Fig. 3

Sir3-dependent hyperclustering of telomeres is the prominent feature of genome folding in long-lived quiescent SP cells. a Chromosome organization of G1 and quiescent cells (HD fraction of SP: G0). ii) Normalized genomic contact matrix obtained for G1 daughter (left) and quiescent (right) cells. The chromosome names are indicated on the top axis. The color scale on the right indicates the frequency of contacts between two regions of the genome (white = rare contacts, dark blue = frequent contacts). Red arrowheads indicate centromere clustering; green and yellow arrowheads point at telomere–telomere contacts between two chromosomes (XIII and XV) in G1 and G0 cells, respectively. The average 3D structures reconstructed from the two contact maps are depicted on the corresponding side (see also Additional files 2 and 4). Each chromosome is represented as a chain of beads (1 bead = 20 kb), and the color code reflects the chromosome arm length, from blue for short arms to red for long arms. Yellow beads = subtelomeric regions; black beads = centromeres; purple beads = boundaries of the rDNA cluster on chromosome 12. b Scaled up view of a region of the matrices corresponding to the contacts between chromosomes XV and XIII in the G0 and G1 stages. c Representation of the distances between all pairs of telomeres as observed in the 3D structures of G1 and quiescent cells. Both structures have been scaled to account for the measured difference in size between nuclei in G0 and G1 daughter cells (unit=10nm, see “Materials and methods”). The 32 telomeres are ordered according to the corresponding chromosome arm length, from the shortest (left) to the largest (right). WT wild type. d Analysis of the contact frequency between sub-telomeres in G1 and G0 quiescent cells. For 3-kb windows starting at the telomere (right) and moving toward the centromeres, the mean of contact from each window with the other subtelomeres is plotted. The blue and pink curves represent the contacts computed between 35-kb segments randomly sampled from the genome in both conditions, to illustrate the absence of coverage bias after normalization in the analysis. e Scaled up view of the contacts between chromosomes XV and XIII in the G0 stage in SIR3 defective (sir3∆, hml∆ to avoid the pseudo-diploid effect due to SIR3 deletion) or WT (hml∆) cells (see Additional file 3 for a genome-wide overview of the contacts in these experiments). f As in (d) for sir3∆ and WT G0 cells