RECQ5 mediates pre-rRNA processing in nucleolus

Abstract

RECQ5 is a member of the RECQ helicase family that maintains genomic stability. However, the molecular mechanism of RECQ5 in this biological process remains elusive. Here, we show that RECQ5 localizes in the dense fibrillar component of nucleolus and associates with several pre-rRNA processing factors. It recognizes pre-rRNA and is able to unwind double-stranded RNA in vitro. Loss of RECQ5 induces the accumulation of 47S, 30SL5', and 30S pre-rRNA, and the reduction of 21S pre-rRNA, suggesting that it regulates the processing of pre-rRNA. Cancer-associated mutations of RECQ5 abolish its nucleolar localization as well as its helicase activities. Moreover, lacking RECQ5 leads to the unprocessed pre-rRNA hybridizing with rDNA, triggering the R-loop formation and ATR activation. Since both RECQ5 and ATR protect genomic stability in nucleolus, suppression of RECQ5 sensitizes tumor cells to the ATR inhibitor treatment. Collectively, this study reveals that RECQ5 plays a crucial role in pre-rRNA processing and ribosomal DNA (rDNA) stability.

Graphical Abstract

Figure 1.

RECQ5 localizes in DFC. (A) A schematic diagram shows the three layers of nucleolus. POLR1A, DKC1, and NPM1 are markers for FC, DFC, and GC, respectively. (B) EGFP–RECQ5 is enriched in DFC. EGFP–RECQ5 was expressed in U2OS cells. EGFP–RECQ5 and nucleolar markers of FC, DFC, and GC were examined in nucleolus by SIM. The fluorescence intensity on the white dash line was plotted (lower panels). Scale bar, 1 μm. (C, D) RECQ5 interacts and colocalizes with proteins in DFC. Co-immunoprecipitation (Co-IP) and western blotting were performed with indicated antibodies (C). EGFP–RECQ5 and DFC-located proteins were examined in nucleolus by SIM. The fluorescence intensity on the white dash line was plotted (lower panels). Scale bar, 1 μm (D). (E) RECQ5 relocates to nucleolar caps upon inhibition of RNA Pol I. U2OS cells were treated with 1 μM BMH-21 for 2 h. RECQ5 and POLR1A were examined by IF. The percentage of the cells with RECQ5 staining in nucleolar caps was calculated. ***P < .001. Scale bar, 10 μm.

Figure 2.

Loss of RECQ5 impairs pre-rRNA processing. (A) Schematic view of the major pre-rRNA precursors detected in human cells. The probes targeting 5′ETS, ITS1, and ITS2 for northern blotting are indicated by the green lines. The cleavage sites at each processing step are indicated by the yellow marks. (B) RECQ5 depletion disrupts pre-rRNA processing. Northern blotting was performed to detect pre-rRNA intermediates in WT and RECQ5-KO HeLa cells. The relative levels of 47S, 30SL5′, 30S, 21S, 32S, and 12S pre-rRNA were measured. GAPDH was used as the loading control. (C) RECQ5 regulates pre-rRNA processing, which is dependent on the helicase activity. Northern blotting was performed in cells deleted endogenous RECQ5 and reexpressing ectopic RECQ5, helicase-dead mutant RECQ5^D157A. The relative levels of 47S, 30SL5′, 30S, and 21S pre-rRNA were measured. (D) RECQ5 inhibitor treatment impairs pre-rRNA processing. HeLa cells were treated with RECQ5-IN-1 (5 μM) for 24 h. Northern blotting was performed to detect pre-rRNA intermediates. The relative levels of 47S, 30SL5′, 30S, and 21S pre-rRNA were measured. Data are represented as means ± SD as indicated. Two-tailed Student’s t-test was used to determine statistical significance. *P < .05; **P < .01; ***P < .001; ns, not significant.

Figure 3.

RECQ5 unwinds dsRNA in vitro. (A) RECQ5 unwinds the dsRNA. Recombinant RECQ5 was incubated with dsDNA, RNA/DNA hybrids, and dsRNA. The reaction products were analyzed using gel electrophoresis. (B) RECQ5 consumes ATP in the presence of RNA. Recombinant RECQ5 was incubated with or without dsRNA. The ATP levels were measured. (C) The D157A mutation abolishes the enzymatic activity of RECQ5. The recombinant D157A mutant was generated from E. coli. (D) The helicase activity of RECQ5 is dependent on ATP. ATP or adenosine 5-[γ-thio] triphosphate (ATP-γ-S) was included in the in vitro helicase assays.

Figure 4.

Characterization of cancer-associated mutations of RECQ5. (A) Schematic of the representative mutations of RECQ5. (B) The truncation mutants, including Q139*, W334*, and S727Pfs*27, do not localize in nucleolus. GFP-tagged RECQ5 mutants and POLR1A were examined by IF. Scale bar, 10 μm. (C) The missense mutations, including S59F, P78L, and S102L, do not affect the nucleolar localization of RECQ5. The GFP-tagged RECQ5 mutants and POLR1A were examined in nucleolus by SIM. The fluorescence intensity on the white dash line was plotted (lower panels). (D) The missense mutations impair the helicase activities of RECQ5. Wild-type RECQ5 or RECQ5 mutants were incubated with the dsRNA substrates, and in vitro helicase assays were conducted. The reaction products were analyzed using gel electrophoresis (lower panels).

Figure 5.

Loss of RECQ5 induces the RNA/DNA hybrid at rDNA loci. (A) Loss of RECQ5 leads to the accumulation of unprocessed pre-rRNA in FC. Nascent pre-rRNA was detected in nucleolus by 5′ETS probe in cells lacking RECQ5. SIM was performed. The staining of POLR1A indicates FC. Scale bars, 1 μm. The relative signal intensity of 5′ETS was calculated in 100 FC (right panel). (B) A model of RECQ5-mediated pre-rRNA releasing from FC. (C) Loss of RECQ5 induces accumulation of the RNA/DNA hybrid in nucleolus. The RNA/DNA hybrids were stained in nucleolus by S9.6 antibody. The staining of POLR1A indicates FC. SIM was performed. 100 nucleoli were examined in each group. Scale bars, 1 μm. The fluorescence intensity on the white dash line was plotted (lower right panel). (D) Ectopically expressed RNaseH1 suppresses the RNA/DNA hybrid. RNaseH1-GFP and empty vector were expressed in HeLa cells lacking RECQ5. The RNA/DNA hybrid was examined in nucleolus by S9.6. 100 nucleoli were examined in each group. Scale bars, 1 μm. (E) The positions of primer pairs at rDNA locus. (F) Loss of RECQ5 causes the RNA/DNA hybrid at rDNA locus. DRIP-qPCR was performed with S9.6 antibody and indicated primers. The assays were triplicated. (G) Wild-type but not cancer-associated mutations of RECQ5 suppress the RNA/DNA hybrid at rDNA locus. Data are represented as means ± SD as indicated. Two-tailed Student’s t-test was used to determine statistical significance. ***P < .001; ns, not significant.

Figure 6.

Loss of RECQ5 induces cancer cell lines hypersensitive to ATR inhibitor treatment. (A) Loss of RECQ5 induces ATR activation in nucleolus. pATR (pT1989) was examined in nucleolus. The staining of POLR1A indicates FC. SIM was performed. 100 nucleoli were examined in each group. Scale bars, 1 μm. The fluorescence intensity on the white dash line was plotted (lower right panel). (B) Ectopically expressed RNaseH1 suppresses ATR activation in the RECQ5-deficient cells. RNaseH1-GFP and empty vector were expressed in HeLa cells lacking RECQ5. pATR was examined in nucleolus. 100 nucleoli were examined in each group. Scale bars, 1 μm. (C) Loss of RECQ5 activates ATR-dependent signal pathway. Western blotting was performed with indicated antibodies. GAPDH was used as a protein loading control. The quantitative analysis (relative to GAPDH) was performed. (D) ATR inhibitor treatment induces apoptosis in RECQ5-deficient cells. The apoptotic cells were examined by flow cytometry with staining of Annexin-V and PI. n = 3. (E) RECQ5-deficient cells are hypersensitive to ATR inhibitor (VE-821) treatment. RECQ5-WT and RECQ5-KO cells were treated with the indicated dose of VE-821 for 4 days. Cell viability was determined using CellTiter-Glo assays. (F) Combination treatment with RECQ5i (RECQ5-IN-1) and ATRi (VE-821) suppresses the growth of cancer cell lines. MDA-MB-231, U2OS, and MCF-7 cells were treated with the indicated dose of ATRi or together with 5 μM of RECQ5i for 4 days. Data are represented as means ± SD as indicated from three independent experiments. Two-tailed Student’s t-test was used to determine statistical significance. *P < .05; **P < .01; ***P < .001; ns, not significant.