Figures and data
Expression of IL1R1 is regulated by telomere length-dependent enrichment of non-telomeric TRF2 on the gene promoter
A. Occupancy of TRF2 was checked by ChIP-qPCR on the IL1R1 promoter spanning +200 to -1000 bp of TSS was reduced ex-vivo HT1080 cells with long telomeres (HL1080-LT) relative to ones with short telomeres (HT1080) cells (N=3 in each case) ;IL1R1-3’UTR or a region 20 kb upstream were used as negative controls for TRF2 binding. Statistical significance was calculated using unpaired T test with Welch’s correction (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
B. mRNA expression of IL1R1 in ex-vivo HT1080 or HT1080-LT cells; GAPDH was used for normalization [N=3]. Statistical significance was calculated using unpaired T test with Welch’s correction (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
C. IL1R1 protein expression was checked by immuno-flow cytometry in HT1080 or HT1080-LT cells in three independent replicates; Mean IL1R1 expression has been plotted along the X-axis in log scale (right panel) [N=3]. Statistical significance was calculated using unpaired T test with Welch’s correction (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
D. IL1R1 levels by immunofluorescence microscopy in HT1080 or HT1080-LT cells; cells were stained with DAPI for marking the cell nucleus; IL1R1 levels from 25 individual cells shown in graph (right panel). [N=25] Statistical significance was calculated using Mann-Whitney’s non-parametric test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
E. Telomere length of the xenograft tumors made in NOD SCID mice using HT1080 cells with long telomeres (HL1080-LT) and short telomeres (HT1080) was checked by flow-cytometry and was higher in HT1080-LT xenograft tumors. Following this, IL1R1 expression was checked by flow cytometry for these samples. Mean Telomeric signal and Mean IL1R1 expression has been plotted respectively ( panels below) [N=5] Statistical significance was calculated using Mann-Whitney’s non-parametric test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
F. Occupancy of TRF2 by ChIP-qPCR at the IL1R1 promoter spanning +200 to -1000 bp of TSS was reduced in xenograft tumors made from HT1080 cells with long telomeres (HL1080-LT) relative to ones with short telomeres (HT1080) cells (N=3 in each case; IL1R1-3’UTR or a region 20 kb upstream were used as negative controls for TRF2 binding). [N=3] Statistical significance was calculated using unpaired T test with Welch’s correction (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
G. mRNA expression of IL1R1 in xenograft tumors (HT1080 or HT1080-LT cells); 18S was used for normalization. [N=5] Statistical significance was calculated using Mann-Whitney’s non-parametric test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
H. Telomere length by flow-cytometry of telomeric signal in HT1080 cells (hTERT-inducible stable line) following 0, 1, 6, 12 or 20 days of hTERT induction; HT1080-LT shown as positive control for enhanced telomere length; mean telomeric signal (FITC) plotted bar graph (left). [N=3] Statistical significance was calculated using unpaired T test with Welch’s correction (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
I. TRF2 occupancy ChIP-qPCR on the IL1R1 promoter spanning +200 to -1000 bp of TSS in HT1080 cells following 0, 1, 6 or 12 days of hTERT induction in ex-vivo culture; primers and normalization as in Figure 2A; IL1R1-3’UTR or a region 20 kb upstream used as negative controls. [N=3] Statistical significance was calculated using unpaired T test with Welch’s correction (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
J. IL1R1 mRNA expression (normalized to GAPDH) in HT1080 cells following 0, 1, 6 or 12 days of hTERT induction in ex-vivo culture. [N=3] Statistical significance was calculated using unpaired T test with Welch’s correction (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
K. TRF2 occupancy by ChIP-qPCR on the IL1R1 promoter in MDAMB23 or MDAMD231-LT cells using primers and normalization described earlier. [N=3] Statistical significance was calculated using unpaired T test with Welch’s correction (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
L. Expression for IL1R1 in MDAMB231 cells or MDAMD231-LT cells; GAPDH was used for normalization. [N=3] Statistical significance was calculated using unpaired T test with Welch’s correction (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
Error bars correspond to standard error of mean from independent experiments.
TRF2 is a transcriptional activator of IL1R1
A. IL1R1 expression by qRT PCR following expression of TRF2-flag or TRF2 silencing ( 48 hrs. of transient transfection) ; GAPDH was used for normalization. [N=3] Statistical significance was calculated using unpaired T test with Welch’s correction (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
B-C. TRF2 ChIP-qPCR spanning +200 to -1000 bp of IL1R1 TSS for TRF2 occupancy (B) and promoter activity (luciferase reporter including -1500 bp TSS; C) following expression of flag-tag TRF2 full-length, or the mutants TRF2-delB or TRF2-delM; anti-flag antibody used for ChIP; IL1R1-3’UTR or a region 20 kb upstream were used as negative controls. [N=3] Statistical significance was calculated using Tukey’s multiple comparisons test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
D. TRF2 and IL1R1 levels by western blots following TRF2 induction with 2 or 4 µg/ml doxycycline (Dox) in HT1080 cells (stably transformed for Dox-inducible TRF2); GAPDH as loading.
E. TRF2 and IL1R1 expression in control or TRF2-induced HT1080 cells by flow-cytometry; TRF2/IL1R1 in x-axis in log scale for 20 thousand cells.
F. IL1R1 levels in control or TRF2-induced HT1080 cells. CD44 used for cell-surface marker and nuclei were stained with DAPI; 25 cells in each condition were scored and plotted in the summary graph. [N=25] Statistical significance was calculated using Mann-Whitney’s non-parametric test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
G. TRF2 (left panel), IL1R1 (right panel) levels from xenograft tumors in NOD-SCID mice (control or doxycycline-induced TRF2 in HT1080 cells; N=6 mice in each group) by immuno-flow cytometry; mean fluorescence signal from individual tumors in control or TRF2-induced tumors plotted in adjacent graphs. [N=6] Statistical significance was calculated using Mann-Whitney’s non-parametric test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
Error bars correspond to standard error of mean from independent experiments.
DNA secondary structure G-quadruplexes are important for TRF2 binding to IL1R1 promoter
A. Schematic depicting G4 motifs (A and B) and their respective positions on the IL1R1 promoter; sequence scheme depicting generic G4 motifs;
B. Circular dichroism profile (220-310 nm wavelength) of oligonucleotides (5 µM) with sequence of the G4 motifs A or B (or respective mutants with base substitutions) in solution.
C. ChIP-qPCR following BG4 ChIP at the IL1R1 promoter (fold enrichment over mock) overlaid with TRF2 occupancy (fold enrichment over IgG) in HT1080 cells. [N=3] Statistical significance was calculated using Šídák’s multiple comparisons test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
D. IL1R1 promoter activity in HT1080 cells from luciferase-reporter without or with G4-deforming substitutions against G4 motif A (IL1R1G4-MUT-A) or G4 motif B (IL1R1-G4 MUT-B) in presence or absence of TRF2 induction; G4 motifs A and B as shown in schematic; specific base substitutions have been indicated in blue font. [N=3] Statistical significance was calculated using Tukey’s multiple comparisons test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
E. IL1R1 promoter activity, mRNA and protein levels in presence/absence of the G4 binding ligand 360A with or without TRF2 induction. For Left and centre panels-[N=3] Statistical significance was calculated using Tukey’s multiple comparisons test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
F. IL1R1 promoter-firefly luciferase reporter cassette, with or without substitutions deforming the G4 motif B, was artificially inserted at the CCR5 locus using CRISPR-cas9 gene editing in HEK293T cells (scheme in left panel). Promoter activity from (right panel). [N=3] Statistical significance was calculated using Tukey’s multiple comparisons test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
G. TRF2 occupancy by ChIP-qPCR (right panel) at the artificially inserted IL1R1 promoter without or with G4-deforming substitutions (IL1R1G4 MUT-B);ChIP-qPCR primers designed specific to the inserted promoter using the homology arms( see methods). [N=3] Statistical significance was calculated using unpaired T test with Welch’s correction (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
Error bars correspond to standard error of mean from independent experiments.
TRF2 recruits the histone acetyl transferase p300 to the IL1R1 promoter
A.H3K27ac occupancy at the IL1R1 promoter spanning +200 to -1000 bp of TSS by ChIP-qPCR in HT1080 (left panel) and MDAMB231 cells (right panel) in control (uninduced), TRF2-induced (up) or TRF2-down conditions; IL1R1-3’UTR or a region 20 kb upstream were used as negative controls. [N=3] Statistical significance was calculated using Tukey’s multiple comparisons test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
A. p300 occupancy on the IL1R1 promoter in HT1080 (left panel) and MDAMB231 cells (right panel) in control (uninduced), TRF2-induced or TRF2-down conditions; IL1R1-3’UTR or a region 20 kb upstream were used as negative controls. [N=3] Statistical significance was calculated using Tukey’s multiple comparisons test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
B. CBP (left panel) and ac-p300/CBP (right panel) occupancy on the IL1R1 promoter in HT1080 cells in control (uninduced) or TRF2-induced conditions; IL1R1-3’UTR or a region 20 kb upstream were used as negative controls. [N=3] Statistical significance was calculated using Šídák’s multiple comparisons test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
C. p300, CBP, acp300/CBP and H3K27Ac occupancy on the IL1R1 promoter in HT1080 and HT1080-LT cells; IL1R1-3’UTR or a region 20 kb upstream were used as negative controls. [N=3] Statistical significance was calculated using Šídák’s multiple comparisons test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
All protein ChIP, other than histones, were normalized to 1% input and fold-change have been calculated over respective IgG. Histone ChIP were normalized to 1% Input and fold change over total H3 have been calculated for individual samples (Further details in the methods section)
Error bars correspond to standard error of mean from independent experiments.
Acetylation of TRF2 at Lysine residue 293 is necessary for p300 recruitment on the IL1R1 promoter
A. Schematic showing positions ( amino acid residues) where acetylation and deacetylation of TRF2 have been reported.
B. IL1R1 mRNA expression post 48 hrs. of transient over-expression of various TRF2 acetylation mutants. GAPDH was used for normalization. [N=3] Statistical significance was calculated using Dunnett’s multiple comparisons test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
C. IL1R1 protein level following post 48 hrs. of transient over-expression of various TRF2 acetylation mutants; GAPDH used as loading control.
D. Immunofluorescence for IL1R1 in HT1080 cells without (control) or with induction of flag-tag TRF2 WT or mutant TRF2-293R; quantification from 29 cells in each case shown in graph. [N=29] Statistical significance was calculated using Mann-Whitney’s non-parametric test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
E. Occupancy of flag-tag TRF2-WT or TRF2-293R on the IL1R1 promoter by ChIP-qPCR in HT1080 cells (left panel); IL1R1-3’UTR or a region 20 kb upstream were used as negative controls. Occupancy by ChIP-qPCR of p300 (middle panel) and acp300/CBP (right panel) without (control) or with induction of TRF2-WT and TRF2-293R. [N=3] Statistical significance was calculated using Tukey’s multiple comparisons test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
F. TRF2-wild type (WT) and TRF2-2293R mutant proteins fused with dCAS9 expressed and targeted to the IL1R1 promoter using IL1R1-specific gRNA in HT1080 and MDAMB231 cells. Following this, expression of IL1R1 or other TRF2 target genes (non-specific with respect to the IL1R1-gRNA) in HT1080 (left) or MDMB231 cells (right). [N=3] Statistical significance was calculated using Tukey’s multiple comparisons test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
Error bars correspond to standard error of mean from independent experiments.
TRF2-dependent IL1 pathway activates NFkappa B (p65) in cancer cells
A-B. NFKappaB activation in presence of IL1B (10 ng/ml) in HT1080 (A) or MDAMB231 (B) cells with or without TRF2 induction; activation signaling was confirmed through NFKappaB-Ser536 phosphorylation (normalized to total NFKappaB); ratio of Ser566-p/total NFKappaB plotted for respective blots from three independent replicates (right panels). [N=2]
C. Expression of NFKappaB targets IL6, IL8 or TNF in presence/absence IL1A, IL1B or TNF-α (10 ng/ml) for 24 hr in control (scrambled siRNA) or TRF2-low (TRF2 siRNA) conditions in HT1080 cells. [N=3] Statistical significance was calculated using Tukey’s multiple comparisons test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
D. Expression of IL6, IL8 or TNF in control or TRF2-induced conditions on treatment with either IL1A or IL1B (10 ng/ml) for 24 hr in absence (left panel) or presence (right panel) of the IL1-receptor-antagonist IL1RA (20 ng/ml) in HT1080 cells. [N=3] Statistical significance was calculated using Tukey’s multiple comparisons test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
E. NFkappaB/phosphor-Ser536-NFKappaB levels in xenograft tumors developed in NOD-SCID mice (control or doxycycline-induced TRF2 in HT1080 cells; N=6 mice in each group) by immuno-flow cytometry; mean fluorescence signal from individual tumors in control or TRF2-induced tumors plotted in adjacent graph; activation shown as ratio of pSer536-p-NFkappaB over total NFkappaB; significance was calculated using Wilcoxon’s non-parametric test. [N=6] Statistical significance was calculated using Mann-Whitney’s non-parametric test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
Error bars correspond to standard error of mean from independent experiments.
IL1R1 expression in Triple Negative Breast cancer (TNBC) samples is sensitive to inter-tumoral variation in telomere length
A. Relative telomere length of Triple Negative Breast Cancer (TNBC) samples (94 patients) by Telo-qPCR as reported earlier; signal from telomere-specific primers was normalized to single copy number gene 36B4 for individual samples. All samples were run with HT1080 DNA as control and telomeric signal from HT1080 cells (telomere length ∼4.5 Kb) was used as reference for relative measurement; median telomere length (3.6 Kb) shown by red bar; samples >50% or <50% of median (dotted lines) were designated as long or short telomere samples respectively. [ N=94]
B. Flow cytometry analysis of telomere length of TNBC tissues from telomeric signal using telomere–specific FITC-labeled PNA probe; quantification of mean telomeric signal for nine TNBC-ST (short telomere, top left panel) and nine TNBC-LT (long telomere bottom left panel) shown in top right panel. Median telomere length has been indicated by dotted lines ( top right). TNBC tissue slides hybridized with telomere-specific PNA probes and counter stained with DAPI. Representative images for long/short telomere TNBC tissue shown ( bottom right) [N=9] Statistical significance was calculated using Mann-Whitney’s non-parametric test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
C. Telomere length was determined using the previously published algorithm Tel-Seq from sequenced genomes of TNBC samples (N=8 for TNBC-ST (short telomere) and N=6 for TNBC-LT (long telomere); and adjacent normal tissue from same patient). Samples identified as long-telomere (TNBC-LT) or short-telomere (TNBC-ST) using telo-qPCR were significantly different in TL, and consistent with telo-qPCR annotation (left); TL of tumor samples was lower that adjacent normal tissue [N=15] (center); and TL from adjacent normal tissues in LT or ST samples did not vary significantly [N= 8; N=6](right) Statistical significance was calculated using Mann-Whitney’s non-parametric test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
D. Paired telomere length assessment in tumor and adjacent normal tissue for 66 patients. Telomere signal was normalized to 36B4 (single copy gene ) by qRT PCR and compared between tumor and adjacent normal samples. The data was plotted pairwise (left) with lines connecting the individual tumor sample to the respective adjacent normal. All samples where tumor telomere length was lower than normal were connected by red lines and the samples where tumor telomeres where higher were connected by blue lines. Relative telomere length(Tumor/Normal) was plotted (middle) with samples with T/N<1 in red and T/N >1 in blue. Out of the 66 samples, 45 samples had T/N < 1, suggesting that about 2/3 rd of the samples tested had shorter telomeres in the tumors. Analysis of whole genome sequences of 14 pairs of samples using tel-seq pipeline revealed that 11 cases showed lower tumor telomere length compared to their adjacent normal counterparts (right). Left panel-Statistical significance was computed by Wilcoxon matched-pairs signed rank test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
E. TNBC tissue (N=6 each for long or short telomere samples) stained with EpCAM (far red) and hybridized with telomere-specific PNA probe (FITC). Mean telomeric signal was plotted in total EpCAM+ve or EpCAM−ve cells. [N=6] Statistical significance was calculated using Mann-Whitney’s non-parametric test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
F. mRNA expression of IL1R1 and other key cytokines: IL1A, IL1B, IL2, IL6, IL8, TNF, IL10, IL4 and IL13 in long or short telomere TNBC tissue,; the 18S gene was used for normalization. [N=6] Statistical significance was calculated using Mann-Whitney’s non-parametric test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
G. Relative telomere length of triple negative breast tumor organoids (TNBO) derived from TNBC samples: five TNBO-ST (short telomere) and five TNBO-LT (long telomere) was estimated by telo-qRTPCR. [N=5] Statistical significance was calculated using Mann-Whitney’s non-parametric test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
H. mRNA expression of IL1R1 and other key cytokines in triple negative breast tumor organoids (TNBO) derived from TNBC samples-TNBO-ST and TNBO-LT; the 18S gene was used for normalization. [N=5] Statistical significance was calculated using Mann-Whitney’s non-parametric test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
I. Secreted IL1B (pg/ml) from organoids TNBO-ST or TNBO-LT by ELISA using media supernatant form organoid culture (see methods).[N=6] Statistical significance was calculated using Mann-Whitney’s non-parametric test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
J. mRNA expression of IL1R1 and other key cytokines using xenograft tumors (made from either short (HT1080) or long telomere (HT1080-LT) fibrosarcoma cells, and ex-vivo HT1080 or HT1080-LT cells. The 18S gene was used for normalization for the tumors ;for ex-vivo cells GAPDH was used for normalization. Top panel-[N=5] Statistical significance was calculated using Mann-Whitney’s non-parametric test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001). ; bottom panel-[N=3] Statistical significance was calculated using unpaired T test with Welch’s correction (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
K. Heat map summarizing the mRNA expression of key cytokines across the models shown above: TNBC tissue, TNBC derived organoids (TNBO), xenograft tumors or ex-vivo cells in long or short telomere cases. Fold change of gene expression (long with respect to short telomeres) color coded as per the reference legend along y-axis.
Error bars correspond to standard error of mean from independent experiments.
Telomere length sensitive IL1R1 expression modulates Tumor associated macrophage (TAM) infiltration in TNBC
A. Key Immune marker expression ( TAM markers, TAM related chemokines, T cell markers ), TERT and TERC were plotted for fold change of transcript level expression ( IL1R1 high / IL1R1 low) and adjusted p-values in breast cancer (BRCA) samples categorized as IL1R1 high or low from TCGA ( See supplementary file 4 ).
B. Percentage tumor-associated-macrophage (TAM) infiltration in TNBC tissue with short (TNBC-ST) or long (TNBC-LT) telomeres using markers CD11b and CD206 (CD11b+CD206+ cells shown in top right quadrant); quantification of TAM infiltration from individual TNBC-ST or TNBC-LT samples have been plotted. [N=9] Statistical significance was calculated using Mann-Whitney’s non-parametric test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
C. TNBC tissue slides stained with TAM-specific marker CD206 for macrophage infiltration within tissue (counterstained with DAPI); representative images for two independent long or short telomere TNBC tissues shown.
D. Labelled (red) M2 macrophages derived from THP1 cells incubated with triple negative breast tissue-derived organoids (TNBO) for 12 hours (scheme in left panel). Infiltration of M2 macrophages in organoids shown as percentage of labelled (red) M2 macrophages in individual flow-cytometry plots (florescence signal (FL3) in x-axis in log scale); percentage infiltration from five TNBO-ST or five TNBO-LT samples shown in right panel. [N=5] Statistical significance was calculated using Mann-Whitney’s non-parametric test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
E. Labelled (red) M2 macrophage infiltration in presence/absence of receptor antagonist IL1RA (20 ng/ml) for tumor organoids with short (TNBO-ST) or long telomeres (TNBO-LT) in three replicates; percentage values for M2 infiltration plotted in right panel. Red florescence (FL3; y-axis in log scale) and respective percentage M2 infiltration values marked on top of individual flow cytometry plots. [N=3] Statistical significance was calculated using unpaired T test with Welch’s correction (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
F. mRNA expression of IL1R1 was checked in presence of varying concentrations of G-quadruplex binding ligands ( 48 hr treatment) using TNBO. 18S gene was used for normalization. Following this, the ligand JD83 was used to check M2 infiltration in TNBO as in (C-D). The M2 infiltration has been plotted for control and JD 83 treated samples. Top panel-[N=3] Statistical significance was calculated using unpaired T test with Welch’s correction (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).; bottom panel-[N=5] Statistical significance was calculated using Mann-Whitney’s non-parametric test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
Error bars correspond to standard error of mean from independent experiments.
Scheme depicting relatively low infiltration of TAM in tumors with relatively long telomeres vis-à-vis tumors with short telomeres. Reduced non-telomeric TRF2 binding at the IL1R1 promoter in tumors with long telomeres, and consequent low IL1R1 activation, attenuated p65-mediated IL1-beta and macrophage infiltration. Figure created with BioRender.com.
Graphical representation of the Telomere Sequestration Partitioning (TSP) model. The binding of TRF2 outside of telomeric regions (non-telomeric) is dependent on telomere length. When telomeres shorten, the presence of TRF2 at telomeres diminishes, resulting in increased TRF2 binding at non-telomeric sites. This shift in TRF2 distribution triggers epigenetic modifications at promoters showing telomere-dependent gene regulation. Described in Mukherjee et al. PLoS Genetics, 2018.
A. Telomer length (TL) by flow-cytometry analysis of telomeric signal (FITC labeled PNA probe) showing increase in TL in HT1080-LT cells relative to HT1080 cells; plotted along x-axis in log scale.
B. hTERT and TERC expression as expected was enhanced in HT1080-LT relative to HT1080 cells (normalized to GAPDH). [N=3] Statistical significance was calculated using unpaired T test with Welch’s correction (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
C. Immuno-histochemical (IHC) staining of previously reported and well-characterized cells with short/long telomeres (fibrosarcoma HT1080 or the long-telomere version HT1080-LT cells in cell-isogenic background) hybridized with telomere-specific FITC-labelled PNA probe, counterstained with DAPI ( left panel); telomeric signal was enhanced in HT1080-LT cells (quantified in right panel). [N=15] Statistical significance was calculated using Mann-Whitney’s non-parametric test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
D. Volcano plot of expressed genes in RNA-seq in HT1080 or HT1080-LT cells; x-axis has log-2 fold change for genes and the y-axis represents adjusted p values in log-10 scale. Key inflammation-related genes that have been marked in figure – most differentially expressed genes like IL1R1,IL1B and IL10 were found to have higher expression in HT1080 cells compared to HT1080-LT cells. TERT which ( as expected) is significantly lower in HT1080 cells compared to HT1080-LT cells have been also marked. Upregulated and downregulated genes listed in supplementary file 1.
E. Gene Ontology (KEGG) for upregulated genes in HT1080 cells over HT1080-LT cells.
F. Gene Ontology (KEGG) for down-regulated genes in HT1080 cells over HT1080-LT cells.
G. Occupancy of TRF2 by ChIP-qPCR checked in HT1080 cells for cytokine-related genes. Enrichment was plotted relative to mock IgG (post normalization to 1% input ) [N=3] Statistical significance was calculated using unpaired T test with Welch’s correction (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
H. TRF2 occupancy on the IL1R1 promoter in HT1080, MDAMB231, HEK293T or MRC5 cells by chromatin-immunoprecipitation (ChIP) followed by qPCR. Primers spanning +200 to -1000 bp of TSS used for TRF2 enrichment on the gene promoter; 3’UTR and -20 Kb upstream of TSS used as negative control for TRF2 binding; fold-change of occupancy was calculated over mock IgG after normalizing signal to 1% input. [N=3] Statistical significance was calculated using unpaired T test with Welch’s correction (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
I. IL1R1 protein levels in HT1080 and HT1080-LT cells by western blot ; GAPDH was used as loading control.
J. mRNA expression of hTERT and TERC (normalized to 18S) in xenograft tumors in NOD-SCID mice injected with HT1080 or HT1080-LT cells. 18S has been used as a normalizing control. [N=5] Statistical significance was calculated using Mann-Whitney’s non-parametric test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
K. hTERT mRNA expression in hTERT-inducible HT1080 cells on treatment with different concentrations of doxycycline [N=2] (left panel) and at different days following induction [N=3] (right panel). GAPDH was used for normalization. Statistical significance was calculated using unpaired T test with Welch’s correction (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
L. Relative telomere length by qRT-PCR in MDAMB231 or MDAMD231-LT (long telomere) cells in three independent experiments. [N=3] Statistical significance was calculated using unpaired T test with Welch’s correction (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
Error bars correspond to standard error of mean from independent experiments.
A. IL1R1 protein levels on TRF2-flag induction (left) or TRF2 silencing (right) in HT1080 cells following 48 hrs. of transfection.
B. IL1R1 mRNA expression in MDAMB231 cells following expression of TRF2 cDNA or TRF2 silencing for 48 hrs; GAPDH used as normalizing control (left). [N=3] Statistical significance was calculated using unpaired T test with Welch’s correction (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001). IL1R1 protein levels in MDAMB231 cells following TRF2 induction for 48 hrs by western blot (centre) and immuno-flow cytometry (right).
C. IL1R1 mRNA expression in HEK293T and MRC5 cells following transfection of TRF2-cDNA for 48 hrs; GAPDH was used as normalizing control. [N=3] Statistical significance was calculated using unpaired T test with Welch’s correction (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
Error bars correspond to standard error of mean from independent experiments.
A.TRF2 ChIP-seq peak with sequence of the G4 motifs (A and B) and their respective positions on the IL1R1 promoter
B. Occupancy of TRF2 by ChIP-qPCR at the IL1R1 promoter at the region of 0-200 bp of TSS in presence of ligand 360A in three independent experiments in HT1080 cells. The dosage of 360A used has been previous standardized (Mukherjee et al, 2019) [N=3] Statistical significance was calculated using Tukey’s multiple comparisons test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
C-D. Base substitutions for the G4 motif B in the IL1R1 promoter (as shown in Figure 3D) were introduced using CRSIPR in HEK293T cells. TRF2 ChIP-qPCR spanning the IL1R1 promoter (left) and IL1R1 mRNA expression (right) in HEK293T cells with or without the G4-disrupting substitution in two replicates. [N=2]
Error bars correspond to standard error of mean from independent experiments.
A. Enrichment of histone marks H3K27ac, H3K4me3, H3K27me3, H3K9me3 (normalized to total H3) by ChIP-qPCR on the IL1R1 promoter in HT1080 cells in TRF2-up (induced) condition or un-induced corresponding control cells. [N=3] Statistical significance was calculated using Šídák’s multiple comparisons test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
B. Transcription factors/histone remodelers on the IL1R1 proximal promoter (-750 bp) across seven cell lines curated from ENCODE was plotted using UCSC genome browser hg19 human genome assembly. P300 enrichment marked in red box.
C. Immunoprecipitation using TRF2 antibody probed for p300, acp300/CBP, TRF2 or TRF1 (positive control) in input, TRF2 IP fraction or mock IP fraction in HT1080 cells.
D. Histone acetyl transferase (HAT) assay using purified histone H3/H4 and full-length p300 as substrate in presence/absence of recombinant TRF2; BSA (bovine serum albumin) was used as a non-specific mock protein. [N=3] Statistical significance was calculated using Tukey’s multiple comparisons test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
Error bars correspond to standard error of mean from independent experiments.
A. TRF2-WT Flag or TRF2 K293R-Flag was stably over expressed in HT1080 cells and immunoprecipitation was performed using Anti-Flag antibody or Mock IgG. Enrichment for p300 in IP faction was tested for TRF2-WT (left) and TRF2 K293R(right)
B. Previously reported TRF2 target genes (Mukherjee et al, 2018) were checked following transient over-expression of TRF2 WT and TRF2 K293R in HT1080 cells. While K293R did not affect TRF2 mediated repression, it affected TRF2 activated genes PDGFR-B and IL1R1 but not WRNIP1. [N=3] Statistical significance was calculated using Tukey’s multiple comparisons test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
Error bars correspond to standard error of mean from independent experiments.
A. TRF2-wild type (WT) and TRF2-2293R mutant proteins fused with dCAS9 expressed and targeted to the IL1R1 promoter using IL1R1-specific gRNA in HT1080 (left) or MDAMB231 cells (right). Following this, NFKB and p-NFKB (Ser536-phosphorylation) levels were assessed by flow-cytometry in ten thousand HT1080 or MDAM231 cells ; mean fluorescence signal for IL1R1 and ratio of p-NFKB/total NFKB plotted in graphs on right panels. [N=1]
B. IL1R1 knockout (KO) HT1080 cells using CRISPR were made and were transduced (lentiviral) to a doxycycline-inducible-TRF2 stable line (top left). TRF2, NFKappa-B and Ser536p-NFKappa-B in IL1R1 knockout or control HT1080 cells with or without TRF2 induction by doxycycline; GAPDH used as loading control (top right). Induction of TRF2 in HT1080 WT and IL1R1 KO cells to check for NFkappaB downstream target gene expression(bottom). Activation of NFKB target genes (∼ 2-5 fold) post-TRF2 over-expression was lower in IL1R1KO cells compared to WT cells (∼ 10 fold ) [N=3] Statistical significance was calculated using Tukey’s multiple comparisons test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
C. Phospho-NF-Kappa-B-(Ser-536p) and NFkappaB in HT1080 or HT1080-LT cells with or without IL1B stimulation (top panel); ratio of p-NFKB to total NFKB from above western blots plotted from two independent experiments (bottom panel). The NFkappa B (p65/RELA) is detected just below the 70 Kda molecular weight marker.
Error bars correspond to standard error of mean from independent experiments.
A. TNBC samples (N= 6 each for long or short telomere samples) stained with EpCAM (far red) and then hybridized with telomere specific PNA probe (FITC). Mean telomeric signal was plotted in total cells. [N=6] Statistical significance was calculated using Mann-Whitney’s non-parametric test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
B. Telomere length was determined using the previously published algorithm Tel-Seq from sequenced genomes of TNBC samples (N=8 for TNBC-ST (short telomere) and N=6 for TNBC-LT (long telomere); and adjacent normal tissue from same patient). Table with sequencing related details of the normal/tumor samples has been provided.
C. Telomerase activity was measured from protein lysates across 64 TNBC samples (long or short telomeres) normalized to the telomerase activity of HT1080 cells for the same amount of protein; spearman correlation was calculated between RTL (relative telomere length) and telomerase activity.
D. mRNA expression of hTERT and TERC in six TNBC samples (long or short telomeres) normalized to the 18S gene in respective samples. [N=6] Statistical significance was calculated using Mann-Whitney’s non-parametric test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
E. TNBC organoids (TNBO) formed from TNBO-ST (short telomere) and TNBO-LT (long telomere) samples.
F. Spearman correlation of,TERT, TERC, IL1R1 and cytokine related gene expression and RTL (Relative telomere length) in 34 TNBC tissue samples with each other and RTL represented as heat map; gene expression was normalized to the18S gene; for telomere length the 36B4 single-copy gene was used for normalization (as in Figure 1D); fold change of TL was calculated with respect to TL in HT1080 cells.
G. Overall Survival for breast cancer patients (BRCA) with high IL1R1 mRNA expression [N=254] or low IL1R1 mRNA expression [N=709] from TCGA datasets ( see Supplementary file 4 for details ) – graph tabulated using cBIOPORTAL with custom queries. Logrank p-value was found to be 0.0149.
H. Frequency of occurrence of samples with high / low IL1R1 mRNA expression in various annotated sub-types of breast cancer from TCGA ( see Supplementary file 4 for details )
Error bars correspond to standard error of mean from independent experiments.
A. Gene Ontology (GO) analysis of top 500 genes correlated with telomere elongation across 31 cancer types( as reported in Barthel et al., 2017) Top 5 hits for ‘biological processes’ represented in descending order of enrichment; respective fold enrichments shown by individual bars (x-axis – top); the false discovery rate (FDR) shown by blue line.
B. Gating strategy for Figure 8B.
C. Gating strategy for CD 45 +ve c gated ells and CD 11B +ve gated cells in TNBC. Percentage of CD 45 +ve and CD11B +ve ( in CD 45 +ve population ) was checked for TNBC-ST and TNBC-LT samples. [N=5] Statistical significance was calculated using Mann-Whitney’s non-parametric test (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).. In the CD11B +ve ( in CD 45 +ve population ), CD 206 (M2 macrophage marker) and CD 86 (M1 macrophage marker ) +ve cells were quantified. Percentage of CD 206 high CD 86 low ( M2 macrophages) and CD86 high Cd206 low (M1 macrophages ) in the myeloid cell population ( CD 45 +ve CD 11B +ve ) was plotted for three TNBC-ST and TNBC-LT samples each. [N=3] Statistical significance was calculated using unpaired T test with Welch’s correction (p values: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001, **** ≤ 0.0001).
D. TNBC tissue (N=6 each for long or short telomere samples) were tested for TERT and TERC expression (The Telomerase activity for these samples was estimated with HT1080 cell lysate as a reference. The normalized values ( to the group average) were plotted in a heat map (left) for telomerase activity, TERT and TERC gene expression in the TNBC samples used in Supplementary Figure 7D along with corresponding % TAM values (top left). The Spearman correlation matrix (top right) and p values (bottom).have been provided.
E. CD206 levels by immuno-flow cytometry in THP1 cells and macrophages differentiated from THP1 cells into M0 or M2 types (described in Methods) and plotted along X-axis in log scale (top panel). Bright-field images at 10X magnification for each cell stage shown in top right panel.
F. Different dosages of IL1RA was checked based on previously published reports and the minimal concentration where change in IL1R1 expression was seen ( 20 ng/ml) was used in other functional assays. [N=1]
G. mRNA expression of IL1R1 was tested in MDAMB-231 cells after a 48-hr. treatment with various reported G4-binding ligands.
GAPDH was used as a normalizing control. [N=2]
Error bars correspond to standard error of mean from independent experiments.
