Fig 2: GSK503 obtains efficient therapeutical effect in RB1 non-mutated tumors.a, b The proliferation ability of RB44 and IM9 cells after GSK503 treatment. **P < 0.01. c A soft agar colony formation assay was performed to determine the tumor formation ability of RB44 and IM9 cells after GSK503 treatment. d Suppressive effects on tumor volume in the GSK503-treated group in the orthotopic xenograft model. Scale bars: 1 mm. e Suppressive effects on tumor weight in the GSK503-injected group in the orthotopic xenograft model. **P < 0.01. f Suppressive effects on tumor volume in the GSK503-injected group in the subcutaneous xenograft model. g Suppressive effects on tumor weight in the GSK503-injected group in the subcutaneous xenograft model. **P < 0.01.

Fig 3: Schematic diagram of the research model.When the CTCF-mediated chromosomal loop between the RB1 promoter and its 25 kb downstream suppressor was formed, EZH2 was then recruited and increased the level of H3K27me3 at the RB1 promoter, thereby inhibiting pRB expression and leading to tumorigenesis. The targeted correction of abnormal chromosomal interactions inhibited tumorigenesis by suppressor deletion or GSK503 inhibition. Every element of this image is new-created by the authors.

Fig 4: A functional chromosomal looping at RB1 locus.A Western blot showed abundance of pRB at the protein level in RB44, IM9, and the control (RPE and HDF) cell lines. **P < 0.01. B A chromosomal conformation capture (3C) assay was performed to detect intrachromosomal interactions between the RB1 promoter and regions in the RB1 locus. Top: Schematic diagram of variant primer sets in 3C assay. Using EcoRI as the enzyme cutting site, 14 sites were selected around the 13q14 locus. E5 was set as 3C bait. Bottom: the intrachromosomal interaction frequency between the E5 and E7 regions was determined by normalizing the 3C PCR signal to that of the positive control (input DNA). **P < 0.01 compared to negative control ARPE19 cells. C A 3C assay was performed to detect the chromosomal looping between E5 and E7 regions in RB44 and the control (RPE) cells by PCR. D The 3C products were confirmed by DNA sequencing. The 3C products derived from the RB1 promoter E5-E7 interaction were cloned and sequenced. The 3C products contained the EcoRI site that was flanked on both sides near the TSS and 25 kb downstream. We set the TSS of RB1 gene as Zero Point, the distance between the TSS of RB1 and each EcoRI cutting site was shown in (B). E5 1753 bp: the distance between the TSS of RB1 and E5 EcoRI cutting site. E7 25898 bp: the distance between the TSS of RB1 and E7 EcoRI cutting site. E A 3C assay was performed to detect the existence of E5-E7 chromosomal looping in IM9 and the control (HDF) cells by PCR. **P < 0.01. F Schematic of pGL3-promoter-RB1-S (pGL3-RB1-S) construction. The E7 fragment was amplified and inserted into pGL3-promoter-Luc. RB1-P, promoter of RB1; RB1-S, suppressor of RB1. G The promoter activity detected in dual luciferase reporter system. The 1.2 kb E7 fragment was amplified and inserted into pGL3-promoter-Luc with firefly luciferase reporter (pGL3-RB1-S). The similar random fragment was amplified and inserted into pGL3-promoter-Luc for negative control (pGL3-NC). Empty pGL3-promoter-Luc vector was used as mock. All the above three groups were transfected into pRL-TK vector with renilla luciferase, which was used as internal control to detect the transfection efficiency. Control: cells were transfected with pRL-TK vector only. All data were calculated as the ratio of firefly to renilla luciferase activity (Fluc/Rluc) in dual luciferase reporter system. For comparison, the ratio of Fluc/Rluc of the mock was arbitrarily set as 1 in the calculation. ***P < 0.001 compared to mock luciferase expression.

Fig 5: Summary of ATO new mode of action in APL cells. ATO phosphorylated pRb at T373 and S608 residues leading to heteromerization with the E2F1promoter, which resulted in reduced phosphorylation of PI3K signaling molecules and expression of E2F1, cyclin E, and inhibition of APL cell proliferation (Figure 5). APL, acute promyelocytic leukemia; ATO, arsenic trioxide; E2F1, E2F Transcriptional factor‐1