Articleshttps://doi.org/10.1038/s41592-019-0436-5

LADL: light-activated dynamic looping for endogenous gene expression controlJi Hun Kim1,5, Mayuri Rege1,4,5, Jacqueline Valeri1, Margaret C. Dunagin1, Aryeh Metzger1, Katelyn R. Titus1, Thomas G. Gilgenast1, Wanfeng Gong1, Jonathan A. Beagan1, Arjun Raj   1,2,3 and Jennifer E. Phillips-Cremins   1,2,3*

1Department of Bioengineering, University of Pennsylvania, Philadelphia, PA, USA. 2Epigenetics Institute, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA. 3Department of Genetics, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA. 4Present address: DST-INSPIRE Faculty, Department of Microbiology, Ramanarain Ruia Autonomous College, Matunga, Mumbai, India. 5These authors contributed equally: Ji Hun Kim, Mayuri Rege. *e-mail: jcremins@seas.upenn.edu

SUPPLEMENTARY INFORMATION

In the format provided by the authors and unedited.

NATuRE METHoDs | www.nature.com/naturemethods

Supplementary Figure 1

Overview of LADL components.

Functional light-induced looping requires three components: the dCas9-CIBN Anchor protein, the CRY2 Bridge and the gRNAs as the Target. We combined these components in a two-plasmid system as follows: (a) the LADL Anchor plasmid with the dCas9-CIBN Anchor protein co-transfected with the LADL Bridge + Target plasmid that has both the CRY2 Bridge and the gRNAs as Target. This plasmid combination is referred to as “LADL (Anchor + Bridge + Target)”. The three negative control combinations exclude one critical component each: (b) The “Empty anchor control” excludes the dCas9-CIBN Anchor protein but includes the CRY2 Bridge and the gRNAs as Target. (c) The “Empty target control” excludes the gRNA expression module but includes the dCas9-CIBN Anchor and CRY2 Bridge. (d) The “Empty bridge control” excludes the soluble CRY2 Bridge but includes the dCas9-CIBN Anchor and the gRNAs as Target. (e) The “One-sided gRNA control” excludes the two gRNAs targeting

the enhancer but includes the two gRNAs targeting the promoter, the soluble CRY2 Bridge and the dCas9-CIBN Anchor. For comprehensive plasmid maps please refer to Supplementary Figure 2.

Supplementary Figure 2

Complete list of intermediate and final LADL plasmids

(a) Plasmid S13.1. Backbone plasmid used to create the Anchor plasmids. Ampicillin resistant. (b) LADL Anchor plasmid with the dCas9-CIBN. CRY2/CIBN heterodimerization based Anchor plasmid with puromycin resistance. The relevant parts are: EF1A promoter (EF1α); three copies of the FLAG tag (3XFLAG); α – importin Nuclear Localization Signal (NLS); Cryptochrome 2 photolyase homology region (CRY2PHR); SV40 Nuclear Localization Signal (NLS); Glycine-Serine Linker (GS); SV40 Nuclear Localization Signal (NLS); dCas9 D10A H840A; Glycine-Serine Linker (GS); α – importin

Nuclear Localization Signal (NLS); Cryptochrome 2 photolyase homology region (CRY2PHR); SV40 Nuclear Localization Signal (NLS); 2A self-cleaving peptide (2A); puromycin resistance (Puro). This plasmid was used in Figures 1- 5 and Supplementary Figures 1, 4-11, 13 and is referred to as the “LADL Anchor” plasmid. (c) Empty anchor control plasmid. Puromycin resistance gene expressed from the EF1a promoter without any dCas9-CIBN Anchor protein. This plasmid was used as the puromycin resistant plasmid in the ‘Empty anchor control’ in Figures 1-2 and Supplementary Figures 1, 5, 7-10. (d) Plasmid S12.1. The modified Yamamoto Ampicillin resistant plasmid (Addgene #58768) capable of cloning a single gRNA when cut with BbsI and multiplexing four gRNAs after Golden Gate assembly with the companion B1, B2, B3 Spectinomycin resistant plasmids. This plasmid was used as a backbone to clone gRNA 129 shown in (e). (e) Single gRNA 129 cloned into the Ampicillin resistant S12.1 plasmid backbone using BbsI digestion. (f-h) Single gRNAs 135, 115 and 117 indicated were cloned into Spectinomycin resistant Yamamoto B1 (Addgene # 58778), B2 (Addgene # 58779), B3 (Addgene # 58780) plasmids respectively using BbsI digestion (Supplementary Table 4). The multiplexed assembly is shown in (i). (i) Empty bridge control plasmid. Multiplexed plasmid with gRNAs 129, 135, 115, 117 (Supplementary Table 5). No soluble CRY2 Bridge is expressed from this plasmid, so it is referred to as the “Empty bridge control” plasmid. This plasmid was used in the “Empty bridge control” condition in Figure 1, 5 and Supplementary Figures 5, 13. (j) CRY2olig-mCherry plasmid with BbsI sites mutated. Plasmid used to PCR amplify the mCherry cassette for use in gRNA cloning. The original CRY2olig plasmid Addgene #60032 was mutated at two BbsI sites to give the CRY2olig-mCherry-mut2-1. Kanamycin resistant. (k) Empty target control plasmid. Soluble CRY2 (CRY2HA-2A-mCherry) cassette cloned into the plasmid S13.1. This plasmid was also used to facilitate insertion of CRY2 into the multiplexed plasmids without soluble CRY2. Schematic shown in Figure 1c. As this plasmid lacks the gRNA expression module it is referred to as the “Empty target control” plasmid and was used in Figures 1, 3-5 and Supplementary Figure 4, 6-11 and 13. (l) LADL Bridge + Target plasmid. Multiplexed plasmid with soluble CRY2 and gRNAs 129, 135, 115, 117 all targeting desert regions (Supplementary Table 6). Schematic shown in Figure 1c and plasmid used in Figures 1-5 and Supplementary Figures 1, 5-11 and 13. (m) LADL Bridge + Promoter Only Target plasmid. Multiplexed plasmid with gRNAs 115 and 117 targeting the Zfp462 promoter region with soluble CRY2 (Supplementary Table 6). This plasmid was used in the “One-sided guide control” condition in Supplementary Figures 7-10.

Supplementary Figure 3

Light box validation

(a) Schematic of the CRY2olig-mCherry (Addgene #60032). (b) Circuit diagram of the light box (c) Experimental setup used to test CRY2 clustering using a customized light box. (d) Fluorescence imaging with a Texas Red filter to visualize CRY2HA-2A-mCherry. Scale bars, 10 µm. Images are representative of 2 independent experiments.

Supplementary Figure 4

Optimizing co-transfection efficiency of LADL constructs by testing different molar ratios and DNA masses

(a) The LADL Anchor and Empty target plasmids were co-transfected with 1.5 µg, 2.0 µg or 2.5 µg of total mass at a 1:1 molar ratio, and imaged using phase contrast and fluorescence imaging in the TXR channel at 24h post puromycin treatment. (b) The LADL Anchor and Empty target plasmids were co-transfected with 2.0 µg of total mass at 1:1, 2:1 and 4:1 molar ratio, and imaged as in (a). Scale bars, 500 µm. Images are representative of 2 independent experiments.

Supplementary Figure 5

Characterization of mouse embryonic stem (ES) cells after co-transfection with LADL constructs and exposure to blue light

Images of mouse ES cells transfected with either the Empty anchor control, the Empty bridge control or the LADL (Anchor + Bridge + Target) combinations (Supplementary Figure 1) and imaged using phase contrast and fluorescence imaging with Texas Red filter to visualize CRY2HA-2A-mCherry at 0, 24 and 36 hours post puromycin treatment. The Empty bridge samples, which should not have any mCherry signal, were used to set the acquisition parameters for the Texas Red channel. Scale bars, 500 µm. Images are representative of > 10 independent experiments.

Supplementary Figure 6

Chromosome-Conformation-Capture-Carbon-Copy (5C) global heatmaps at the larger Zfp462 and Klf4 gene loci in LADL-engineered mouse embryonic stem (ES) cells compared to relevant controls

Heatmaps represent 5C signals in relative interaction frequency for an 804 kb genomic region around the Klf4 and Zfp462 genes (chr4:54,799,136-55,603,136) from mouse ES cells co-transfected with LADL (Anchor + Bridge + Target) constructs after 24 hour blue light illumination at 5 mW/cm2 intensity (left column) and compared with two relevant controls: LADL–engineered cells in dark (middle column) and Empty target control in dark (right column). Zoom regions for engineered de novo loop (Box 1) and Klf4-stretch enhancer (SE) loop (Box 2) are shown in green boxes.

Supplementary Figure 7

5C heatmaps and boxplots of the engineered de novo loop after 24 hours of blue light illumination in independent experiments

(a-e) Zoomed-in 5C heatmaps represent engineered de novo loop between the targeted Zfp462 promoter (chr4:54,911,136-54,983,136; vertical) and pluripotency-specific Klf4 SE (chr4:55,451,136-55,523,136; horizontal) regions in LADL-engineered mouse ES cells after 24 hours of (a, c-e) 5 mW/cm2 or (b) 1.5 mW/cm2 intensity blue light illumination, or in dark, from independent replicate experiments. (a-e) The loop is represented in relative interaction frequencies (top rows) and distance-corrected interaction scores (bottom rows). (a, c-d) Empty target controls (Anchor + Bridge) in dark are included in replicate 1, replicates 3-4. (b) Empty anchor control (Bridge + Target) in replicate 2 and (c) one-sided guide control (Anchor + Bridge + Zfp462 promoter only Target) in replicate 3 were exposed to the blue light for 24 hours, or were in dark. See Supplementary Figure 1 for details of the plasmid combinations used. (f) Boxplots of relative interaction frequencies in the de novo engineered loop pixels (green box in a-e) from each independent experiment, respectively. Boxplots: central tendency = median, box minima = 25th percentile, box maxima = 75th percentile, notches = 95% confidence interval, whiskers = 1.5x interquartile range.

Supplementary Figure 8

Looping efficiency analysis highlights the engineered de novo loop from both gRNA anchors after blue light illumination.

(a) Zoomed-in 5C heatmaps representing the engineered de novo loop between the gRNA-targeted Zfp462 promoter (chr4:54,911,136-54,983,136; vertical) and pluripotency-specific Klf4 SE (chr4:55,451,136-55,523,136; horizontal). Zoom corresponds to Box 1 in the global heatmaps (see Figure 3a-b, Supplementary Figure 6). The brown box represents the 4C viewpoint at the Zfp462 promoter in (b-d), and Figure 3d. The blue box indicates the 4C viewpoint at Klf4 SE in (c, d left) and Figure 3e. The orange box represents the anchor point at the Klf4 SE used in (d right). See Supplementary Figure 1 for details of the plasmid combinations used. (b-c) Classic 4C looping efficiency plots. The line graph represents normalized interaction frequency of (b) the Zfp462 promoter anchor (brown vertical line) or (c) the SE anchor (brown vertical line) for five biological replicates in LADL+ blue light (blue line), LADL + dark (black line) and additional controls in each replicate as listed: Empty target control in dark (grey line) is included in replicates1, 3 and 4; Empty anchor control + dark (grey line) and + blue light (light blue line) are included in replicate 2; and one-sided guide control +dark (light gray line) and + blue light (light blue line) is included in replicate 3. The single pixel of interest (engineered de novo loop) at (b) the targeted site in Klf4 SE and (c) the Zfp462 promoter are shaded in yellow. (d) Plots show the percent looping efficiency of the SE interactions at the Zfp462 promoter for each condition in all five replicates (left) after 24 hours (intersection of brown and blue boxes in panel (a)) and (right) 4 hours (intersection of brown and orange boxes in panel (a)) of blue light illumination. Red bars indicate medians of each condition. P-values computed using the unpaired two-sided Mann Whitney U test.

Supplementary Figure 9

5C heatmaps and boxplots of the endogenous loop between the pluripotency-specific Klf4 stretch enhancer (SE) and the Klf4 gene after 24 hours of blue light illumination in independent experiments.

(a) Zoomed-in 5C heatmaps represent Klf4-SE loop in LADL-engineered cells after 24 hours of 5 mW/cm2 blue light illumination, or in dark. (b) Klf4-SE loop in one additional replicate using 1.5 mW/cm2 and (c-e) three additional replicates using 5 mW/cm2 blue light illumination for 24 hours. (a, c-d) Empty target controls (Anchor + Bridge) in dark are included in replicates 1, 3-4. (b) Empty anchor control (Bridge + Target) in replicate 2 and (c) one-sided guide control (Anchor + Bridge + Zfp462 promoter only Target) in replicate 3 were exposed to 1.5 mW/cm2 or 5 mW/cm2 blue light for 24 hours, respectively, or were in dark. See Supplementary Figure 1 for details of the plasmid combinations used. (a-e) Klf4-SE loop is represented in relative interaction frequencies (top rows) and distance-corrected interaction scores (a-b bottom rows). (f) Boxplots of relative interaction frequencies in the Klf4-SE loop pixels (green box in a-e) after 24 hours of blue light illumination from each independent experiment, respectively. Boxplots: central tendency = median, box minima = 25th percentile, box maxima = 75th percentile, notches = 95% confidence interval, whiskers = 1.5x interquartile range.

Supplementary Figure 10

Looping efficiency analysis highlights the effect on the endogenous Klf4-SE loop after engineering a loop with 24h blue light illumination

(a) Zoomed-in 5C heatmaps representing the endogenous loop between pluripotency-specific Klf4 SE and Klf4 gene (chr4:55,447,136-55,583,136). The blue box represents the 4C viewpoint at Klf4 promoter in (b-c) and Figure 3g. The green box indicates the 4C viewpoint at Klf4 SE in (b, c) and Figure 3g. See Supplementary Figure 1 for details of the plasmid combinations used. (b) Classic 4C looping efficiency plots. The line graphs represent normalized interaction frequency of the Klf4 promoter (brown vertical line) for five biological replicates in LADL+ blue light (blue line), LADL + dark (black line) and additional controls in each replicate as listed: Empty target control in dark (grey line) included in

replicates 1, 3 and 4; Empty anchor control + dark (grey line) and + blue light (light blue line) included in replicate 2; and one-sided guide control +dark (light gray line) and + blue light (light blue line) included in replicate 3. The single pixel of interest (endogenous Klf4-SE loop) at the SE is shaded in yellow. (c) Strip charts show the % looping efficiency of the (left) Klf4 promoter interactions at the SE and (right) SE interactions at Klf4 promoter for each condition in all five replicates after 24h of blue light illumination. Values derived from the pixel at the intersection of green and blue boxes as shown in panel (a). Red bars indicate medians of each condition.

Supplementary Figure 11

5C heatmaps of the looping interactions between the Zfp462 gene and its pluripotency-specific enhancers (E1-E4) after 24 hours illumination with 5 mW/cm2 intensity blue light.

(a-b) LADL-engineered mouse embryonic stem cells after 5 mW/cm2 blue light illumination (left column) after 24 hours were compared with LADL-engineered cells in dark (middle column) and Empty target control in dark (right column). Heatmaps represent the regions around Zfp462 gene and its four pluripotency-specific enhancers E1, E2, E3 and E4 in chr4:54,859,136-55,351,136. (a) Relative interaction frequency and (b) distance-corrected interaction score heatmaps.

Supplementary Figure 12

Schematic of LADL experimental design with 4 and 24 hours of blue light illumination Mouse embryonic stem cells were co-transfected with LADL (Anchor +Bridge + Target) plasmids or Empty target control (Anchor + Bridge only) plasmids. At the time of harvesting, the cells have undergone puromycin selection for at least 36 hours in either 4 or 24 hours of blue light exposure.

Supplementary Figure 13

RNA-FISH analysis of additional independent experiments

LADL-engineered mouse embryonic stem cells and three other controls (LADL+dark, Empty target control+dark, Empty bridge control+dark) were exposed to 5 mW/cm2 blue light for 24 hours in two independent replicate experiments. See Supplementary Figure 1 for details of the plasmid combinations used. (a-d) Strip charts representing (a, c) the number of total mRNA transcripts per cell and (b, d) the estimated level of nascent transcripts per allele for Zfp462 (upper row) and Klf4 (lower row) for two additional independent experiments. Red bars indicate means of each condition. (e, f) Histograms represent the proportion of cells with a specific number of actively expressing Zfp462 alleles in two additional biological replicates. (a, c, e, f) n = the number of cells, or (b, d) the number of active transcription alleles per condition. P-values computed using the unpaired one-tailed Mann Whitney U test.

Supplementary Tables

Supplementary Table 1. Primers used to clone LADL plasmids See attached .xls file.

Supplementary Table 2. Primers and matched plasmids for cloning LADL constructs.

Name of insert Primers PCR template Size (bp)

LADL Anchor Control plasmid

EF1a MRP177, MRP 188 Addgene #47457 1237

3XFLAG-dCas9 MRP189, MRP 190 dCas9 plasmid 4300

GS-CIBN MRP198, MRP 199 Addgene #47457 649

2A-Puro MRP200, MRP 182 Addgene #62987 681

Empty anchor Control plasmid

EF1a MRP036, JV002 Addgene #47457 1237

Puro JV001, MRP 051 Addgene #62987 681

Empty target Control plasmid

EF1a MRP177, MRP 178 Addgene #47457 1247

CRY2PHR MRP179, MRP 183 Addgene #47457 1700

2A-mCherry MRP184, MRP 185 CRY2olig mut 2-1 681

Supplementary Table 3. List of gRNA primer sequences

Primer number

gRNA Primer name gRNA sequence

Backbone plasmid

(single clone)

Multiplex

129 Klf4_Enh_1_F CACCGTACATGCAGTAGTACTAAGT S12.1 Desert gRNAs (Klf4/

Zfp462)

130 Klf4_Enh_1_R AAACACTTAGTACTACTGCATGTAC S12.1 135 Klf4_Enh_2_F CACCGTTTGTGTTTTAGTGTAGATT B1 136 Klf4_Enh_2_R AAACAATCTACACTAAAACACAAAC B1 115 Zfp462_Prom_2_F CACCGTAAAGAAAAGTGTTTATCGA B2 116 Zfp462_Prom_2_R AAACTCGATAAACACTTTTCTTTAC B2 117 Zfp462_Prom_1_F CACCGAAGTGTTTATCGAGGGAAAG B3 118 Zfp462_Prom_1_R AAACCTTTCCCTCGATAAACACTTC B3 115 Zfp462_Prom_2_F CACCGTAAAGAAAAGTGTTTATCGA S12.1 Promoter

only Target plasmid (Zfp462

only)

116 Zfp462_Prom_2_R AAACTCGATAAACACTTTTCTTTAC S12.1 117 Zfp462_Prom_1_F CACCGAAGTGTTTATCGAGGGAAAG B1 118 Zfp462_Prom_1_R AAACCTTTCCCTCGATAAACACTTC B1

Supplementary Table 4. List of plasmids with individual gRNAs without soluble CRY2 gRNA

in plasmid

gRNA Primer name

BbsI digested plasmid backbone

Targeting region

Genomic feature of

target

Refer to

129 Klf4_Enh_1_F S12.1 Engineered Loop Anchor 2

Desert near SE

Supplementary Figure 2e

135 Klf4_Enh_2_F B1 (Addgene # 58778)

Engineered Loop Anchor 2

Desert near SE

Supplementary Figure 2f

115 Zfp462_Prom_2_F

B2 (Addgene # 58779)

Engineered Loop Anchor 1

Desert near Zfp462 TSS

Supplementary Figure 2g

117 Zfp462_Prom_1_F

B3 (Addgene # 58780)

Engineered Loop Anchor 1

Desert near Zfp462 TSS

Supplementary Figure 2h

Supplementary Table 5. List of plasmids with multiplexed gRNAs without soluble CRY2

gRNAs in

plasmid

gRNA Primer name

BsaI digest plasmid

backbone

Targeting region

Multiplex Plasmid name

Refer to

129 Klf4_Enh_1_F S12.1 Desert near SE and

Desert near Zfp462 TSS

Empty bridge control plasmid

Supplementary Figure 2i 135 Klf4_Enh_2_F

115 Zfp462_Prom_2_F 117 Zfp462_Prom_1_F

Supplementary Table 6. List of plasmids with multiplexed gRNAs with soluble CRY2

gRNAs in

plasmid

gRNA Primer name

BsaI digest plasmid

backbone

Targeting region Multiplex Plasmid name

Refer to

129 Klf4_Enh_1_F S12.1 Desert near SE and Desert near Zfp462 TSS

LADL Bridge + Target

Supplementary Figure 2l 135 Klf4_Enh_2_F

115 Zfp462_Prom_2_F 117 Zfp462_Prom_1_F 115 Zfp462_Prom_2_F S12.1 Desert near Zfp462 TSS One-sided

guide Control Supplementary

Figure 2m 117 Zfp462_Prom_1_F

Supplementary Table 7. List of primers used for RNA qRT-PCR Name Sequence Nanog_F_MR137 TGCCTGCAGTTTTTCATCCC Nanog_R_MR138 TAGAAGAATCAGGGCTGCCTTG Mouse_GAPDH_F_MR141 GCACAGTCAAGGCCGAGAAT Mouse_GAPDH_R_MR142 GCCTTCTCCATGGTGGTGAA Sox2_F_MR143 GCACATGAACGGCTGGAGCAACG Sox2_R_MR144 TGCTGCGAGTAGGACATGCTGTAGG Klf4_F_MR147 AGACCAGATGCAGTCACAAGTC Klf4_R_MR148 TTTTGCCACAGCCTGCATAG Oct4_F_MR151 TGTGGACCTCAGGTTGGACT Oct4_R_MR152 TTTCATGTCCTGGGACTCCTC MRP213_qPCR_Zfp462_F GCCAACTGATGTTGCCGAGGACAATG MRP214_qPCR_Zfp462_R CCTGAAGTAGCGTACGCAGAACTTG MRP215_qPCR_Zhang_dCas9_F GCACAGCATCAAGAAGAACCTG MRP216_qPCR_Zhang_dCas9_R CGTTGCTGAAGATCTCTTGCAG MRP149_CRY2_qPCR_F AATGCCTCGACATGTCCATC MRP150_CRY2_qPCR_F AGCGCGTTACTGGGTTTTTC Nestin_Fwd_JB AGGCCACTGAAAAGTTCCAG Nestin_Rev_JB TAAGGGACATCTTGAGGTGTGC Supplementary Table 8: Summary of external sequencing libraries analyzed in this study See attached .xls file.

Supplementary Table 9. List of primers used for Chromatin Immunoprecipitation

Primer number ChIP Primer name Sequence Target region

MRP217 MRP217_IP_Down_Y

_ED_129to135_F TGGGCCTACTTAGTACTACTGC Engineered site at

SE

MRP218 MRP218_IP_Down_Y

_ED_129to135_R GCTGGGTAAGTAGCCCTCTAC Engineered site at

SE

MRP221 MRP221_IP_Down_H

_ED_115to117_F AAGCCCCTTTCCCTCGATAAAC Engineered site at Zfp462 promoter

MRP222 MRP222_IP_Down_H

_ED_115to117_R ACACTAGGAGGATGGGGATAGTC Engineered site at Zfp462 promoter

MRP223 MRP223_IP_H_CTCF

_149to155_F GCTCTATGTTCTAACACCTCTCC Negative control

MRP224 MRP224_IP_H_CTCF

_149to155_R CGTGCTTGTACACACACACAG Negative control

Supplementary Table 10. 5C Primer Sequences. See attached .xls file. Supplementary Table 11. 5C Primer Genomic Coordinates. See attached .xls file. Supplementary Table 12: Summary of mapped 5C sequencing reads See attached .xls file. Supplementary Table 13. Fluorescence-labeled oligonucleotide sequences for RNA FISH. See attached .xls file.