Insertion of an array of G4s in an isolated locus.

(A) Schematic showing the insertion of the G4-array and the G4-mutated control at chr12:79,872,423-79,872,424 (hg19). (B) PCR of the insertion locus showing the successful insertion of the 275bp long insert sequence.

Changes in chromatin upon G4-array insertion.

The top panel shows the positions of the PCR amplicons used in the Histone ChIP experiments. Changes in chromatin-modifying histone modifications in the insert region represented by calculating the ratio of occupancy of different histone marks in the G4-array insert cells over the G4-mutated insert (control) cells-enhancer mark, H3K4Me1 (A); active enhancer/promoter mark, H3K27Ac (B); facultative repressor mark, H3K27Me3 (C) and constitutive repressor mark, H3K9Me3 (D). Mean ± SD (n=2/3); unpaired, two-tailed t-test.

Insertion of the G4-array led to enhancer function.

(A) Long-range G4-dependent changes in mRNA expression are represented by calculating the ratio of expression of surrounding genes in the G4-array insert cells over the G4-mutated insert (control) cells. Top panel shows the positions of the gene promoters with respect to the insertion site. (B) Ratio of Pol2 Phospho-Ser5 Occupancy at the promoters of the surrounding genes in the G4-array insert cells over the G4-mutated insert (control) cells. (C) A UCSC genome browser snapshot showing the 3C looping interactions between the insertion and the surrounding gene promoters. Fold change in 3C looping interactions between the insertion and the surrounding gene promoters in the G4-array insert cells over the G4-mutated insert (control) cells.

Top panel shows the positions of the PCR amplicons used in the ChIP experiments. The ratio of occupancy of p300 (D) and Ac p300/CBP (E) in the G4-array insert cells over the G4-mutated insert (control) cells. Mean ± SD (n=3); unpaired, two-tailed t-test.

LNA-mediated disruption of the inserted G4s reverses enhancer phenotype.

(A) Schematic showing the inserted G4 structures disrupted using LNA probes (details in methods). (B) Effects of LNA treatment in the G4-array insert cells on the expression of surrounding genes which showed enhanced expression when compared against the G4-mutated insert (control) cells in Figure 3A-represented by the ratio of expression of surrounding genes in the LNA-treated over the vehicle-treated (control) cells. Effects of LNA treatment in the G4-array insert cells on the levels of H3K4Me1 (C) and H3K27Ac (D) histone modifications at the insert locus represented by the ratio of occupancy of the histone marks in the LNA-treated over the vehicle-treated (control) cells. The top panel shows the positions of the PCR amplicons used in the Histone ChIP experiments. Mean ± SD (n=3); unpaired, two-tailed t-test.

Insertion of the G4-array increased Hi-C interactions.

Circos plots showing raw Hi-C contacts across the genome originating from a ±10 kb window with the insertion site at the middle across 3 samples-(A) G4-array insert cells, (B) G4-mutated insert (control) cells and (C) HEK293T control cells (taken from GSE44267). (D) Table showing the number of genome-wide raw Hi-C contacts and normalized contacts (normalized against the total raw Hi-C contacts to normalize for the sequencing depth) originating from the ±10 kb window with the insertion site at the middle across the 3 samples.

G4-dependent changes in local chromatin architecture.

Juicebox Hi-C matrices showing Hi-C contacts in the (A) G4-array insert cells, (B) G4-mutated insert (control) cells in a 3.6Mb region of chromosome 12 with the insertion site at the middle of the matrices. The arrows at the top of the Hi-C matrices indicate the site of insertion. (C) Juicebox Hi-C matrix showing normalized Hi-C contacts in the G4-array insert cells over the G4-mutated insert (control) cells as a heatmap. The region of interest (i.e., interactions associated with the immediate vicinity of the insert) is marked with a box. The arrow at the top of the Hi-C matrix indicates the site of insertion. (D) A line histogram displaying the differences in interaction frequency across G4-array insert cells and G4-mutated insert (control) cells in regions up to 100 kb away from the insertion site. As seen interactions downstream of the insertion site are more enriched than upstream in the G4-array insert cells as compared to the G4-mutated control.

(E) Circos plot showing differential interactions (fold enrichment >=2) originating from a ± 100 kb window with the insertion site at the middle, in the G4-array insert cells over the G4-mutated insert (control) cells. (F) A UCSC genome browser snapshot showing the more significant differential interactions (fold enrichment >=2, interaction reads >20) originating from a ± 50 kb window with the insertion site at the middle, in the G4-array insert cells over the G4-mutated insert (control) cells. The color intensity of the arcs indicating the interacting bins is proportional to the fold enrichment.

An isolated locus was chosen for insertion.

(A) 3D Genome Browser (Wang et al., 2018) snapshots showing the Hi-C interaction matrices from 4 cell lines (NHEK, IMR90, HUVEC and HMEC) of a section of chromosome 12 with chr12:79,870,000-79,875,000 (hg19) locus in the middle (marked by arrows); the chr12:79,870,000-79,875,000 locus has very low interaction with its surrounding loci, indicated by the lack of red dots or very faint red dots.

Insert sequences.

Sequences of the G4 array and the G4-mutated array that were inserted; three or more runs of G/Cs (stem of a potential G4) are shown in bold and the G/Cs (marked in blue) that were changed to T/As (marked in red) in the G4-mutated array are marked.

Insertion confirmed by Sanger sequencing.

Representative snapshots of the Sanger sequencing chromatogram of the reverse strand of the insertion locus PCR product showing some of the specific G/C to T/A substitutions in the insert.

The chromosomal architecture of the insertion locus in the G4-array insert cells is broadly similar to uninserted cells except for the increase in looping interactions.

Comparison of Hi-C contact matrices around the insertion site in the G4-array insert cells (A) and the HEK293T control cells (taken from GSE44267) (B) shows that the broad chromatin organization is conserved. The TADs appear to be otherwise unaltered upon insertion.

Comparative analysis of the G4-dependent increase in Hi-C interactions upon insertion.

Table showing the number of genome-wide raw Hi-C contacts, actual number of contacts originating from the ±10 kb window with the insertion site at the middle and the comparative analysis (mean, standard deviation and z-score) of these number of contacts with contacts across 10,000 random 20kb windows across the genome across the 3 samples.