Repression of rRNA gene transcription by endothelial SPEN deficiency normalizes tumor vasculature via nucleolar stress
Zi-Yan Yang, … , Tian Xiao, Hua Han
Zi-Yan Yang, … , Tian Xiao, Hua Han
Published August 22, 2023
Citation Information: J Clin Invest. 2023;133(20):e159860. https://doi.org/10.1172/JCI159860.
View: Text | PDF
Research Article Vascular biology

Repression of rRNA gene transcription by endothelial SPEN deficiency normalizes tumor vasculature via nucleolar stress

Abstract

Human cancers induce a chaotic, dysfunctional vasculature that promotes tumor growth and blunts most current therapies; however, the mechanisms underlying the induction of a dysfunctional vasculature have been unclear. Here, we show that split end (SPEN), a transcription repressor, coordinates rRNA synthesis in endothelial cells (ECs) and is required for physiological and tumor angiogenesis. SPEN deficiency attenuated EC proliferation and blunted retinal angiogenesis, which was attributed to p53 activation. Furthermore, SPEN knockdown activated p53 by upregulating noncoding promoter RNA (pRNA), which represses rRNA transcription and triggers p53-mediated nucleolar stress. In human cancer biopsies, a low endothelial SPEN level correlated with extended overall survival. In mice, endothelial SPEN deficiency compromised rRNA expression and repressed tumor growth and metastasis by normalizing tumor vessels, and this was abrogated by p53 haploinsufficiency. rRNA gene transcription is driven by RNA polymerase I (RNPI). We found that CX-5461, an RNPI inhibitor, recapitulated the effect of Spen ablation on tumor vessel normalization and combining CX-5461 with cisplatin substantially improved the efficacy of treating tumors in mice. Together, these results demonstrate that SPEN is required for angiogenesis by repressing pRNA to enable rRNA gene transcription and ribosomal biogenesis and that RNPI represents a target for tumor vessel normalization therapy of cancer.

Authors

Zi-Yan Yang, Xian-Chun Yan, Jia-Yu-Lin Zhang, Liang Liang, Chun-Chen Gao, Pei-Ran Zhang, Yuan Liu, Jia-Xing Sun, Bai Ruan, Juan-Li Duan, Ruo-Nan Wang, Xing-Xing Feng, Bo Che, Tian Xiao, Hua Han

×

Figure 9

Spen ablation–mediated tumor vessel normalization requires p53.

Options: View larger image (or click on image) Download as PowerPoint
Spen ablation–mediated tumor vessel normalization requires p53. (A) pRN...
(A) pRNA expression in TECs, as determined by strand-specific RT-qPCR (n = 6). (B and C) Expression of (B) pre-rRNA, 18S, 5.8S, and 28S rRNAs as well as (C) Rpl5, Rpl11, and Rpl23 in TECs was determined by RT-qPCR (n = 5). (D and E) Expression of p53 and p21 in TECs was determined by (D) RT-qPCR (n = 4) and (E) immunoblotting (n = 6). β-Actin served as the loading control. (F) Mice with different genotypes were inoculated with LLC cells. Tumors were dissected on day 21 after inoculation (Supplemental Figure 8E). Tumor size and weight were quantified (n = 9, 9, 10, and 9 for control, eSpen–/–, ep53+/–, and eSpen–/–ep53+/–, respectively). (G) LLC tumors on day 21 after inoculation were stained with Hypoxyprobe, immunofluorescence, or assayed for vessel perfusion and leakage with FITC dextran 2MD or Texas Red–dextran 70 KD. Scale bar: 100 μm. (H) The hypoxia (n = 8, 6, 4, and 4 for control, eSpen–/–, ep53+/–, and eSpen–/–ep53+/–, respectively), (I) vessel density (CD31+) (n = 9, 7, 5, and 4 for control, eSpen–/–, ep53+/–, and eSpen–/–ep53+/–, respectively), (J) pericyte coverage (CD31+NG2+) (n = 9, 6, 5, and 4 for control, eSpen–/–, ep53+/–, and eSpen–/–ep53+/–, respectively) as well as (K) vessel perfusion and (L) leakage (n = 5, 6, 3, and 5 for control, eSpen–/–, ep53+/–, and eSpen–/–ep53+/–, respectively) were quantified. Data represent mean ± SEM. Unpaired 2-tailed Student’s t test was used for A–E and 1-way ANOVA with Tukey’s multiple comparisons test was used for F and H–L.