Long Noncoding RNA MPRL Promotes Mitochondrial Fission and Cisplatin Chemosensitivity via Disruption of Pre-miRNA Processing

Abstract

Purpose: The overall biological roles and clinical significance of most long noncoding RNAs (lncRNA) in chemosensitivity are not fully understood. We investigated the biological function, mechanism, and clinical significance of lncRNA NR_034085, which we termed miRNA processing-related lncRNA (MPRL), in tongue squamous cell carcinoma (TSCC).

Experimental design: LncRNA expression in TSCC cell lines with cisplatin treatment was measured by lncRNA microarray and confirmed in TSCC tissues. The functional roles of MPRL were demonstrated by a series of in vitro and in vivo experiments. The miRNA profiles, RNA pull-down, RNA immunoprecipitation, serial deletion analysis, and luciferase analyses were used to investigate the potential mechanisms of MPRL.

Results: We found that MPRL expression was significantly upregulated in TSCC cell lines treated with cisplatin and transactivated by E2F1. MPRL controlled mitochondrial fission and cisplatin sensitivity through miR-483-5p. In exploring the underlying interaction between MPRL and miR-483-5p, we identified that cytoplasmic MPRL directly binds to pre-miR-483 within the loop region and blocks pre-miR-483 recognition and cleavage by TRBP-DICER-complex, thereby inhibiting miR-483-5p generation and upregulating miR-483-5p downstream target-FIS1 expression. Furthermore, overexpression or knockdown MPRL altered tumor apoptosis and growth in mouse xenografts. Importantly, we found that high expression of MPRL and pre-miR-483, and low expression of miR-483-5p were significantly associated with neoadjuvant chemosensitivity and better TSCC patients' prognosis.

Conclusions: We propose a model in which lncRNAs impair microprocessor recognition and are efficient of pre-miRNA cropping. In addition, our study reveals a novel regulatory network for mitochondrial fission and chemosensitivity and new biomarkers for prediction of neoadjuvant chemosensitivity in TSCC.These findings uncover a novel mechanism by which lncRNA determines mitochondrial fission and cisplatin chemosensitivity by inhibition of pre-miRNA processing and provide for the first time the rationale for lncRNA and miRNA biogenesis for predicting chemosensitivity and patient clinical prognosis.

Figure 1.. Differential expression of lncRNAs induced by cisplatin in TSCC cells and chemosensitive or nonsensitive tumors.

(A) Schematic flowchart depicting the strategy for microarray analysis and validation of lncRNAs. (B) Scatter plot of lncRNA expression in TSCC cells with or without cisplatin treatment. The red dots indicate differentially expressed lncRNAs (fold change ≥2); black dots indicate a fold change less than two in both cell lines, while green dots and blue dots indicate fold changes less than two in the SCC-9 and CAL-27 cells, respectively. X and y-axes, normalized sample signal values (log₂ scale). (C) Thirty-eight lncRNAs upregulated by cisplatin treatment were identified by qRT-PCR in both CAL-27 and SCC-9 cells. (D) Eleven lncRNAs were significantly upregulated in TSCC tumors from chemosensitive and nonchemosensitive patients. (E) Schematic annotation of the MPRL genomic locus on chr5:43,016,152–43,019,003 in humans. Green rectangles represent exons. (F) The expression of five splice variants of LOC648987 in CAL-27 and SCC-9 cells. (G) Representative images (MRI scanning) of tumor response (upper panels) and MPRL expression (lower panels) in tissue specimens from patients with chemosensitive (S) and nonchemosensitive (NS) tumors. Scale bar, 20 μm. (H) Prediction of TFs of MPRL by overlaying ChIP-seq data from UCSC and the virtual laboratory of PROMO. (I) Binding motif analysis showing enriched E2F1 motifs in the MPRL promoter; the arrow indicates the binding sites with the highest score. The green to red color gradation is based on the ranking of each binding site from the minimum (green) to maximum (red) scores, which were analyzed by JASPAR. (J) ChIP-qPCR analysis of the E2F1 genomic occupancy in the MPRL promoter in CAL-27 and SCC-9 cells, as indicated. (K) Luciferase assay showing that E2F1 knockdown inhibits MPRL promoter activity in CAL-27 cells. (L) Cisplatin induced upregulation of E2F1 and MPRL expression, while silencing E2F1 decreased MPRL levels in both TSCC cell lines. ^#P<0.05, *P<0.01 and **P<0.001 versus control, 2-tailed Student’s t tests (D and J); **P<0.001, 1-way ANOVA followed by Dunnett’s tests (K and L).

Figure 2.. MPRL promotes mitochondrial fission and cisplatin sensitivity in TSCC through the miR-483–5p-FIS1 axis.

(A-D) Knockdown of MPRL attenuated mitochondrial fission and apoptosis in CAL-27 and SCC-9 cells. Mitochondrial fission was detected by staining with MitoTracker Red (left panel) and quantified (right) (A); Scale bar, 3 μm; cell apoptosis was detected using flow cytometry (B), TUNEL (C), and caspase-3/7 activity assays (D). (E) Target miRNAs of MPRL were screened by microarray in cells treated with cisplatin. Heat map (left panel) and Venn diagrams (right) depicted differentially expressed miRNAs in cisplatin-treated CAL-27 and SCC-9 cells stably expressing shMPRLs (fold change ≥1.5). (F) The inhibitory effect of MPRL knockdown on mitochondrial fission was attenuated after inhibiting miR-483–5p levels. Mitochondrial fission was detected by staining with MitoTracker Red. (G) Western blot analysis showed that the inhibitory effect of MPRL knockdown on FIS1 expression was attenuated after inhibiting miR-483–5p levels. *P<0.01 and **P<0.001, 2-tailed Student’s t tests.

Figure 3.. MPRL can directly bind to the loop of pre-miR-483.

(A) Predicted binding sites between MPRL and pre-miR-483; △G, free energy. (B) Northern blotting (left panel) and FISH (right panel) revealed that MPRL was located in both the nuclear and cytoplasm but was predominantly located in the cytoplasm, while pre-miR-483 and miR-483–5p were mainly located in the cytoplasm; scale bar, 3 μm. (C) MPRL can directly bind to pre-miR-483 in vivo. CAL-27 cells were transfected with wild-type or mutant MPRL (wt-MPRL or mut-MPRL) or empty vector. Twenty-four hours post-transfection, cells were harvested for biotin-based pull-down assays. The seed region of wt and mut-MPRL is shown (upper panel). Pre-miR-483 was analyzed by qRT-PCR (lower panel). (D) Pre-miR-483 can bind to MPRL in vivo. Cytoplasmic lysates of CAL-27 cells were incubated with magnetic beads coated with biotin-labeled probes specific to pre-miR-483 or random probes. After the beads were washed and the bead/RNA complexes were enriched, RNA was eluted from the streptavidin beads and analyzed by Northern blot. I, input (10% of each sample); P, pellet (100% of each sample). (E) EMSA showed that pre-miR-483 selectively binds in vitro to wt-MPRL (lanes 1–3) but not mut-MPRL (lanes 4–6). (F and G) Overexpression of wt-MPRL but not mut-MPRL reduced the coimmunoprecipitated pre-miR-483 with TRBP or DICER and increased pre-miR-483 levels in CAL-27 cells. (H) Luciferase reporter assays to test miR-483–5p functionality upon MPRL overexpression. CAL-27 cells were cotransfected with the indicated reporter vector, pre-miR-483, and with either wt-MPRL or mut-MPRL. Relative firefly luciferase/renilla activity was determined and enhanced by MPRL overexpression compared with the control vector (pGL3-Control). (I and J) Knockdown of MPRL increased the coimmunoprecipitated pre-miR-483 with TRBP or DICER and reduced pre-miR-483 levels in CAL-27 cells. (K) Luciferase activity was downregulated by silencing MPRL. **P<0.001, 1-way ANOVA followed by Dunnett’s tests for multiple comparisons.

Figure 4.. MPRL inhibits pre-miR-483 recognition and cleavage by the TRBP-DICER complex in CAL-27 cells.

(A) Schematic representation of inhibition of pre-miRNA recognition and cleavage by DICER-TRBP complex in the presence of MPRL. (B) Serial deletions of MPRL were used in RNA pull-down assays to identify valid length of MPRL that is required for physically masking the recognition of pre-miR-483 by the TRBP-DICER complex. (C) Site-directed mutagenesis of MPRL to 1100 nt resulted in the inability of MPRL to mask the recognition of pre-miR-483 by TRBP and DICER. (D and E) Overexpression of truncated isoforms of MPRL increased pre-miR-483 levels (D) and subsequently reduced miR-483–5p levels (E), while mutant MPRL (1–1100) had no effect. (F) Luciferase reporter assays showed that miR-483–5p function was inhibited by overexpression of truncated MPRL. (G) Forced expression of the truncated MPRL (1–1300) but not the MPRL (1–1100) abolished the increase in pre-miR-483 immunoprecipitated with TRBP and DICER by depletion of endogenous MPRL. (H and I) Overexpression of MPRL (1–1300) and MPRL (1–1100) increased pre-miR-483 levels (H) and reduced miR-483–5p levels (I). (J) Luciferase reporter assays demonstrated that MPRL(1–1300) and MPRL(1–1100) overexpression abolished the increase in miR-483–5p functionality by MPRL depletion. (K-L) qRT-PCR demonstrated that overexpressing MPRL or the truncated isoforms (1–1300) and (1–1100) prevented the reduction in pre-miR-484 (K) and attenuated the increase in miR-483–5p (L) induced by DICER, while silencing MPRL had the opposite effects. *P<0.01 and **P<0.001, 1-way ANOVA followed by Dunnett’s tests for multiple comparisons.

Figure 5.. MPRL knockdown inhibits apoptosis and cisplatin sensitivity in CAL-27 cell xenografts in BALB/c nude mice.

(A) Tumor growth curves for CAL-27 tumors. BALB/c nude mice bearing xenografts of CAL-27 cells with stable knockdown of MPRL or negative control (Ctl) were treated with saline or cisplatin (n=6 per group). The results are expressed as the mean ± SEM. (B) Photomicrographs of tumors from each group at day 35. (C) Tumor weight for each group. (D) TUNEL assays showed that apoptosis in response to cisplatin was attenuated by MPRL knockdown; Scale bar, 20 μm. (E and F) MPRL knockdown decreased pre-miR-483 expression (E) but upregulated miR-483–5p expression (F) in CAL-27 cell xenografts upon treatment with cisplatin. (G) Western blot showing the inhibitory effect of MPRL knockdown on FIS1 expression but not PCNA expression in cisplatin-treated tumors. (A) **P<0.001, 2-way ANOVA followed by Bonferroni’s post test; (C, E, F) *P<0.01 and **P<0.001, 1-way ANOVA followed by Dunnett’s tests for multiple comparisons.

Figure 6.. MPRL, pre-miR-483 and miR-483–5p expression correlates with chemosensitivity and prognosis in TSCC patients.

(A) MPRL, pre-miR-483 and miR-483–5p expression and apoptosis were demonstrated in chemosensitive versus nonsensitive TSCC samples. MPRL, pre-miR-483 and miR-483–5p expression was analyzed using in situ hybridization (×200); apoptosis was detected using TUNEL assays; Scale bar, 20 μm. (B) Quantification of MPRL, pre-miR-483 and miR-483–5p expression in chemosensitive versus nonsensitive TSCC tumors. (C) Associations among MPRL, pre-miR-483 and miR-483–5p expression in TSCC were analyzed via Spearman rank order correlation. (D) Kaplan-Meier survival curves for TSCC patients were plotted for MPRL, pre-miR-483 and miR-483–5p expression, and survival differences were analyzed using a log-rank test. **P<0.001, 2-tailed Student’s t tests. (E) Schematic diagram depicting the proposed model in which MPRL regulates pre-miR-483 processing and determines mitochondrial fission and cisplatin sensitivity. (Left panel) TRBP recruits pre-miRNA-483 to the TRBP-DICER complex and ensures efficient pre-miRNA-483 processing. (Right panel) E2F1 transactivates MPRL expression under cisplatin treatment conditions. MPRL directly binds to the loop of pre-miR-483 and inhibits pre-miRNA recognition by physically masking the TRBP binding sites and blocking its stable association and cleavage by the TRBP-DICER complex, which compromises miRNA-483 generation and enhances FIS1 expression along with upregulated mitochondrial fission and cisplatin sensitivity.