Extended Data Fig. 5: Strategy to relocate R4-5 and R2-3 to an iLAD environment or close to LAD4 and screening, validation and additional analysis of R4-5 and R2-3 relocation clones.
From: Interactions between the genome and the nuclear lamina are multivalent and cooperative
(a) Simulated NL interaction tracks to illustrate two-step relocation of R4-5. First, upon recombination i8, R4-5 is flanked on one side by iLAD sequences. Second, to totally isolate R4-5 in an iLAD environment, the remaining LAD1 sequences and part of the flanking iLAD were deleted by two Cas9-mediated cuts in the CAST allele of recombined clones: at the R4-5-LAD1 junction and in the middle of the following iLAD. To isolate R4-5 close to LAD4, the second cut was induced closer to LAD4. (b) Simulated NL interaction tracks to illustrate two-step relocation of R2-3. Upon recombination i9, R2-3 is flanked on one side by iLAD sequences. Next, to isolate R2-3 in an iLAD or close to LAD4, we used the same strategy as in (a). (c) Screening of clones from recombination i8 and i9 by PCR. Clones were selected if rec1 and rec2 products only were amplified by PCR2 and PCR3 respectively. For PCR design, see Extended Data Fig. 2A. PCR1 was not informative because these clones carry an extra copy of SB. (d) Bar plots showing the total CAST and 129S1 allele-specific sequencing read counts per 25 kb from pA-DamID experiments, at the Cas9-mediated deletion sites and flanking regions, in clones in which R4-5 (top) or R2-3 (bottom) were isolated in an iLAD environment. Read counts are the sum of Dam-LaminB1 and Dam-only reads; note that this is different from the pA-DamID tracks in Fig. 2, which show the (z-score transformed) log2 ratio of Dam-LaminB1 and Dam-only. Vertical dashed lines indicate the expected deletions on the CAST allele. (e) Domainogram (top) and pA-DamID tracks (middle) of 129S1 allele of cells with hopped but not recombined loxP (“Control”, blue) and rearranged cells (“Experimental”, red) in which R4-5 (top) and R2-3 (bottom) regions have been isolated within an iLAD. pA-DamID tracks and domainograms are plotted using WT chromosomal coordinates. (f) Left panel: R4-5 contacts the NL as frequently as its neighbouring LAD. pA-DamID score (z-score) for NL interactions of iLAD, R4-5 or LAD GATC fragments from the window plotted in Fig. 3d, left. Middle panel: R2-3 is not as strongly associated with the NL as its neighbouring LAD. Same but for iLAD, R2-3 or LAD GATC fragments from the window plotted in Fig. 3d, right. Right panel: R4-5 has a higher NL contact frequency than R2-3. For these analyses, R4-5 and R2-3 GATC fragments more than 2.5 kb apart were normalised over the 5th percentile of genome-wide pA-DamID values and their normalised pA-DAmID score distributions plotted. Because the resolution of DamID is estimated to be about 1 kb, we conducted this analysis on a subset of GATC fragments that are at least 2.5 kb apart, thus ensuring independent measurements. Black lines show median values. Statistical significance was assessed using one-sided Wilcoxon tests. (g) Bar plots showing the total CAST and 129S1 allele-specific sequencing read counts per 25 kb from pA-DamID experiments, at the Cas9-mediated deletion sites and flanking regions, in clones in which R4-5 (top) or R2-3 (bottom) were isolated close to LAD4. Read counts are the sum of Dam-LaminB1 and Dam-only reads; note that this is different from the pA-DamID tracks in Fig. 2, which show the (z-score transformed) log2 ratio of Dam-LaminB1 and Dam-only. Vertical dashed lines indicate the expected deletions on the CAST allele. (h) Domainogram (top) and pA-DamID tracks (middle) of 129S1 allele of cells with hopped but not recombined loxP (“Control”, blue) and rearranged cells (“Experimental”, red) in which R4-5 (top) and R2-3 (bottom) regions have been isolated close to LAD4. pA-DamID tracks and domainograms are plotted using WT chromosomal coordinates. Domainogram colour-key is displayed in (e).
