies
(
46
)
.
PCR primer sequences
for RNA seqFISH a
nd DNA seqFISH
+
primary probes
were selected from previous studies
(
17
,
35
)
.
Note that RNA seqFISH primary probes
were
amplified from multiple pools with different primer binding sites
and
then
combined for
experiments
.
The primary probe sequences used in this study are
publicly available
(
18
).
RNA seqFISH, DNA seqFISH+ and sequential immunofluorescence readout probes (5’ amine
-
modified DNA oligonucleotides, Integrated DNA Technologies) of 12
-
15
-
nt in length were
designed and conjugated to Alexa Fluor 647
-
NHS ester (Invitrogen A20006) or Cy
3B
-
NHS ester
(GE Healthcare PA63101) or Alexa Fluor 488
-
NHS ester (Invitrogen A20000) in
-
house as
described previously
(
17
)
.
For
RNA seqFISH readout probes
,
subset
s
of
DNA seqFISH+ readout
probes were selected.
The readout probe sequences used in this study are
publicly available
(
18
).
2
DNA
-
antibody conjugation
Preparation of the oligonucleoti
de
-
conjugated antibodies was performed, as described previously
(
17
,
47
)
. The BSA
-
free primary antibodies were purchased from commercial vendo
rs as listed
below. As a conjugation strategy, the crosslinking of 5’ thiol
-
modified 18
-
nt DNA
oligonucleotides (Integrated DNA Technologies) to lysine residues on antibodies was performed
via PEGylated SMCC cross
-
linker (SM(PEG)2) (Thermo Scientific Therm
o Scientific 22102).
As an alternative conjugation strategy, SiteClick R
-
PE Antibody Labeling Kit (Life
Technologies S10467) was also used to crosslink 5’ DBCO
-
modified 18
-
nt DNA
oligonucleotides (Integrated DNA Technologies) to the specific sites on prima
ry antibodies. The
oligonucleotide
-
conjugated primary antibodies were individually validated by SDS
-
PAGE gel
and immunofluorescence, and stored in 1× PBS at −80 °C as small aliquots.
The oligonucleotide DNA
-
conjugated primary antibodies used were as follo
ws: mH2A1 (Abcam
ab232602), H3K27ac (Active Motif 39133), H3K27me2 (Cell Signaling 9728BF), H3K27me3
(Cell Signaling 9733BF), H3K4me2 (Cell Signaling 9725BF), H3K9me3 (Diagenode MAb
-
146
-
050), H4K20me3 (Active Motif 39671), MeCP2 (Cell Signaling 3456BF), RN
A polymerase II
(phospho S5) (Abcam ab5408), SF3a66 (Abcam ab77800). Two antibodies (H3K27ac, RNA
polymerase II (phospho S5)) were excluded from the downstream analysis owing to the signal
dimness (H3K27ac) and incomplete penetration of the antibody (RNA p
olymerase II (phospho
S5)).
Tissue slice preparation
All animal care and experiments were carried out in accordance with Caltech Institutional
Animal Care and Use Committee (IACUC) and NIH guidelines. 6
-
7
-
week
-
old C57BL/6J female
mice were
obtained from The Jackson Laboratory.
To attach tissue sections, the 24 × 60 mm coverslips (VWR 16004
-
312) were cleaned by
sonication in 1 M sodium hydroxide and 100% ethanol in Ultrasonic Cleaner (Fisher Scientific
FS20) three times for 10 minutes each,
followed by a 15 minute sonication in 100% acetone. The
coverslips were then immersed in 2% (v/v) (3
-
Aminopropyl)triethoxysilane (Sigma A3648)
prepared in acetone for 2 minutes at room temperature. Then the coverslips were rinsed twice in
water and heat tr
eated at 90°C for 20 minutes. Next the coverslips were treated with 90 μg/ml of
Poly
-
D
-
lysine (Sigma P7280) in 1× PBS (Invitrogen AM9625) for 16 hours at room temperature,
followed by 3 times rinsing in nuclease
-
free water. The coverslips were then air
-
dri
ed and
attached to a microscope slide (VWR 48312
-
004) to facilitate the collection of cryo
-
sections.
The coverslips were freshly prepared as required for sectioning below.
Mice were perfused for 4 minutes with perfusion buffer (10 U/ml Heparin (Sigma
-
Aldr
ich
H3149), 0.5% Sodium nitrite (w/v) (Sigma
-
Aldrich 237213) in 1xPBS) upon isoflurane
anesthesia, followed by fresh 4% PFA (EMS 15714) in 1× PBS buffer for 4 minutes with a flow
rate of 5 mL/min through the peristaltic pump (Masterflex MP
-
07557
-
00). The m
ouse brains
were dissected out of the skull and immediately placed in 4% PFA in 1× PBS for 16 hours at
3
4°C. The brains were then immersed in 10% (w/v) RNAse
-
free Sucrose (Amresco 0335
–
2.5KG)
in 1x PBS for 1 hour at room temperature, then 20% (w/v) RNAse
-
fr
ee Sucrose in 1xPBS, until
the brains sank. Next the samples were incubated for 16 hours in 30% (w/v) RNase
-
free Sucrose
in 1× PBS at 4°C. After the brains sank, they were embedded in OCT (Sakura 4583) individually
and frozen in a bath of dry ice and ethan
ol. The samples were stored at −80°C until 10
-
15 μm
coronal sections were cut using a cryostat (Leica CM3050). The sections were immediately
placed on the functionalized coverslips described above. The tissue slices were stored at −80°C
until the tissue sl
ice experiment.
Tissue slice experiment
The combined sequential immunofluorescence, RNA seqFISH and DNA seqFISH+ sample
preparation was performed similarly to
our previous study
(
17
)
with some modifications for
tissue slice experiments. The tissue slice samples were dried, and permeabilized with 70%
ethanol pre
-
chilled to −20°C at room tempera
ture for 15 minutes. The samples were then dried
and further permeabilized with 8% Triton
-
X (Sigma
-
Aldrich 93443) in 1× PBS at room
temperature for 30 minutes after attaching a sterilized silicon plate (McMASTER
-
CARR
86915K16) with a punched hole to the co
verslip to use it as a chamber. The samples were
washed three times with 1× PBS and blocked at room temperature for 15 minutes with blocking
solution consisting of 1× PBS, 10 mg/mL UltraPure BSA (Invitrogen AM2616), 0.3% Triton
-
X,
0.1% dextran sulfate (Sig
ma D4911) and 0.5 mg/mL sheared Salmon Sperm DNA (Invitrogen
AM9680). Then oligonucleotide DNA
-
conjugated primary antibodies (see ‘DNA
-
antibody
conjugation’) with an estimated each concentration of 1
-
5 ng/μl were incubated in the blocking
solution with 100
-
fold diluted SUPERase In RNase Inhibitor (Invitrogen AM2694) at room
temperature for 18
-
24 hours. The samples were washed with 1× PBS three times and incubated
at room temperature for 15 minutes, fixed with freshly made 4% formaldehyde in 1× PBS at
room t
emperature for 5 minutes, and washed with 1× PBS six times and incubated at room
temperature for 15 minutes. The samples were then further fixed with 1.5 mM BS(PEG)5
(PEGylated bis(sulfosuccinimidyl)suberate) (Thermo Scientific A35396) in 1× PBS at room
te
mperature for 30 minutes, followed by quenching with 100 mM Tris
-
HCl pH7.4 (Alfa Aesar
J62848) at room temperature for 5 minutes. Then the samples were washed with 1xPBS and 70%
ethanol three times, and air dried by removing the custom silicon chamber.
Af
ter the immunofluorescence
preparation steps, custom
-
made flow cells (fluidic volume about
40 μl), which were made from glass slide (25 × 75 mm) with 1
-
mm thickness and 1
-
mm
diameter holes and a PET film coated on both sides with an acrylic adhesive with t
otal thickness
0.25 mm (Grace Bio
-
Labs RD481902), were attached to the coverslips. The samples were rinsed
three times with a 50% denaturation buffer consisting of 50% formamide (Invitrogen AM9342)
and 2× SSC, and incubated at room temperature for 15 minut
es. Then the samples were
optionally heated at 80°C for 3 minutes (this step was performed for replicate 1 and 2, and
omitted for replicate 3), and washed with 2× SSC twice. Then RNA seqFISH primary probe
pools (1
-
10 nM per probe) including mRNA, intron an
d non
-
coding RNA targets were
hybridized in a 50% hybridization buffer consisting of 50% formamide, 2× SSC and 10% (w/v)
dextran sulfate (Millipore 3710
-
OP). The hybridization was performed at 37°C for 3 days in a
humid chamber. After the hybridization ste
p, the samples were washed with a 40% wash buffer
consisting of 40% formamide, 2× SSC and 0.1% Triton X
-
100 at 37°C for 15 minutes, followed
4
by three rinses with 4× SSC, and stored at 4°C until imaging. Then the samples were imaged as
described below (see
‘seqFISH imaging’). Note that H4K20me3 immunofluorescence imaging
was also performed at this step for validation and alignment.
After the completion of RNA seqFISH imaging, the samples were prepared for DNA seqFISH
+
imaging. The samples were rinsed with 1× PBS, and incubated with 100
-
fold diluted RNase
A/T1 Mix (Thermo Fisher EN0551) in 1× PBS at 37°C for 1 hour. Then samples were rinsed
three times with 1× PBS, followed by three rinses with the 50% denaturation buf
fer and an
incubation at room temperature for 15 minutes. Then the samples were heated on the heat block
at 90°C for 7 minutes in the 50% denaturation buffer, by sealing the holes of the custom chamber
with aluminum sealing tapes (Thermo Scientific 232698)
. After the heating, the samples were
rinsed with 2× SSC and hybridized with a powder of DNA seqFISH+ primary probe pools,
which were dried by speed
-
vac and consisted of about 1 nM each DNA seqFISH+ probe, 1 μM
LINE1, 1 μM SINEB1 and 100 nM 3632454L22Rik l
ocus fiducial marker probes
(
17
)
,
resuspended in a 40% hybridization buffer consisting of 4
0% formamide, 2× SSC and 10% (w/v)
dextran sulfate. For the rDNA FISH experiment, about 20 nM each rDNA probe and 100 nM
3632454L22Rik locus fiducial marker probe were resuspended in the 40% hybridization buffer.
The DNA seqFISH+ primary probe hybridizatio
n was performed at 37°C for 3
-
5 days in a humid
chamber. After hybridization, the samples were washed with a 30% wash buffer consisting of
30% formamide, 2× SSC and 0.1% Triton X
-
100 at room temperature for 15 minutes, followed
by three rinses with 4× SSC.
Then samples were further processed to ‘padlock’ DNA seqFISH+ primary probes on the
genomic DNA in order to increase the stability of primary probes and prevent the loss of signals
during 80 rounds of DNA seqFISH+ imaging routines (see ‘seqFISH imaging’)
. A 31
-
nt global
ligation bridge oligonucleotide (Integrated DNA Technologies, 5′
-
TCAGTTGCAGCGCATGCTCGACCAAGGCTGG) was hybridized in a 20% hybridization
buffer consisting of 20% formamide, 10% dextran sulfate (Sigma D4911) and 4× SSC at 37°C
for 2 hours. T
he global ligation bridge was designed to hybridize to the 15
-
nt sequence of the
DNA seqFISH+ primary probes at the 5′ end and the 16
-
nt sequence at the 3′ end. Then, samples
were washed with a 12.5% wash buffer consisting of 12.5% formamide, 2× SSC and 0.
1% Triton
X
-
100 three times and incubated at room temperature for 5 minutes, followed by three rinses
with 1× PBS. The samples were then incubated with 20
-
fold diluted Quick Ligase in 1× Quick
Ligase Reaction Buffer from Quick Ligation Kit (NEB M2200) supp
lemented with an additional
1 mM ATP (NEB P0756) at room temperature for 1 hour to allow ligation reaction between the
5′
-
and 3′
-
ends of the DNA seqFISH+ primary probes. Then the samples were washed with the
12.5% wash buffer, followed by three rinses wit
h 1× PBS.
The samples were further processed for amine modification and post
-
fixation to increase the
stability of the primary probes across imaging rounds along with the ligation
(
17
)
. First, the
samples were rinsed with 1× labelling buffer A, followed by incubation with tenfold diluted
Label IT amine modifying reagent in 1× labelling buffer A
from Label IT nucleic acid modifying
reagent (Mirus Bio MIR 3900) at room temperature for 45 minutes. Second, after three rinses
with 1× PBS, the samples were fixed with 1.5 mM BS(PEG)5 in 1× PBS at room temperature for
30 minutes, followed by quenching w
ith 100 mM Tris
-
HCl pH7.4 at room temperature for 5
minutes. Then the samples were washed with the 55% wash buffer at room temperature for 5
-
15
5
minutes, followed by three rinses with 4× SSC, and stored at 4°C until imaging. Then the
samples were imaged for
DNA seqFISH+ and sequential immunofluorescence as described
below (see ‘seqFISH imaging’).
After the completion of DNA seqFISH+ and sequential immunofluorescence imaging,
Concanavalin A (ConA) staining and imaging
(
48
)
were performed for nuclear segmentation.
The samples were incubated with a ConA solution consisting of 20
-
100 μg/mL Alexa
Fluor 488
conjugate of ConA (Invitrogen C11252), 1× PBS and 0.3% Triton X
-
100 at room temperature
for 2 hours, and washed with the 55% wash buffer at room temperature for 2 minutes, followed
by a rinse with 4× SSC. Then the samples were imaged as described
below (see ‘seqFISH
imaging’).
Cell culture experiment
E14 mouse ES cells (E14Tg2a.4) from Mutant Mouse Regional Resource Centers were
maintained under serum/LIF condition as previously described
(
17
,
39
)
. NIH/3T3 cells (ATCC
CRL
-
1658) were cultured as previousl
y described
(
40
,
49
)
.
The combined sequential immunofluorescence, RNA seqFISH and DNA seqFISH+ protocol was
performed as previously described
(
17
)
for E14 mouse embryonic stem cells and NIH/3T3
mouse fibroblast cell line with ITS1 RNA FISH pro
be and rDNA FISH probes used in the tissue
slice experiments.
Image acquisition
Microscope setup
All imaging experiments were performed with the confocal fluorescence imaging platform and
fluidics delivery system similar to previous studies
(
17
,
39
,
40
)
. The microscope (Leica DMi8)
was equipped with a c
onfocal scanner unit (Yokogawa CSU
-
W1), a sCMOS camera (Andor Zyla
4.2 Plus), a 63× oil objective (NA = 1.40, Leica 11506349), and a motorized stage (ASI
MS2000). Fiber coupled lasers (635, 561, 488 and 405 nm) from CNI and Shanghai Dream
Lasers Technology
and filter sets from Semrock were used. The custom
-
made automated
sampler was used to move to the well of the designated hybridization buffer corresponding to
each hybridization round from a 2.0
-
mL 96
-
well plate (Corning 3960) and hybridization buffers
we
re moved through a multichannel fluidic valve (IDEX Health & Science EZ1213
-
820
-
4) to the
custom
-
made flow cell using a syringe pump (Hamilton Company 63133
-
01). Other buffers used
for the imaging routine were also moved through the multichannel fluidic va
lve to the custom
-
made flow cell using the syringe pump. The control of imaging and the automated fluidics
delivery system was achieved by a custom
-
written script in μManager
(
50
)
.
seqFISH imaging
The sequential hybridization and imaging routines were performed similarly to those previously
described
(
17
,
39
,
40
)
with some modifications. Briefly, the sample with the custom
-
made flow
cell was first connected to the automated fluidics system on the motorized stage on the
microscope. Fields of view (FOVs) (4 FOVs in replicate 1, 5 FOVs in replicate 2, an
d 8 FOVs in
replicate 3 of the mouse cortex) were registered using nuclei signals stained with DAPI solution