A phosphatidic acid-binding lncRNA SNHG9 facilitates LATS1 liquid-liquid phase separation to promote oncogenic YAP signaling

Abstract

Long noncoding RNAs (lncRNAs) are emerging as a new class of important regulators of signal transduction in tissue homeostasis and cancer development. Liquid-liquid phase separation (LLPS) occurs in a wide range of biological processes, while its role in signal transduction remains largely undeciphered. In this study, we uncovered a lipid-associated lncRNA, small nucleolar RNA host gene 9 (SNHG9) as a tumor-promoting lncRNA driving liquid droplet formation of Large Tumor Suppressor Kinase 1 (LATS1) and inhibiting the Hippo pathway. Mechanistically, SNHG9 and its associated phosphatidic acids (PA) interact with the C-terminal domain of LATS1, promoting LATS1 phase separation and inhibiting LATS1-mediated YAP phosphorylation. Loss of SNHG9 suppresses xenograft breast tumor growth. Clinically, expression of SNHG9 positively correlates with YAP activity and breast cancer progression. Taken together, our results uncover a novel regulatory role of a tumor-promoting lncRNA (i.e., SNHG9) in signal transduction and cancer development by facilitating the LLPS of a signaling kinase (i.e., LATS1).

Fig. 1. The genome-wide analysis identified a lncRNA SNHG9 upregulated in breast tumors compared to cultured cancer cells.

a Schematic illustration showing the analysis of lncRNA profiles from subcutaneously xenografted tumors grown in vivo and tumor cells cultured in dishes in vitro. b A Venn Diagram showing upregulated lncRNAs within RNA-seq data, TCGA data (https://ibl.mdanderson.org/tanric/_design/basic/analysis.html) and our previous TNBC profiling data. c The list of 11 differentially expressed lncRNAs involved in 3D-cultured cells. d qRT-PCR detection of SNHG9 expression in breast cancer tissues and paired adjacent normal tissues (n = 48, Sun Yat-sen cohort). Horizontal black lines represent median values (**P < 0.01, Student’s t-test). e qRT-PCR detection of SNHG9 expression in breast cancer tissues from patients with (R⁺, n = 30) or without (R^–, n = 40) recurrence (Sun Yat-sen cohorts). Horizontal black lines represent median values (**P < 0.01, Student’s t-test). f Kaplan–Meier survival analysis of SNHG9 expression in breast cancer patients (n = 208, Kaplan–Meier analysis with Gehan–Breslow test, P = 0.005). g The colony formation assay was performed on wild-type MDA-MB-231 cells and SNHG9 KO MDA-MB-231 cells. Error bars, SEM of three independent experiments (n.s., not significant; ***P < 0.001, Student’s t-test). h Subcellular localization of RNAs was detected by RNA FISH in MDA-MB-231 cells. Scale bar, 10 μm. i qRT-PCR detection of RNA expression in cytoplasmic and nuclear fractionations of MDA-MB-231 cells. Error bars, SEM of three independent experiments. j Spatial analysis of LATS1 and SNHG9 colocalization. Representative confocal images of LATS1 (green) and SNHG9 (red) in MDA-MB-231 cells are shown. Scale bar, 10 μm. k In vitro 3D culture and sphere formation assays were performed using wild-type MDA-MB-231 cells and SNHG9 KO MDA-MB-231 cells. The representative pictures were shown (bottom) and the diameter of the sphere was measured (top). Scale bar, 200 μm. Error bars, SEM of three independent experiments (n.s., not significant; ***P < 0.001, Student’s t-test). l, m SNHG9 induces YAP nuclear translocation. YAP subcellular localization was detected in the formed SNHG9 WT or SNHG9 KO tumor sphere (l). Scale bar, 50 μm. Cells from ten different fields were randomly selected and quantified for YAP localization (m). Error bars, SEM of three independent experiments. n qRT-PCR detection of YAP target gene expression. Heatmap shows significant expression changes induced by SNHG9 KO in MDA-MB-231 cells.

Fig. 2. SNHG9 interacts with PA and LATS1.

a Immunoblot detection of proteins retrieved by in vitro-transcribed SNHG9 sense (sen.) or antisense (as.) transcripts from MDA-MB-231 cell lysates. b In vitro transcribed biotinylated SNHG9-sen. or SNHG9-as. transcripts were incubated with indicated recombinant proteins for in vitro RNA pull-down analysis. c RIP assays were performed using the indicated antibodies in MDA-MB-231 cell lysates. Error bars, SEM of three independent experiments (n.s., not significant; ***P < 0.001, Student’s t-test). d PA interacts with SNHG9. In vitro-transcribed SNHG9-sen. or SNHG9-as. transcripts were subjected to the lipid dot-blot assay. PC was used as a negative control. e RIP assays were performed using anti-GFP antibody in MDA-MB-231 cells harboring indicated vectors. Error bars, SEM of three independent experiments (n.s., not significant; ***P < 0.001, Student’s t-test). f The predicted secondary structure of SNHG9 was made by RNAfold software, and schematic illustration of SNHG9 mutations used in this study was shown. g Immunoblot detection of proteins retrieved by indicated in vitro transcribed biotinylated SNHG9 deletion mutants from recombinant GST-LATS1-CT. h PA binds with the sixth loop of SNHG9. In vitro-transcribed SNHG9 WT and its deletion mutants were subjected to the lipid dot-blot assay. i Schematic illustration of LATS1 protein and its truncations used in this study. j Two fragments in the C1 (601–830 aa) of LATS1 were named as C1a and C1b. Immunoblot detection of proteins retrieved by in vitro-transcribed biotinylated SNHG9 from indicated recombinant GST-LATS1-C1 mutants was shown. k Schematic illustration of the C1b fragment of LATS1 protein and its deletion mutants. l Four deletion mutants in the C1b (698–830 aa) were named as d1, d2, d3 and d4. Immunoblot detection of proteins retrieved by in vitro-transcribed biotinylated SNHG9 from indicated recombinant GST-LATS1-C1b mutants was shown. m PA interacts with LATS1. In vitro-purified GST-LATS1 was subjected to the lipid dot-blot assay. n SNHG9 promotes the interaction between PA and LATS1. In vitro-transcribed SNHG9 and in vitro-purified GST-LATS1 were subjected to the lipid dot-blot assay. o PA increased the levels of SNHG9-enriched LATS1. In vitro-transcribed biotinylated SNHG9 transcripts were incubated with GST-LATS1-CT proteins and 300 µM PA for in vitro RNA pull-down analysis. Immunoblot detection of GST-LATS1-CT proteins retrieved by in vitro-transcribed biotinylated SNHG9 was shown. p In vitro-transcribed biotinylated SNHG9-FL or SNHG9-D4 transcripts were incubated with GST-LATS1-C1 or GST-mutPA-LATS1-C1 proteins and 300 µM PA for in vitro RNA pull-down analysis. Immunoblot detection of GST-LATS1-C1 or GST-mutPA-LATS1-C1 proteins retrieved by in vitro-transcribed biotinylated SNHG9 was shown. q Schematic illustration of LATS1 protein with MOB1-, PA- and SNHG9-binding domains. r Overexpression of SNHG9 enhances the inhibitory effect of PA on the LATS1–MOB1 complex association. SFB-MOB1 was overexpressed in HEK293A and SNHG9-overexpressed HEK293A cells. Serum-starved above cells were treated with PA (100 μM) for 1 h. S protein beads were used to pull down the transfected SFB-MOB1, and the co-precipitated LATS1 was detected. s Knocking down the expression of SNHG9 attenuated the inhibitory effect of PA on the formation of the LATS1–MOB1 complex. SFB-MOB1 was overexpressed in HEK293A and SNHG9 KD HEK293A cells. Serum-starved above cells were treated with PA (100 μM) for 1 h. S-beads were used to pull down the transfected SFB-MOB1, and the co-precipitated LATS1 was detected.

Fig. 3. SNHG9 promotes PA-mediated LATS1 inactivation.

a In vitro LATS1 phosphorylation assay using recombinant LATS1-CT, eukaryotic purified MST1, MOB1 proteins and in vitro-transcribed RNA transcripts as indicated in NETN buffer with the presence of 500 µM ATP. Bacterially purified GST-LATS1-CT protein was used as the substrate. Immunoblots were used to detect the p-LATS1 (S909). b Flag-FL-LATS1, Flag-mutPA-LATS1 and Flag-ΔSNHG9-LATS1 were expressed in HEK293A cells, respectively, and purified by anti-Flag M2 magnetic beads. MOB1 and MST forming complex with above proteins were also pulled down. In vitro kinase assay was performed using above proteins, in vitro-transcribed SNHG9-FL, SNHG9-D12, SNHG9-D4 transcripts and PA as indicated in NETN buffer with 500 µM ATP. Bacterially purified GST-YAP protein was used as the substrate. Densitometry analysis of p-YAP/YAP levels (means ± SEM, n = 3 experiments) was shown. (n.s., not significant; *P < 0.05; ***P < 0.001; ****P < 0.0001, Student’s t-test). c, d Knockdown of SNHG9 largely revoked PA-mediated inhibition on the phosphorylation of LATS1 and YAP. Serum-starved wild-type and SNHG9 KD HEK293A cells were treated with PA (100 μM) for 1 h, and the levels of p-LATS1 (S909), p-YAP (S127) were detected using immunoblotting (c). Serum-starved wild-type and SNHG9 KD HEK293A cells were treated with PA (100 μM) for 4 h, and the expression levels of YAP target genes including CTGF and CYR61 were detected by qRT-PCR (d). Error bars, SEM of three independent experiments (n.s., not significant; ****P < 0.0001, Student’s t-test). e, f Overexpression of SNHG9 promotes PA-mediated inhibition on the phosphorylation of LATS1 and YAP. Serum-starved wild-type and SNHG9-overexpressed HEK293A cells were treated with PA (100 μM) for 1 h, and the levels of p-LATS1 (S909), p-YAP (S127) were detected using immunoblotting (e). Serum-starved wild-type and SNHG9-overexpressed HEK293A cells were treated with PA (100 μM) for 4 h, and the expression levels of YAP target genes including CTGF and CYR61 were detected by qRT-PCR (f). Error bars, SEM of three independent experiments (**P < 0.01, Student’s t-test). g, h Knockdown of SNHG9 largely revoked PA-mediated inhibition on the phosphorylation of LATS1 and YAP. Serum-starved wild-type and SNHG9 KD MDA-MB-468 cells were treated with PA (100 μM) for 1 h, and the levels of p-LATS1 (S909), p-YAP (S127) were detected using immunoblotting (g). Serum-starved wild-type and SNHG9 KD MDA-MB-468 cells were treated with PA (100 μM) for 4 h, and the expression levels of YAP target genes including CTGF and CYR61 were detected by qRT-PCR (h). Error bars, SEM of three independent experiments (n.s., not significant; ***P < 0.001, Student’s t-test). i, j Knockout of SNHG9 largely revoked PA-mediated inhibition on the phosphorylation of LATS1 and YAP. Serum-starved wild-type and SNHG9 KO MDA-MB-231 cells were treated with PA (100 μM) for 1 h, and the levels of p-LATS1 (S909), p-YAP (S127) were detected using immunoblotting (i). Serum-starved wild-type and SNHG9 KO MDA-MB-231 cells were treated with PA (100 μM) for 4 h, and the expression levels of YAP target genes including CTGF and CYR61 were detected by qRT-PCR (j). Error bars, SEM of three independent experiments (n.s., not significant; **P < 0.01; ***P < 0.001, Student’s t-test). k, l SNHG9 KO largely revoked PA-mediated YAP nuclear translocation. Serum-starved wild-type and SNHG9 KO MDA-MB-231 cells were treated with PA (100 μM) for 1 h. Representative images of YAP subcellular localization were shown (k). Cells from five different fields were randomly selected and quantified for YAP localization (l). Scale bar, 50 μm. Error bars, SEM of three independent experiments.

Fig. 4. LATS1 undergoes LLPS.

a PrLD prediction of LATS1. Top, schematic illustration of LATS1 showing domains. Bottom, predictions of PrLDs and disordered regions by Prion-like Amino Acid Composition (http://plaac.wi.mit.edu/), IUPred (https://iupred2a.elte.hu/) and D2P2 algorithms (http://d2p2.pro/). b HEK293A cells transfected with FL LATS1-GFP or ΔPrLD-LATS1-GFP (0.5 µg/well, 24-well) were analyzed by confocal microscopy. Representative pictures were shown (left panel), and the numbers of LATS1-GFP puncta per cell were counted in ten random fields (right panel). ****P < 0.0001, Student’s t-test. Scale bar, 10 μm. c Time-series fluorescence microscopy analysis of LATS1-GFP puncta. Bottom row shows zoom-in view of two fusing puncta. Scale bar, 10 μm (top) and 1 μm (bottom). d Representative micrographs of LATS1-GFP puncta before and after photobleaching. Scale bar, 0.2 μm. e Quantification of fluorescence intensity recovery in the bleached region of LATS1-GFP puncta. Error bars, SEM of three independent experiments. f Endogenous LATS1 puncta were detected using anti-LATS1 antibody in HEK293A cells permeabilized with 0.05% saponin. Scale bar, 10 μm. g Phase-separation assay of truncation mutants of LATS1 in vitro. The FL protein and PrLD fragment phase separated into liquid-like droplets, whereas ΔPrLD-LATS1-GFP failed. Scale bar, 2 μm. h Small droplets fused into larger ones over time in vitro. Scale bar, 5 μm. i LATS1-GFP fused into liquid droplets in a concentration-dependent manner in vitro. Scale bar, 5 μm. j LATS1-GFP droplets fused during the in vitro phase separation process. Scale bar, 1 μm. k The fluorescence intensity of LATS1-GFP droplets recovered after bleaching during FRAP assay. Time 0 indicates the photobleaching pulse. Scale bar, 1 μm. l Quantification of fluorescence intensity recovery in the bleached region of LATS1-GFP droplets. Error bars, SEM of three independent experiments. m Schematic illustration of LATS1 protein and its mutants. n Phase-separation assay of truncation mutants of LATS1 in vitro. FL-LATS1-GFP, FUS_IDR-LATS1-GFP and Tia1_IDR-LATS1-GFP phase separated into liquid-like droplets, whereas ΔPrLD-LATS1-GFP failed. Scale bar, 2 μm. o LATS1 KO HEK293A cells were transfected with FL-LATS1-GFP, ΔPrLD-LATS1-GFP, FUS_IDR-LATS1-GFP and Tia1_IDR-LATS1-GFP (0.3 µg/well, 24-well), then the cells were analyzed by confocal microscopy. The representative pictures were shown (left panel). The numbers of puncta per cell were counted in ten random fields (right panel). ****P < 0.0001, Student’s t-test. Scale bar, 10 μm. p LATS1 KO HEK293A cells were transfected with Flag-FL-LATS1 and its mutants (Flag-ΔPrLD-LATS1, Flag-FUS_IDR-LATS1 and Flag-Tia1_IDR-LATS1), respectively, and treated with PA (100 µM) for 1 h after serum starvation, then the subcellular localization of YAP was detected using immunofluorescence. Cells from five different fields were randomly selected and quantified for YAP localization. Error bars, SEM of three independent experiments. q, r LATS1 KO HEK293A cells were reconstituted with Flag-FL-LATS1, Flag-ΔPrLD-LATS1, Flag-FUS_IDR-LATS1 and Flag-Tia1_IDR-LATS1, respectively. qRT-PCR was used to detect the YAP target gene levels in indicated cells (q). The cell proliferation rates from 24 h to 72 h were assessed by OD density (450 nm) (r). Data are means ± SEM. n.s., not significant; ***P < 0.001; ****P < 0.0001, Student’s t-test.

Fig. 5. SNHG9 promotes LATS1 phase separation.

a Schematic diagram of subcellular fractionation procedures. b Subcellular fractionation assay was performed in the presence or absence of RNase A (50 μg/mL) in HEK293A cells. After serum starvation, cells were treated with PA (100 μM) for 1 h. Immunoblot analysis was used to probe the cytosolic fraction (S100), membrane fraction (P100), and nuclear fraction (P20) for the indicated proteins. c Serum-starved wild-type HEK293A cells and SNHG9 KD HEK293A cells, were treated with PA (100 μM) for 1 h. Then subcellular fractionation assay and immunoblot analysis were performed to show the presence of LATS1 in the nuclear fraction (P20). d Serum-starved wild-type HEK293A cells and HEK293A cells overexpressed with SNHG9 were treated with PA (100 μM) for 1 h. Then subcellular fractionation assay and immunoblot analysis were performed to present LATS1 in the nuclear fraction (P20). e HEK293A cells with or without SNHG9 overexpression were transfected with LATS1-GFP (0.5 µg/well, 24-well), and then analyzed by confocal microscopy. Representative pictures were shown (left panel) and the numbers of LATS1-GFP puncta per cell were counted in 10 random fields (right panel). Error bars, SEM of three independent experiments. **P < 0.01, Student’s t-test. Scale bar, 10 μm. f HEK293A cells expressing LATS1-GFP were transfected with exogenous SNHG9 labeled with 546-UTP or CamK-A labeled with 546-UTP, and then analyzed by confocal microscopy. Representative pictures were shown (left panel) and the numbers of LATS1-GFP puncta per cell were counted in 10 random fields. n.s., not significant; ***P < 0.001, Student’s t-test. Scale bar, 10 μm. g HEK293A cells expressing LATS1-GFP were transfected with SNHG9 siRNA or control (Ctrl) siRNA. Cells were transfected with exogenous SNHG9-sen. or SNHG9-as. labeled with 546-UTP, and then analyzed by confocal microscopy. The representative pictures were shown (left panel) and the numbers of LATS1-GFP puncta per cell were counted in 10 random fields (right panel). **P < 0.01, Student’s t-test. Scale bar, 10 μm. h HEK293A cells expressing LATS1-GFP were transfected with SNHG9 siRNA or control RNA. Cells were then transfected with exogenous SNHG9-sen. or SNHG9-as. labeled with 546-UTP, permeabilized with 0.05% saponin and analyzed by confocal microscopy. The representative pictures were shown (left panel) and the numbers of LATS1-GFP puncta per cell were counted in 10 random fields. **P < 0.01, Student’s t-test. Scale bar, 10 μm. i In vitro phase separation assay of LATS1-GFP and in vitro-transcribed SNHG9 labeled with 546-UTP. Scale bar, 2 μm. j In vitro phase separation assay showing droplet formation of LATS1-GFP at different concentrations in the presence of 100 nM SNHG9-sen. (top), 100 nM SNHG9-as. (middle) or no RNA (bottom). Scale bar, 2 μm. k In vitro phase separation assay showing that SNHG9 promotes LATS1 phase separation in a dose-dependent manner. Scale bar, 5 μm.

Fig. 6. SNHG9 and PA promote LATS1 phase separation and inhibit its activity.

a, b WT and SNHG9 KO MDA-MB-231 cells were starved overnight, treated with PA (100 μM) for 1 h, and then analyzed by confocal microscopy. The representative pictures were shown (left panel), and the numbers of LATS1 puncta per cell were counted in ten random fields (right panel). n.s., not significant; **P < 0.01; ****P < 0.0001, Student’s t-test. Scale bar, 10 μm. c, d FL-LATS1-GFP, mutPA-LATS1-GFP and ΔSNHG9-LATS1-GFP were re-expressed in LATS1 KO HEK293A cells, respectively. After serum starvation, indicated cells were treated with PA and/or transfected with SNHG9. Finally, all the cells were analyzed by confocal microscopy. Representative pictures were shown (c) and the numbers of FL-LATS1-GFP, mutPA-LATS1-GFP or ΔSNHG9-LATS1-GFP puncta per cell were counted, respectively, in 10 random fields (d). Error bars, SEM of three independent experiments. n.s., not significant; ****P < 0.0001, Student’s t-test. Scale bar, 10 μm. e, f SNHG9-FL, SNHG9-D12, SNHG9-D4 were respectively re-expressed in SNHG9 KD and LATS1 KO HEK293A cells with LATS1-GFP overexpression. After serum starvation, indicated cells were treated with PA, and then all cells were analyzed by confocal microscopy. Representative pictures were shown (e) and the numbers of LATS1-GFP puncta per cell were counted, respectively, in 10 random fields (f). Error bars, SEM of three independent experiments. n.s., not significant; **P < 0.01; ***P < 0.001; ****P < 0.0001, Student’s t-test. Scale bar, 10 μm. g Quantification of LATS1 kinase activity towards GST-YAP (pmol/min/μg) in the presence of the in vitro-transcribed SNHG9 and/or PA as indicated. The levels of released phosphate ions were measured using luminescent detector (mean ± SD; n = 3 biological replicates). n.s., not significant; **P < 0.01; ***P < 0.001, Student’s t-test. h Eukaryotic purified recombinant proteins Flag-FL-LATS1 was eluted by 3× Flag peptide. In vitro kinase assay was performed using above proteins, in vitro-transcribed SNHG9-sen./as. and PA as indicated in physiological LLPS buffer with 500 µM ATP. Bacterially purified GST-YAP protein was used as the substrate. The immunoblots were used to detect p-LATS1 and p-YAP. i Eukaryotic purified recombinant proteins Flag-FL-LATS1, Flag-mutPA-LATS1 and Flag-ΔSNHG9-LATS1 were eluted by 3× Flag peptide. In vitro kinase assay was performed using above proteins, in vitro-transcribed SNHG9-sen./as. and PA as indicated in physiological LLPS buffer with 500 µM ATP. Bacterially purified GST-YAP protein was used as the substrate. The densitometry analysis of p-YAP/YAP levels (means ± SEM, n = 3 experiments) was shown. n.s., not significant; **P < 0.01; ***P < 0.001, Student’s t-test. j Eukaryotic purified recombinant protein Flag-FL-LATS1 was eluted by 3× Flag peptide. In vitro kinase assay was performed using FL-LATS1, in vitro-transcribed SNHG9-FL, SNHG9-D12, SNHG9-D4 and PA as indicated in physiological LLPS buffer with 500 µM ATP. Bacterially purified GST-YAP protein was used as the substrate. The densitometry analysis of p-YAP/YAP levels (means ± SEM, n = 3 experiments) was shown. n.s., not significant; **P < 0.01; ***P < 0.001, Student’s t-test. k qRT-PCR detection of the expression level of SNHG9 in indicated MDA-MB-231 cell lines. Data are means ± SEM. n.s., not significant; ***P < 0.001, Student’s t-test. l, m LATS1 KO MDA-MB-231 cells were reconstituted with FL-LATS1 or ΔPrLD-LATS1. The cells were transfected with SNHG9 as indicated, and qRT-PCR was used to analyze the levels of YAP target genes (l). Cell proliferation rates from 24 h to 72 h were assessed by OD density (450 nm) (m). Data are means ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001, Student’s t-test.

Fig. 7. SNHG9 promotes tumor growth by decreasing LATS1 activity.

a In vivo analysis of tumors in mice that were subcutaneously injected with wild-type or SNHG9 KO MDA-MB-231 cells. b Representative IHC images of randomly selected tumors were shown. Scale bar, 50 µm. c The relative intensities of IHC staining were quantified by Image-pro plus 6.0 software (Media Cybernetics). Error bars, SEM of three independent experiments. **P < 0.01; ***P < 0.001, Student’s t-test. d Immunoblot detection of p-YAP (S127) and p-LATS1 (S909) in indicated xenograft tumors. e qRT-PCR detection of YAP target genes, including CTGF and CYR61 in indicated subcutaneous xenograft tumors. Data are means ± SEM. *P < 0.05; ***P < 0.001, Student’s t-test. f Immunofluorescence analysis of LATS1 puncta in the indicated xenograft tumor tissues permeabilized with 0.5% saponin (left panel). Scale bar, 10 μm. The numbers of endogenous LATS1 puncta were counted in 10 random fields (right panel). **P < 0.01, Student’s t-test. g Mice were subcutaneously injected with wild-type MDA-MB-231 cells, and xenograft tumors were analyzed at indicated time points. h qRT-PCR detection of SNHG9 expression in indicated subcutaneous xenograft tumors. Data are means ± SD. *P < 0.05, Student’s t-test. i Immunofluorescence analysis of LATS1 puncta in the indicated xenograft tumor tissues permeabilized with 0.5% saponin (left panel), Scale bar, 10 μm. The numbers of endogenous LATS1 puncta were counted out in 10 random fields (right panel). Data are means ± SD. *P < 0.05; **P < 0.01, Student’s t-test. j qRT-PCR detection of YAP target genes, including CTGF and CYR61 in indicated subcutaneous tumors. Data are means ± SD. *P < 0.05, Student’s t-test.

Fig. 8. Increased SNHG9 expression correlates with poor clinical outcomes in breast cancer patients.

aSNHG9 level was reversely associated with p-LATS1 (S909), and positively correlated with the expression of YAP, Ki-67, F4/80, CD31 and the formation of LATS1 puncta in primary human breast cancer specimens pretreated with 0.5% saponin (Sun Yat-sen cohorts, n = 100). Two representative cases are shown. Scale bar, 100 µm for HE and IHC, 50 µm for IF. b Percentages of specimens with low or high SNHG9 expression relative to levels of p-LATS1, YAP, Ki-67, F4/80, CD31 and LATS1 puncta are shown. *P < 0.05; **P < 0.01, χ² test. c Correlations between expression levels of SNHG9 and YAP target genes, including CTGF and CYR61 in breast cancer tissues (Sun Yat-sen cohorts, n = 100). RNA levels were determined using qRT-PCR and normalized to B2M. The r values and P values were calculated using Pearson’s correlation analysis.