Reconstruction of the Global Neural Crest Gene Regulatory Network in Vivo

Reconstruction of the global neural crest gene regulatory network in vivo

Ruth M Williams1, Ivan Candido-Ferreira1, Emmanouela Repapi2, Daria Gavriouchkina1,4, Upeka Senanayake1, Jelena Telenius2,3, Stephen Taylor2, Jim Hughes2,3, and Tatjana Sauka-Spengler1,∗

Supplemental Material

∗Lead and corresponding author: Tatjana Sauka-Spengler ([email protected]) 1University of Oxford, MRC Weatherall Institute of Molecular Medicine, Radcliffe Department of Medicine, Oxford, OX3 9DS, UK 2University of Oxford, MRC Centre for Computational Biology, MRC Weatherall Institute of Molecular Medicine, Oxford, OX3 9DS, UK 3University of Oxford, MRC Molecular Haematology Unit, MRC Weatherall Institute of Molecular Medicine, Oxford, OX3 9DS, UK 4Present Address: Okinawa Institute of Science and Technology, Molecular Genetics Unit, Onna, 904-0495, Japan A

25 25 25 25 25

20 20 20 20 20

15 15 15 15 15

10 10 10 10 10 log2(R1_5-6ss) log2(R1_5-6ss) log2(R1_8-10ss) log2(R1_8-10ss) log2(R1_non-NC) 5 5 5 5 5

0 r=0.92 0 r=0.99 0 r=0.96 0 r=0.99 0 r=0.96 0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 15 20 25 log2(R2_non-NC) log2(R2_5-6ss) log2(R3_5-6ss) log2(R2_8-10ss) log2(R3_8-10ss) 25 25 25 25 25

20 20 20 20 20

15 15 15 15 15

10 10 10 10 10 log2(R1_5-6ss) log2(R2_5-6ss) log2(R1_8-10ss) log2(R2_8-10ss) log2(R1_non-NC) 5 5 5 5 5

0 r=0.94 0 r=0.96 0 r=0.95 0 r=0.96 0 r=0.95 0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 15 20 25 log2(R3_non-NC) log2(R4_5-6ss) log2(R3_5-6ss) log2(R4_8-10ss) log2(R3_8-10ss)

25 25 25 25 25

20 20 20 20 20

15 15 15 15 15

10 10 10 10 10 log2(R2_5-6ss) log2(R3_5-6ss) log2(R2_8-10ss) log2(R3_8-10ss) log2(R3_non-NC)

5 5 5 5 5

0 r=0.97 0 r=0.95 0 r=0.98 0 r=0.95 0 r=0.98 0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 15 20 25 log2(R2_non-NC) log2(R4_5-6ss) log2(R4_5-6ss) log2(R4_8-10ss) log2(R4_8-10ss)

B C Downregulated 8-10ss vs 5-6ss Upregulated 8-10ss vs 5-6ss vi vii viii 10 10 10 CLDN1 35 CYP26A1 9 9 9

8 8 8 30 OTOGL ed counts

mali z 7 7 7 CadM1 ADAMTS1 Col6A1 No r 25 Col2a1 EDNRB CYP26A1 6 6 6 Lmo4 Pax3 PPP1R1C FGFR2/3 Wnt7a/b 234 DACT1 PCDH18 202 FZD6 244 alue) FST 5-6ss 8-10ss non-NC 5-6ss 8-10ss non-NC 5-6ss 8-10ss non-NC 20

PTN ix x xi 10 10

BPIFB3 10 −log10(p v

15 AGTRAP

MEF2C

9 9 9 ADAMTS1

NOXO1 8 10 LTK ADAMTSL1 8 8 BMP5 ed Counts DMBX1 CITED4 MYL4 WNT7B BRINP2 CXCR4 RXRG PPL

mali z

EDNRB 7 7 NEMP1 7 HES5 SOX10 BMP5 Oct1 BRPF3 FBXO2 No r Draxin TFAP2d 5 COL9A1 LHX9 FZD7 PCDH12 Notch2 Sox2 Chd2,4,5 Id2 TFAP2e 6

6 FGF8 Zic2 Cramp1 ILK Wnt8c 6 Kit Wnt’s3-9b LIMK1/2 WISP1 393 FST 341 Pcdh8 Zic1 762 5-6ss 8-10ss non-NC 5-6ss 8-10ss non-NC 5-6ss 8-10ss non-NC −3 −2 −1 0 1 2 3 log2FoldChange xii xiii 10 10 9 9 8 8 ed Counts FGF22 RARB Cldn16-18 Wnt5 7 mali z

FGFBP1 Hand1 7 Cx3cr1 FZD8

NR2E3 E2F1 No r SFRP5

6 6 223 323 5-6ss 8-10ss non-NC 5-6ss 8-10ss non-NC

Supplementary Figure S1. RNA-seq quality control and reproducibility. Related to figure 1 (A) Scatter plots showing correlation of RNA-seq replicas. r= Pearson correlation co-efficient. (B) Volcano plot of genes up and downregulated at 8-10ss compared to 5-6ss. (C) Clusters (vi-xiii) of highly correlated genes identified by WGCNA.

2 Stacked Cl7 Cl7 (D) 20 15 Scatter plots ● Cl6 Cl6 Cl5 Cl5 k-Cluster-7 k-Cluster-6 k-Cluster-2 k-Cluster-3 1 0.95 0.9 0.85 0.8 0.75 0.7 5-6ss 5-6ss_neg 8-10ss 8-10ss_neg HH4 ● Cl4 Cl4 ● ● ● ● ● 5-6ss_negs 15 Heatmap and merged 10 Cl3 3 Cl3 5-6ss_negs ● Cl2 Cl2 (E) ● 5-6ss_negs ● ● Cl1 Cl1

●

5 0 5 0 log(readcount_closest_TSS_8-10ss) ● log(readcount_closest_TSS_non-NC) 10 10 5-6ss NC 10 5 iance

8-10ss_negs ar 0 ● Percentage of genes with multiple Cl7 Cl7 ●

8-10ss_negs ● Cl6 Cl6 Number of enhancers per gene 8-10ss_negs 5 (G)

Cl5 Cl5

PC1: 36% v

10 5 HH4 0 Cl4 HH4 Cl4 -3 ●

Cl3 Cl3

60 50 40 30 20 10 5-6ss NC 0

Cl2

Cl2 8-10ss NC enhancers multiple with genes % -Cluster is also shown. ● k Cl1 G

8-10ss NC Cl1

5 0

log(readcount_closest_TSS_5_6ss) xpressed_non-NC) e 10 log(readcount_closest_ 10 9.0 7.5 6.0 4.5 3.0 1.5 0.0 12.0 10.5 -6 HH4 HH4 Cl7 5-6ss NC 5-6ss NC F’ e 1.0kb 8-10ss NC 8-10ss NC 5-6ss_negs 5-6ss_negs 5-6ss_negs HH4 Cl6 8-10ss_negs 8-10ss_negs 8-10ss_negs -Cluster elements and gene expression levels in NC cells (F,

Matrix presenting the correlation coefﬁcients to all possible

2.5 0.0

k -2.5 iance r a v PC2: 20% 20% PC2: -1.0 centr B Cl5 C (B) Cl4 e 1.0kb Cl3 15 15 15 HH10 Somites Cl2 -1.0 centr Cl1 3 10

e 1.0kb

10 10 5 0

10 xpressed_5_6ss) e log(readcount_closest_ s non-NC r=0.98 r=0.97 r=0.99 8-10s -1.0 centr F 5 5 5 log2(R2_HH4) e 1.0kb log2(R3_5-6ssNegs) s non-NC log2(R3_8-10ss_Negs)

0 0

5-6s

15 10 5 0 15 10 5 0 15 10 5

0 100% -1.0 centr

log2(R2_5-6ss_Negs) log2(R1_8-10ss_Negs) log2(R1_HH4) s e 1.0kb 15 15 15 80% R3 8-10s -1.0 centr ) annotated to closest expressed gene and closest TSS. s 0 10 10 10 PCA comparing NC and non-NC cells at both stages and HH4 ATAC-seq samples. e 1.0kb r=0.98 r=0.90 r=0.97 R2 8-10s (C) 60% -1.0 centr 5 5 5 log2(R3_5-6ss_Negs) s Violin plots showing correlation between -Cluster elements. The number of elements in each log2(R2_8-10ss) e 1.0kb k ) log2(R3_8-10ss_Negs) on 0

0 R1 8-10s 15 10 5 0 TTS 5' UTR e 3' UTR 15 10 5 0 15 10 5 0

-1.0 centr log2(R1_5-6ss_Negs)

log2(R2_9ss_Negs) 40% log2(R1_8-10ss) (F, F ax7) e 1.0kb 15 s (P 15 15 TSS - R3 5s 20% -1.0 centr genic on 10 s omoter 10 10 e 1.0kb r=0.98 r=0.98 r=0.92 pr intr Inter R2 5-6s -1.0 centr 5 5

5 0% 6 3 1 9 2 6 1 2 8 7 3 5 log2(R2_5-6ss) 8 6 6 4 8 s 1 log2(R2_5-6ss_Negs) 7 6 7 6 5 -means linear enrichment clustering of ATAC signal across all samples/stages analysed. Pax7 sample is NC cells isolated using 0 _ _ _ _ _ 2 e 1.0kb k 2 3 4 6 7 _ log2(R2_8-10ss_Negs) l l l l l

1 C C C C C

Cl5_7740

0 0 l

R1 5-6s 15 10 5 0

15 10 5 0 15 10 5 0 C

log2(R1_8-10ss_Negs)

log2(R1_5-6ss) -1.0 centr log2(R1_5-6ss_Negs) 7740 el. 7740 el. 6479 el. 5832 7681 el. 7681 7816 el. 7816 el. 6623

6 2 el. 20156

-Cl6 -Cl7 -Cl5 -Cl3 -Cl4 -Cl2 -Cl1 k k k k k k k E A D k-Cl1 k-Cl2 k-Cl3 k-Cl4 k-Cl5 k-Cl6 k-Cl7 bar plot showing genomicproﬁles annotation depicting of the Pax7-195 enhancer (Fig.green, S3). 5-6ss, purple, 8-10ss) and non-NC cells (F showing correlation of ATAC-seq replicas,pairwise comparisons r= of Pearson replicates/samples. correlation co-efﬁcient. associated enhancers. Supplementary Figure S2. ATAC-seq quality control and chromatin accessibility dynamics. Related to Figures 2 and 3. (A) A B

putative enhancer 12 Nanotagged Dissect cranial Nanostring Assay putative enhancers regions, (inc.controls) extract RNA

BsmBI/T4 Ligase TK C Cerulean tag# 1-16 pA pro 1000 TK lacZ Citrine pA 900 pro tag#17-32 800 TK Cherry tag#33-48 pA pro 700 HH4 6-10ss 600

500

400

300 Nanostring count

200

100

0 D Nanotagged Enhancers Snai2 Ets1 Msx1

enh-332 enh-332 enh-241 enh-242 enh-370 enh-117 enh-264

Foxd3 Sox10 Tfap2a

enh-372 enh-372 enh-193 enh-84 enh-84 enh-99a enh-185

Pax7 Tfap2a

enh-194 enh-195 enh-195 enh-195 enh-199 enh-74 enh-249 Pax7 Tfap2b

enh-216 enh-218 enh-143 enh-368 enh-32 enh-36 enh-226

Supplementary Figure S3. Multiplexed high-throughput enhancer screening. Related to Figures 3 and 4. (A) Schematic depiction of enhancer cloning strategy. (B) Cartoon showing ex ovo electroporation technique and Nanostring assay. (C) Bar graph representing typical Nanostring results. Nanostring count (of nanotag transcripts) above 50 (green) was determined to reﬂect in vivo enhancer activity. (D) In vivo activity of selected enhancers.

4 A 500kb

NC 5-6ss RNA-seq uc_338 tfap2b tfap2d PKHD1

RBC

diff. mean e-C

aptu r stats C

log padj.

50kb tfap2b tfap2d

5-6ss diff.bind C-seq 8-10ss

AT A diff.bind

non-NC

Cl4 Cl9 Cl3

B pax7

500 kb

RBC e-C

diff. aptu r

C mean

stats

pax7 50kb

5-6ss diff.bind

C-seq 8-10ss

AT A diff.bind

non-NC

Cl4 Cl9 Cl3

Supplementary Figure S4. Capture-C at Tfap2b and Pax7 loci. Related to Figure 3. Genome browser views of chromosome conformation capture (Capture-C) from the Tfap2b (A) and Pax7 (B) promoters and associated statistical analysis of the differences between Capture-C profiles in neural crest (NC, n=4) and red blood cells (RBC, n=4). Raw counts of unique interactions mapped to each restriction fragment were analysed using the bioconductor package DESeq2. (A) shows ∼1Mb around the Tfap2b locus, the topologically associating domain mostly spanning the region downstream of the Tfap2b coding region. (B) shows ∼2Mb around the Pax7 locus, the topologically associating domain mostly spanning the gene body and downstream of the Pax7 coding region. The red line denotes the position of the capture probe. The top track shows the NC RNA-seq output and the second and third tracks show the Capture-C normalised counts from raw count per restriction fragment from NC cells (in blue) and RBCs (in grey), respectively. Purple track shows differential mean between NC and RBC profiles specifically highlighting proximal and distal interaction blocks, with NC-specific interactions overlapping distal cis-regulatory elements (ATAC-seq tracks and mapped analysed k-Cluster and Diffbind elements). The majority of differences are with elements that interact more strongly in NC than in RBCs and the DESeq2 analysis highlighted these interactions as being statistically significant. Statistical significance is presented in the form of the DEseq2 Wald statistics track (in red), which determines significance of difference in interactions between NC and RBCs, and is calculated as a ratio of LogFoldChange values and their standard errors. Blue track showing the log transformed p-values mapped across the locus indicates the significant differences between NC and RBC profiles, and points to both proximal and distal interactions. 5 A

10 Fzd3 (19) Chd7 (27) 5 Sox10 (73) H3K27Ac signal (x1000) Tfap2b (243, 720)

0 0 5 10 15 20 25

Enhancers ranked by H3K27Ac signal (x1000)

B 84 85 87

1.2 1.2 1.2

1.0 1.0 1.0

0.8 0.8 0.8

0.6 0.6 0.6

0.4 0.4 0.4

0.2 0.2 0.2

0 0 0 5 5 6 6 7 6 6 6 7 7 8 6 6 7 8 ss ss ss

89 99 10E2 1.2 1.2 1.2

1.0 1.0 1.0

0.8 0.8 0.8

0.6 0.6 0.6

0.4 0.4 0.4

0.2 0.2 0.2

0 0 0 6 6 7 7 8 8 5 6 6 7 7 8 5 6 6 7 7 8 ss ss ss

84+99 84+85+99 1.2 1.2

1.0 1.0

0.8 0.8

0.6 0.6

0.4 0.4

0.2 0.2 ng Sox10 right ng Sox10 left 0 0 5 6 7 7 8 8 6 6 7 7 8 8 ss ss

Supplementary Figure S5. Decommissioning super-enhancer elements controlling Sox10 expression. Related to Figure 4. (A) Enhancers ranked by H3K27ac signal from 5-6ss, using the ROSE algorithm, top-ranked genes are annotated. (B) qPCR for Sox10 following dCas9-Krab mediated decommissioning of associated enhancers using bilateral electroporation (Fig. 4K-M). Sox10 on the left (experimental) side of embryos shown in magenta, right (control) side shown in blue. Error bars show standard deviation.

6 A 0.4 B

DiffBind_5-6ss Frequency in DiffBind 5-6ss DiffBind_8-10ss Frequency in DiffBind 8-10ss xA 0.3

Frequency in background X Frequency in background

o H Dmbx1 Oct4 Ho xK1 ESSRA/ xC O

0.35 o s TF H A REST MAFF A F XD13 RXRA T HO A EBF1 G Elf ELF5 xA1 MEIS1 o X SO F FO X9 Oct4 Stat X/ 0.25 SO T 0.3 F ZIC1/2/3 AP2a E2F6 AX2/3/5 P NR2 Zbtb3 Zic1/2/3 ZBTB3 0.25 0.2 SIX1 Zic2 NR3C1 RFX5 Sox5 AX8 TEAD1_2 P Mzf1 Zic3 STAT 0.2

P Nr2E3 HIC1

ax7 0.15 TF

AP2

0.15 10 X Tcf Lhx2 0.1 SO

S O 0.1 X Msx2 8

So x10 0.05 ATF 0.05 SMAD3 Lef1 Sox2 NT GCR AR

f f f

f 0

2 2 3 2 3 5 3 2 2 2 2 3 1 1 2 2 2 2 2 2 2 2 2 2

7 7 7

f f

at at

x x x F c c c

x x x x x

8 0 2 0 2 8 2 3 0 3 1 8 8 5 5 1 6 0 1 2 8 5 3 9 5 9 6 8 9 3 0 0 3 3 0 3 2 3 1 3 0 2 x

c t

x x x x x

b3 b3 b3 b3 b3 b3

ic ic

T T

z z

i ic ic

f f

T /

o o o

x ox x x x x x x x x x x x x t t t t

F F

P P P X P X

P 1 1 b d 1 1 b 1 1 1 1 P2a 1 C o

l l

Z Z Z

i i

t A 2

F F

a a 7-T 1-T 2-

bt b b b

o o

o L o o o o o

x x

x x x

F F

2 S 2 S P

2 AP2a AP2a AP2a

A E A E A A A

-So - -Sox -

D 1

- / / OX10 / / /

2-St S 2-S S -So -Sox - -Pax

2-Z

2- 1 2- - -

F x x

P P

S S

S S S t S S S

D E E F F F

o o o

o o o o

-Z - - T F F F

F f F

R R

e-So

1-Lhx c R

2-Sox

2-So

x 2-So 1-Pa 2-Pax

2-M 1-M

x x

R A C

R x

2 A 2

1 - 1 - -

------

- -

i 2-M x x

m F

b3-Z

b Z Z

-SOX10 -

4_Sox

z T T T t

T T

e-Z

S S

A-Z 1-Z

2-Z S x x S n S S S S

- -T

A t

x x x t

c c -

K1-Lh o o

F F n

N K1-So

F -

C - T a -

- -

Pax

Sox N 3 -

0 - F -

3 - 3 0 - - -

R S 0

1-SOX10

2-SOX10

i k

s o o

s t

Elf

1-T

2-T

ox E

x b

Lhx t m

n c t -

1 /

/ T at

i Lh Lhx

S S S

8 So

-lik

Lef F

A 8 H

3 3

x r

3 3 8 1 1

2-Z

l A

Lhx

Lhx 1 T

bx Lhx

EBF

R o -

Sox - -

o n

x T

- F

z F

n A

Lh T

H -

M M x n

D H R T

2 A

x x

A Sox

a t

A 1

GC F l

D F

T F A T r

R f

8 o

M 0

EBF

a a

2-Oc Lhx

o A / EBF

x E

A x A 4_So

A A

a 1 x d

t S M x

P D S

P P 1

S 36799-SOX10 S

AT a

A x S

o x

S -

r x 4_So

T H

S H

s Lh

o o

m c

s A H

S S

E A G C Diffbind 5-6ss Diffbind 8-10ss k-Cluster 3 k-Cluster 9 Sox10-Sox8/9 Sox10−Sox8/9 Otx2-Pax2 Lhx2−Sox2 0.35 0.25 0.20 0.30 0.15

0.20 Arnt−Sox8/9

Sox8-TFAP2 Sox8/9−TFAP2

0.25 ATF/Sox8/9

ATF−Sox8 0.15

ATF−Sox10 ATF-Sox10 Sox8/9−Sox/Fox Sox8/9-Sox/Fox Pax2−Sox9 0.15

0.20 Arnt−Sox10 Arnt-Sox10 0.10

Sox10−Sox/Fox Sox10-Sox/Fox Smad3−Sox8/9 Smad3-Sox8/9

Lhx2−Zbtb3 0.10 Lhx2−Tcf NR2F2−Pax2

0.15 Sox10−TFAP2 Sox10-TAFP2 0.10

Frequency of co-occupancy events Arnt−TFAP2 Arnt-TFAP2 Smad3-Sox10 Otx2−Sox9 Arnt−ATF Smad3−Sox10 Arnt-ATF Lhx2−SOX10 Arnt−Sox/Fox TFAP2-Sox/Fox Sox2−Zbtb3 TFAP2−Sox/Fox ATF-TFAP2 ATF−TFAP2 NR2F2−Otx2 0.10 GCR−Lhx2 ATF−Sox/Fox ATF-Sox/Fox Sox2−Tcf Arnt−Smad3 0.05 Arnt-Smad3 Lhx2−TFAP2a Lhx2−Pax7 ATF-Smad3 Lef1−Lhx2 ATF−Smad3 Smad3-TFAP2 0.05 Lhx2−Msx2 Smad3−TFAP2 SOX10−Sox2 Smad3−Sox/Fox

0.05 Smad3-Sox/Fox GCR−Sox2 NR2F2−Sox9 Lef1−Sox2 GATA−Pax2 Sox2−TFAP2a Sox2−Pax7 0.05 Msx2−Sox2 Lhx2−Mzf1 Pax7−Zbtb3 Msx2−Zbtb3 GATA−Otx2 Lhx2−Zic1/2/3 Sox8−Sox9 Sox8-Sox9 Pax7−TcfTcf−TFAP2a TFAP2a−Zbtb3 Sox10−Sox9 Sox10-Sox9 NR2C2−Pax2 Sox2−Zic1/2/3 Sox/Fox−Sox9 Zbtb3−Zic1/2/3GCR−Pax7 TFAP2−Sox9 Arnt-Sox9 NR2C2−Otx2 Lhx2−Zic2Elf−Lhx2Tcf−Zic1/2/3 Arnt−Sox9 Sox/Fox−Otx2 REST−Sox2 Sox2−Zic2Lhx2−REST Arnt−Pax2/3/5 Rxra/Essra−Pax2 0.00 0.00 ATF−Sox9 0.00 Pax3-Sox/Fox 0.00 Msx2−Pax7 GCR−Zic1/2/3 Pax3−Sox8

0 20 40 60 80 0 50 100 150 200 250 300 0 50 150 150 0 10 20 30 40 50 P −log10(FDR-corrected -value)

Supplementary Figure S6. Combinatorial transcription factor binding in DiffBind elements. Related to Figure 6. (A, B) Motif co-occurrence frequencies and circular networks plots showing putative TF combinatorial binding interactions in DiffBind 5-6ss (A) and DiffBind 8-10ss (B). (C) Showing frequency of TF co-occupancy events and statistical signiﬁcance in DiffBind 5-6ss, 8-10ss and k-Cluster-3 and k-Cluster-9

7 A HH8 - HH10 D 4 2 0 −2 −4

z-score sgRNAs Control X10 X2 targeting scrambled x9 AP2A AP2B AP2C TX2 TF1 TF2 TF3 TF4 AX2 selected factors + sgRNA + ARNT ARNT2 A A A A BCHE BMPR1A CXCR7 ETS1 NR2C1 NR2C2 NR2F2 O P PDGFRA PLXNA4 PTPRM RARB S O S O S o T F T F T F Cas9 Cas9 RNA-seq RNA-seq HH4

B Differential expression analysis

12 FOXG1

Arnt2

10 Atf2 Sox9 Sox10 TFAP2b

PLK2 NR2C1 (padj) 10 −log

FZD5

SIX3

WNT8B DACT2

C8orf22 SIX6 COL9A1 DRAXIN PAX3 SOX10 HESX1 NR2E1 PLXNA4 PTGFRN ALX1 MSX1 ALDH1A1 PCDH8 LHX9 BMPER ENC1COL9A3 Meis2a.1 PCDH18 SOX2 PICK1 LHX2 0246 8

−10 −5 0 5 10

log2FoldChange C 12

WNT4

Otx2

10 Pax2 Lrrn1 INPP5E NR2F2 TIMM1 Arnt2 Atf2 FOXD3

(padj) EFNB2 OST4 10 AVL9 TMEM258 PLAGL2 FGFR3 PAQR3 PTEN −log LRRTM3 septin 2 TMEM141 SMARCAD1

BMPR1A SOX11 ADAMTS14 S1PR1 Pax2 ADAMTS19 WDR1 NKX6.2 PARK2 ENC1 SOX4 TFAP2C FGFR1 PAX3 SOX2 PSEN1 CXCR4 NKX3.2 EPHA4 TFAP2B SEMA6D CLASP2 PCDH7 LRP8 SEMA3D NR2F2 GBX2 KLF7 PLK2 SEMA3B SOX21 TWIST2 ADAM33 NR2C1 TCF7L2 FOXG1 MYB PLXNA4 TWIST1 BCHE TFAP2A TNNI2 FOXJ2 BMP2 CCL17 FOXD1 0246 8

−10 −5 0 5 10 x9 X2 TF1 TF2 TF3 TF4 o TX2 AX2 X10 A A A A ETS1 S P O ARNT AP2A S O AP2C AP2B RARB log FoldChange O 2 BCHE ARNT2 S CXCR7 NR2F2 PTPRM NR2C1 NR2C2 T F T F T F PLXNA4 PDGFRA BMPR1A

Supplementary Figure S7. Perturbation of core NC-GRN transcription factors. Related to Figure 6. (A) Schematic of bilateral electroporation assay for CRISPR mediated perturbation experiments. (B, C) Volcano plots showing up- and down-regulated genes following CRISPR knock-out of core TFs associated with k-Cluster-3 (B) and k-Cluster-9 (C) (D) Heatmap showing single-cell co-expression of targeted TFs and selected dysregulated genes following CRISPR knockout of core TFs..

8 Table S1. Capture-C targets and oligo sequences

GalGal4 location Capture oligo sequence

Sox10 chr1:50,912,325-50,912,444 TTTTCAAATCAGGGGACAGTGATGCTGTGGCAGGGACTTACAGAGGTGACTGCAGACAGTGAGAGAGGAGGGGTGCACAGGGCAGGCAGCTCAGGTCCTGGGCTTCTCTTCAAGTTGATC GATCAAGGACATGCTGGGGATGTGGATACCACAAGCGTGTTGGTGGTGGCGGGGGTAAGGAAGGGGAGCTCAGCCCTGGGTTGCACAACTTCACATCCCTTCCTGTGACCAGCAACTCCA Sox9 chr18:9,068,402-9,068,521 AGCCGGGCTGCGCGCTGGTGGAGACTCCGTCTCTGCCGGCTTTACTTCTTGTTTTTAACCCTTCCCCGCCCCTCAGCCGCCCGGTTGTTTTTTTTCTCTCCGTTTTCTCCTCCCCTGATC GATCCGCGGGAACCCCTCCGGCACGCAGCGCACGGACTTCGGCGCCTGGGAAGCCCGAAGCCGTCGCGGGGCGGAGCGAAGAGAAGCGCAACGGCTCCCACCGCCCCGCGCCCCGCCCCG Lmo4 chr8:14,676,982-14,677,101 GCTTTTTTAATGGAGACGGAGGGAGGGACGCGCGGAGAGCTGGCAATTTGTAGGACGAAAATGGATGCTTAATTCACGTCTCGGTTTTAATTAGGTGATTCACCGGATTTCTCCCGGATC GATCTTCCCCCCGCAGCCGGCGCACCTCTTCCACGAGAGGGAGCCCGCCGTCACCGGGGGCGGCTGCGCGCTGCCGCCGGGGTTCACCATGGCCTGCGGGAGAGCGGCGCGGGTCAGGGG Pax7 chr21:4,443,114-4,443,233 TTTGGGGCGGTTGGAGCTCCTTTCCCACGCCGCGCCTTTCCCCGAGCAGCTGTGCCGCTTTGCTCTTTATTTCTCCTCCCGTTTCAAGTAGTGAGGAGCCGGCTTTCAGAAGCCAGGATC GATCCAATTCATTAAGGATGCTAATGAAGGAGGTGCGTCGGGAGCCGCCGAGCGCAGAGCGGAGGGGTTTGGGTCTCATTTCGGCCCTATATACGGGGGGGTGATGTGGAAGGAATGCTG Snai2 chr2:108,126,987-108,127,106 GATCCTCTTTGAATAACTGAGTTCAAGTGGATGAACAAACTTCATGATTCATTCCGAGCAGCGCTGACATATTTGTCCGAACTGCCTCACTGTAAGCACGGCAAAAAAACAGAGCTAGCG GGGAGAAGCGCAGCTCTTCACAGCACTGAGGGCAAAGCTGCTGCTTTCCTTCACTGTACAGAAACGATTAAATCCACTTTTGGAAGGGACGTTCGTGACGCCGGTCGCCTCTCAGCGATC TFAP2b chr3:107,873,143-107,873,262 GATCCATCTATAATTGGAAATGGGGGACAGACACCAAATCCGACGTTCCTCTTCCATCGCAACTTATATCTGTTGTCTGAAACAATAGGCTGCAGAAGTAAACCTCAATCGGATAGTAAA GACATATATAAAGTGACCCATTTATATATGTAAATTATATATATTCTCGCTTATGTATGGATTTACATAGGCACATGTATGCTACACGTTACATATGCATATATGCATACAAACCTGATC Zfhx4 chr2:119,024,967-119,025,086 TTTTGGAACAGCTGTAAATTAGTGATGAGCTATTAGTGAGCTGTGTCATTATTTAATAAAAATGGCTTCTCTCACCTTATTTTTTATCCAGGTCCCTGACAGGCTGGATGAAATGAGATC GATCTGCAAGGACTCGGAGTAGCTTAGGCTGTAATCAGGCTACTATTACAGTGGCTGGAGCCTTGCAGGCTCCCAAAAAAATGAAGGAAAGCAGACTTTTAACCAATGTGTGACCAACTT Globin chr14:12,097,459-12,097,578 GATCCTAACACTAACCCCAGCTCGCGTCGGGGTCCAACCCCCCCAGCCTGCGCAGTATCGTGGGTGGGGCAGGGCAGCAGCCCTGCCTGGCTGGGGTCCAGAATCTATGGGGCGGGCTGG ACAAGAACAACGTCAAGGGCATCTTCACCAAAATCGCCGGCCATGCTGAGGAGTATGGCGCCGAGACCCTGGAAAGGTAGGTGTCCTTCTCTGTCCTCCGGCTGCCTCTCTCCCCTGATC

9 Table S2. Primer design for mulitplex enhancer cloning and Nanostring screening.

BsmBI binding site Vector speciﬁc overhangs Target speciﬁc seqeunce ~20nt Cerulean vectors F TTTTTT CGTCTC ccatgg nnnnnnnnnnnnnnnnnnn Cerulean vectors R TTTTTT CGTCTC ggtcct nnnnnnnnnnnnnnnnnnn

Citrine vectors F TTTTTT CGTCTC gccagg nnnnnnnnnnnnnnnnnnn Citrine vectors R TTTTTT CGTCTC caacag nnnnnnnnnnnnnnnnnnn

Cherry vectors F TTTTTT CGTCTC gtgcag nnnnnnnnnnnnnnnnnnn Cherry vectors R TTTTTT CGTCTC caccgt nnnnnnnnnnnnnnnnnnn

Cerulean neg control oligo F TTTTTT CGTCTC ccatgg AGCTGGATCGATgatatcCGATCGATCGTAGCAC Cerulean neg control oligo R TTTTTT CGTCTC ggtcct GTGCTACGATCGATCGgatatcATCGATCCAGCT

Citrine neg control oligo F TTTTTT CGTCTC gccagg AGCTGGATCGATgatatcCGATCGATCGTAGCAC Citrine neg control oligo R TTTTTT CGTCTC caacag GTGCTACGATCGATCGgatatcATCGATCCAGCT

Cherry neg control oligo F TTTTTT CGTCTC gtgcag AGCTGGATCGATgatatcCGATCGATCGTAGCAC Cherry neg control oligo R TTTTTT CGTCTC caccgt GTGCTACGATCGATCGgatatcATCGATCCAGCT

10 Table S3. Guide RNA sequences

Epigenome engineering of Sox10 enhancers Targeted knock-out of core TFs

Target/sgRNA sgRNA seqeunce Target/sgRNA sgRNA seqeunce 84_sgRNA_1 AGTCTGCCACCCATCAAAGC ATF2_sgRNA_1 TCAACAACTGAAACACCGgt 84_sgRNA_2 CCATTGTATCATGCTGGACA ATF2_sgRNA_2 cttgctgttttcagGCATCA 84_sgRNA_3 CTCCACTGAACGAGTCCATG 84_sgRNA_4 ATTAATTCCTGCGAACAGAA TFAp2b_sgRNA_1 GGAGGAGTGCTGAGAAGgta 84_sgRNA_5 CCCTTTGTGTATGGGCTCAC TFAp2b_sgRNA_2 accctcgcttacCTTCCACC 85_sgRNA_1 AGATGTGCTTATGGGCTCCT 85_sgRNA_2 TGGGAACAATGTCAACTCCG Sox10_sgRNA_1 ttccctccccagTGAGAAGA 85_sgRNA_3 GCACAGAGCGGCCCCGTCG Sox10_sgRNA_2 ggtaggaaaacttacATTGC 85_sgRNA_4 TTCAGTACAGCTACTTACAG 85_sgRNA_5 TCTTTCCACCCGCCCAGGGC Arnt2_sgRNA_1 AGGGACCCAGCAAATTTTCA 87_sgRNA_1 CAGGAAGAAATGCGTAGTGA Arnt2_sgRNA_2 tcttttgtttatagGTATGA 87_sgRNA_2 GAGCGAGCAGAGAGTGGAGC 87_sgRNA_3 TCTTTGTTCCCTGCCTTTAA NR2C1_sgRNA_1 CTCTTTACCGCAGCGTATAC 87_sgRNA_4 CTCTAAAACACCCGATTGTC NR2C1_sgRNA_2 AGACAACTCTCCCAATGAGC 87_sgRNA_5 GCAGGGAAGGAGGATTCTGA 89_sgRNA_1 AGGGCATCCCCATGCACAAC Sox9_sgRNA_1 CTCTCATTCAGCAGCCTG 89_sgRNA_2 GAGGCAACAAATCTTTTCCA 89_sgRNA_3 AGGCAACTCACTGAGCATGA 89_sgRNA_4 GGGGAGAGTAAATGAGACAG 89_sgRNA_5 CAGTCAGTTGGGCTGCAGAG 99_sgRNA_1 GGTGAGAAATGTTGAAAACG 99_sgRNA_2 GTGTGTGACTCTTTTGTTCC 99_sgRNA_3 TGGACAACTTTGCTAGGCCC 99_sgRNA_4 GCACAAAGGAAATGGATAAA 99_sgRNA_5 AGGAAAGAGGGAAGGAAGGC 10E2_sgRNA_1 AGCAGGAGCAGGGAAACAAT 10E2_sgRNA_2 AAACATAAGCACAAACTAGG 10E2_sgRNA_3 TGGTAAGGATGGCCTGGATC 10E2_sgRNA_4 GCTGGGGAGGGGAGGCGGGC 10E2_sgRNA_5 CCATATCAACCATTCTCCAG

Scrambled control TGCAGTGCTTCAGCCGCT