The functional features of all the constructs generated during the present study are summarized in Supplementary Table S1 . The wild-type constructs (e.g., pmiR-617, p3’UTR-S, pDDX27, pGL3-PrMIR617-F1, pGL3-PrMIR617-F2, and pGL3-PrDDX27) were generated using human genomic DNA or cDNA as template and gene specific primers ( Supplementary Table S2 ) by standard laboratory protocols. The constructs harboring mutations (pmiR-617-M and pGL3-PrDDX27-M) were generated by site-directed mutagenesis using specific primers ( Supplementary Table S3 ), according to Sambrook et al. (16). pPr-null-DDX27 was generated by excising the luc ORF from the promoter-null pGL3-Basic vector and replacing it with the DDX27 ORF. Thereafter, the wild-type and target site mutated DDX27 promoter fragments were cloned upstream to DDX27 ORF in pPr-null-DDX27 to generate pPr-DDX27 and pPr-M-DDX27 respectively. pDDX27–3’UTR was generated by cloning 3’UTR of DDX27 downstream to pDDX27 using specific primers ( Supplementary Table S2 ). All the constructs were sequenced on an ABI PRISM A310-automated sequencer (Thermo Fisher Scientific, Waltham, MA) to confirm directionality and error-free sequence of the inserts.

Human oral squamous cell carcinoma cell lines, UPCI: SCC084 and UPCI: SCC131, were a kind gift from Prof. Susanne Gollin, University of Pittsburgh, Pittsburgh, PA (17). These cell lines were generated following Institutional Review Board guidelines from consenting patients undergoing surgery for squamous cell carcinoma of the oral cavity at the University of Pittsburgh Medical Centre (17). SCC131 cells were derived from a T₂N_2bM₀ lesion on the floor of the mouth of a 73-year-old Caucasian male, and SCC084 cells were derived from a T₂N₂M₀ lesion of the retromolar trigone of a 52-year-old Caucasian male (17). Cells were cultured in DMEM supplemented with 10% FBS and 1X Antibiotic Antimycotic solution (Sigma-Aldrich, St. Louis, MO) at 37°C in 5% CO₂.

SCC131 or SCC084 cells were seeded at a density of 2 × 10⁶ cells/well in a 6-well plate and transiently transfected with an appropriate construct or a combination of constructs using the Lipofectamine 2000 Transfection Reagent (Thermo Fisher Scientific, Waltham, MA). After 48 h, cells were harvested for either total RNA isolation using TRI-Reagent or total protein lysate preparation using the CelLytic M Cell Lysis Reagent (Sigma-Aldrich, St. Louis, MO).

For the dual-luciferase reporter assay, 5 × 10⁴ cells/well in a 24-well plate were transfected with different constructs as per experimental requirement. The assay was carried out after 48 h of transfection of SCC131 cells, using the dual-luciferase reporter assay system (Promega, Madison, WI). The pRL-TK control vector, coding for renilla luciferase, was co-transfected for normalizing the transfection efficiency. Each bar representing the relative light unit (RLU) is an average of 3 biological replicates.

The putative promoter sequences of MIR617 and DDX27 were retrieved by an in silico search using following promoter prediction databases: DBTSS (http://dbtss.hgc.jp), Promoter 2.0 (https://services.healthtech.dtu.dk/services/Promoter-2.0/), and PROMO (http://alggen.lsi.upc.es).

SCC131 cells were grown for 24 h and treated with 5-Azacytidine (Sigma-Aldrich, St. Louis, MO, USA) at a final concentration of 5 µM and the vehicle-control DMSO (Sigma-Aldrich, St. Louis, MO, USA) separately for 5 days. Total RNA, protein, and genomic DNA were isolated from 5-Azacytidine- and DMSO-treated cells for further studies.

The sodium bisulphite-treated genomic DNA samples from 5-Azacytidine- and DMSO-treated cells were used as templates to amplify MIR617 promoter with methylation-specific primers designed ( Supplementary Table S4 ) according to the MethPrimer database (http://www.urogene.org/methprimer; 18). Following amplification, PCR products were ligated to the pTZ57R TA cloning vector (Thermo Fisher Scientific, Waltham, MA, USA). Ten clones from each experimental set were sequenced on an ABI PRISM A310-automated sequencer (Thermo Fisher Scientific, Waltham, MA, USA) to identify the methylation status of the CpG sites in the MIR617 promoter in 5-Azacytidine- and DMSO-treated cells.

Total RNA was isolated using TRI-Reagent (Sigma-Aldrich, St. Louis, MO), and quantitated using the NanoDrop 1000 Spectrophotometer (Thermo Fischer Scientific, Waltham, MA). First-strand cDNA was synthesized using 2 µg of total RNA and a Verso cDNA Synthesis Kit (Thermo Fischer Scientific, Waltham, MA).

The expression level of miR-617 was determined by RT-qPCR as suggested by Sharbati-Tehrani et al. (19). Details of the primers are given in Supplementary Table S5 . The RT-qPCR analysis was carried out using the DyNAmo ColorFlash SYBR Green qPCR Kit in a QuantStudio3 Real-Time PCR System (Thermo Fischer Scientific, Waltham, MA). GAPDH and 5S rRNA were used as normalizing controls (20). The following equation ΔCt _gene = Ct _gene-Ct _{normalizing control,} was used to calculate the relative expression of gene normalized to GAPDH or 5S rRNA. Ct represents cycle threshold value. Each bar representing the relative expression of a gene of interest normalized to its appropriate house-keeping gene is an average of 2 technical replicates.

Proteins were resolved on an SDS-PAGE and then transferred to a PVDF membrane (Pall Corp., Port Washington, NY). The signal was visualized using an appropriate antibody and the Immobilon Western Chemiluminescent HRP substrate (Milipore, Billerica, MA). The following antibodies were used in this study: anti-DNMT1 antibody (1:5,000 dilution, cat# NB100–56519; Novus Biologicals, Centennial, CO, USA), anti-DDX27 antibody (1:1,000 dilution; cat# ab177950; Abcam, Cambridge, MA, USA), anti-mouse β-actin antibody (1:10,000 dilution; cat# A5441; Sigma-Aldrich, St. Louis, MO, USA), anti-phospho-p70 S6 Kinase (Thr421/Ser424) antibody (1:1,000 dilution; cat# 9204; Cell Signaling Technology, Danvers, MA, USA) and anti-p70 S6 Kinase antibody (1: 1,000 dilution; cat# 9202; Cell Signaling Technology, Danvers, MA, USA). The anti-mouse HRP-conjugated secondary antibody (cat# HP06) and anti-rabbit HRP-conjugated secondary antibody (cat# HP03) were purchased from Bangalore Genei, Bangalore, India.

To predict the gene targets of miR-617, we used a consensus approach by employing five different mRNA target prediction algorithms like TargetScanHuman 8.0 (https://www.targetscan.org/vert_80/), miRDB (https://mirdb.org/), MiRanda (https://cbio.mskcc.org/miRNA2003/miranda.html), microT-CDS (Diana Tools) (https://dianalab.e-ce.uth.gr/html/dianauniverse/index.php?r=microT_CDS), and PITA (https://genie.weizmann.ac.il/pubs/mir07/mir07_prediction.html) ( Supplementary Table S6 ).

All the 36 matched normal oral tissue and OSCC patient samples used in the present study were ascertained at the Kidwai Memorial Institute of Oncology (KMIO), Bangalore, Karnataka. The samples were obtained as surgically resected tissues from oral cancerous lesions and adjacent normal tissues (taken from the farthest margin of surgical resection) in RNALater™ (Sigma-Aldrich, St Louis, MO, USA) and transferred to -80°C until further use. The tumors were staged according to the UICC’s (union for international cancer control) TNM (tumor-node-metastasis) classification (21). The details of the clinicopathological parameters obtained from the patients are summarized in Supplementary Tables S7, S8 .

The total viable cell count was assessed by employing trypan blue dye exclusion assay as described in Karimi et al. (22). In brief, 24 h post transfection, OSCC cells were trypsinized and re-seeded at a density of 3 × 10⁴ cells/well in triplicates for each transfection experiment in 24-well plates. At each time point, cells from each well were trypsinized and pelleted by centrifugation and resuspended in complete DMEM. The cell suspension was mixed with an equal volume of 0.4% trypan blue dye (Sigma-Aldrich, USA) prepared in 1X PBS. For cell counting, 10 µL aliquot of the stained cells was added to a hemocytometer. The number of viable cells/mL was calculated using the following formula: number of viable cells/mL= mean of the total number of cells counted in the four quadrants (unstained cells) X dilution factor X 10⁴. Each data point representing the total viable cell count is an average of 3 technical replicates.

The rate of cell proliferation was determined by the CHEMICON BrdU Cell Proliferation Assay Kit (Millipore Corporation, Billerica, MA) (20). BrdU cell proliferation assay is used for the in vitro quantitative detection of newly synthesized DNA of actively proliferating cells. In brief, 2 × 10³ cells/well were seeded in triplicates in 96-well plates and transiently transfected with different plasmid constructs. The BrdU label was added on days 1, 2, and 3, and incubated for 6 h in a humidified CO₂ incubator. The remaining protocol was followed as per the manufacturer’s instructions. The absorbance was measured at 450 nm using Infinite 200 PRO Plate reader (Tecan Group Ltd, Mannedorf, Switzerland). Each bar representing absorbance at 450 nm is an average of 3 biological replicates.

Synergism between two drugs is calculated by CDI (coefficient of drug interaction); a CDI <1 indicates a synergistic relationship (23). In the present study, we employed the CDI formula to calculate the synergistic effect of the interaction between miR-617 and DDX27 on cell viability by co-transfecting their overexpression constructs in SCC131 cells. For our analysis, we designated it as the coefficient of interaction (CI). The CI was calculated as follows. CI = AB/(A × B), where AB is the ratio of the viable cell count post 72 h of cells co-transfected with overexpression constructs of miR-617 (pmiR-617) and DDX27 (pPr-DDX27) to the viable cell count post 72 h of cells co-transfected with pPr-null-DDX27 (a promoter-less construct with DDX27 ORF) and pcDNA3-EGFP, and A or B is the ratio of the viable cell count post 72 h of cells transfected with pmiR-617 or pPr-DDX27 to the viable cell count post 72 h of cells co-transfected with the pPr-null-DDX27 and pcDNA3-EGFP. Each data point representing the total viable cell count is an average of 3 technical replicates.

The CaspGLOW Fluorescein Active Caspase-3 Staining Kit (Biovision, Mountain View, CA) was used to quantify the apoptosis of cells transfected with the appropriate constructs (24). After transfection, FITC-DEVD label was added to cells, and the rest of the steps were followed as per the manufacturer’s protocol. The fluorescence intensity was measured, using an Infinite 200 PRO plate reader (Tecan Group Ltd, Mannedorf, Switzerland). Each bar representing the caspase-3 activity is an average of 4 technical replicates.

The ability of cells to grow in an anchorage independent manner was assessed by the soft agar colony-formation assay (25). After transfecting cells with appropriate constructs, they were harvested and 7 × 10³ cells were plated in 1 mL of 0.35% Noble Agar (Difco, Mumbai, India) diluted with culture media in a 35 mm dish. After 14 days, colonies were imaged using the Leica Inverted Microscope Dmi1 (Leica Microsystems, Wetzlar, Germany) and counted by the OpenCFU software (https://opencfu.sourceforge.net/). Each bar representing the number of colonies per microscopic field is an average of 4 technical replicates.

To investigate the effect of miR-617-mediated targeting of DDX27 on tumor growth, 2×10⁶ SCC131 cells were transfected with 400 nM of a synthetic miR-617 mimic or 400 nM of a mimic control separately. After 24 h of transfection, cells from both groups were suspended separately in 200 μL of 1:1 dilution of DPBS and matrigel [Corning^® Matrigel^® Growth Factor Reduced (GFR) Basement Membrane Matrix, LDEV-free, cat # 354230, New York, USA]. A total of six 4–6 weeks old female mice (NU/J (athymic nude), three per group, were injected with either miR-617 mimic or mimic control transfected cells on both flanks. Tumor growth was monitored by measuring its volume using a digital caliper every 2 days until 27 days. The following equation was used to measure the tumor volume V = (W ² _X L)/2, where L and W represent length and width, respectively (20). Excised tumors were weighed at the end of 27 days. All mice were maintained on a 12:12 h light/dark cycle, in proper cages with sufficient food and water. The miRIDIAN microRNA hsa-miR-617-Mimic (cat# C-300943–01-0020) and miRIDIAN microRNA Mimic Negative Control #1 (cat # CN-001000–01-20) were purchased from Dharmacon, Inc., Lafayette, CO, USA.

The statistical significance of the comparison between any two experimental data sets was calculated using the student’s t test with Welch’s correction in the GraphPad Prism 8 software (Boston, MA, USA). The statistical significance of comparisons between multiple data sets was calculated by one-way analysis of variance (ANOVA). Multiple comparisons were carried out by Dunnett’s and Sidak’s multiple comparisons tests as per experimental requirement. Comparisons among data sets were considered significant when p-values were ≤0.05 (*), <0.01 (**), <0.001 (***), <0.0001 (****) or non-significant (ns) when the p-value was >0.05.

Tumor suppressor genes are characterized by their ability to regulate the rates of cell proliferation and programmed cell death (e.g., apoptosis, autophagy etc.) to prevent cancer. The total viable cell count in a population is a balance between the rates of cell proliferation and programmed cell death. This balance is crucial for maintaining homeostasis in multicellular organisms and is also a fundamental aspect of various physiological and pathological processes. Thus, to assess if miR-617 has tumor suppressive role in OSCC cells, pmiR-617 and the vector control (pcDNA3-EGFP) were transiently transfected separately in SCC131 and SCC084 cells, and the cell viability assay was performed by cell counting using the trypan blue dye exclusion method. The results showed that miR-617 reduces the total viable cell count of OSCC cells, indicating its tumor suppressive role in OSCC ( Supplementary Figure S1A ). Furthermore, the anti-proliferative role of miR-617 in OSCC cells was established by BrdU cell proliferation assay ( Supplementary Figure S1B ).

In cancer, one of the mechanisms for silencing/downregulation of tumor suppressor miRNAs is DNA hypermethylation at their independent promoters or host gene promoters or promoters of their transcription activators. Thus, we hypothesized that the tumor suppressor microRNAs, which are silenced by promoter DNA hypermethylation, can be reactivated by demethylating their respective promoters upon treatment of cells with the passive DNA demethylating agent 5-Azacytidine. To this end, we treated the SCC131 cells with 5-Azacytidine and DMSO separately. After validating the efficacy of the 5-Azacytidine treatment by checking the protein level of DNA methyltransferase 1 (DNMT1) using Western blot analysis and the transcript level of a positive control MCPH1 by RT-qPCR ( Supplementary Figure S2 ), we checked for the transcript level of miR-617 by miR-Q RT-qPCR and found that it was significantly reactivated in 5-Azacytidine-treated cells compared to the DMSO-treated cells ( Figure 1A ).The MIR617 gene is located in the 4^th intron (34,045 bp) of the host gene LIN7A ( Supplementary Figure S3 ). Since miRNAs located in introns of >5,000 bp length are likely to have their independent transcription units, we speculated that MIR617 could be transcribed by an independent promoter. To characterize its promoter, we retrieved a short F1 (-365 bp to +460 bp relative to TSS) and a long F2 (-2925 bp to +250 bp relative to TSS) fragments using DBTSS, Promoter 2.0 and PROMO databases and characterized these cloned fragments (pGL3-PrMIR617-F1 and pGL3-PrMIR617-F2) for promoter activity, using the dual-luciferase reporter assay in SCC131 cells ( Supplementary Figures S4A, B, S5 ). The results showed that only pGL3-PrMIR617-F2 has a significant promoter activity compared to pGL3-Basic, suggesting that the F2 fragment represents the MIR617 promoter ( Supplementary Figure S5 ). To determine the methylation status of the MIR617 promoter post 5-Azacytidine treatment, we chose a CpG-rich region and divided it into two segments (218 bp long F2-I and 344 bp long F2-II) and performed the bisulphite sequencing PCR (BSP) analysis ( Supplementary Figure S6 ). The results showed a significant decrease in cumulative methylation percentage of the MIR617 promoter from 88% to 56% in 5-Azacytidine-treated cells compared to DMSO-treated cells ( Figures 1B, C ). This indicated that the upregulation of miR-617 following the treatment of SCC131 cells by 5-Azacytidine ( Figure 1A ) is due to hypomethylation of the MIR617 gene promoter, further suggesting that the MIR617 promoter hypermethylation is one of the mechanisms for miR-617 downregulation in OSCC.

miR-617 demonstrated an anti-proliferative role in OSCC cells, thus we expected its gene targets to have pro-proliferative functions. We employed five miRNA target prediction tools and obtained eight putative protein-coding gene targets which have putative miR-617 binding sites in their 3’UTRs ( Supplementary Table S6 ). Since DDX27 is a known oncogene in other cancers, we chose to investigate it as a potential gene target for miR-617. To determine if miR-617 regulates DDX27, we transfected the vector control, pcDNA3-EGFP, and pmiR-617 in increasing quantities in SCC131 cells and observed that miR-617 upregulates DDX27 at both transcript and protein levels in a dose-dependent manner, instead of downregulating it ( Figure 2A ). To ascertain the specificity of DDX27 upregulation post miR-617 overexpression, we transfected SCC131 cells with the vector control, pmiR-617, and pmiR-617-M (miR-617 overexpression construct with the deleted seed region) separately and performed RT-qPCR and Western blot analysis. The results showed that the miR-617 level was upregulated in the cells transfected with pmiR-617 compared to those transfected with the vector control or pmiR-617-M. As expected, DDX27 level was specifically upregulated in cells transfected with pmiR-617 compared to those transfected with the vector control or pmiR-617-M ( Figure 2B ). Further, the tumor-suppressive role of miR-617 was also specific to miR-617 levels, wherein total viable cell count significantly reduced in SCC131 cells transfected with pmiR-617 in comparison to those transfected with pcDNA3-EGFP or pmiR-617-M ( Supplementary Figure S7 ).

To explain the upregulation of DDX27 by miR-617, we speculated that there could be a noncanonical interaction of miR-617 with the DDX27 promoter to upregulate it. We therefore bioinformatically retrieved the putative DDX27 promoter fragment (-642 bp to +67 bp relative to TSS) and characterized the cloned fragment pGL3-PrDDX27 for promoter activity by dual-luciferase reporter assay in SCC131 cells ( Supplementary Figures S8A, B ). The results showed that it has a significant promoter activity compared to pGL3-Basic ( Supplementary Figure S8B ). Subsequently, we observed a putative miR-617 binding site in the DDX27 promoter as well ( Figure 3A ). Thus, we aimed to determine if miR-617 interacts with the DDX27 promoter or 3’UTR or both ( Figure 3B ). To this end, we transfected p3’UTR-S, pGL3-PrDDX27, and pGL3-PrDDX27-M separately or in combination with pmiR-617 in SCC131 cells and performed the dual-luciferase reporter assay. The results showed that miR-617 interacts with the DDX27 promoter only ( Figure 3B ). Next, pGL3-PrDDX27 was co-transfected with an increasing dosage of pmiR-617 in SCC131 cells and the results showed that miR-617 positively regulates the DDX27 promoter activity in a dose-dependent manner ( Figure 3C ). Further, co-transfection of SCC131 cells with pmiR-617 and pGL3-PrDDX27-M or pmiR-617-M and pGL3-PrDDX27 showed that miR-617 regulates the DDX27 promoter activity in a sequence-specific manner ( Figure 3C ).

A DDX27 overexpression construct driven by its own promoter (pPr-DDX27) was generated to mimic the endogenous transcription unit ( Supplementary Figure S9 ). When pPr-DDX27 was co-transfected with pmiR-617 in SCC131 cells, DDX27 was upregulated at transcript and protein levels compared to those co-transfected with pPr-DDX27 and the vector control (pcDNA3-EGFP) due to the interaction of miR-617 with the DDX27 promoter in pPr-DDX27 ( Figure 4 ). As expected, no change was observed in DDX27 transcript and protein levels in cells co-transfected with pDDX27–3’UTR and pmiR-617 compared to those co-transfected with pDDX27–3’UTR and the vector control ( Figure 4 ). Thus, the in vitro construct system recapitulated the DDX27 upregulation observed upon miR-617 overexpression.

Further, to determine the biological relevance of the interaction between miR-617 and the DDX27 promoter, we evaluated their levels in 36 matched normal oral tissue and OSCC patient samples by RT-qPCR. The results showed that miR-617 was significantly downregulated in 11/36 (viz., patient no. 3,11,36,39,62,71,73,75,81,86, and 40) OSCC patient samples in comparison to their matched normal oral tissue samples ( Figure 5 , top panel). Alongside, DDX27 was significantly downregulated in 19/36 (viz., patient no. 3,11,36,39,62,71,73,75,81,86,35,74,76,80,82,83,87,47, and 67) OSCC patient samples when compared to their matched normal oral tissue samples ( Figure 5 , bottom panel). Moreover, there was a significant upregulation in the levels of miR-617 in 18/36 (viz., patient no. 12,25,41,43,72,77,35,74,76,80,82,83,87,38,42,78,88, and 89) and of DDX27 in 8/36 (viz., patient no. 12,25,41,43,72,77,79, and 85) OSCC patient samples compared to their matched normal oral tissue samples ( Figure 5 ). The levels of miR-617 and DDX27 did not show any change in 7/36 (viz., patient no. 37,47,67,79,84,85, and 90) and 9/36 (viz., patient no. 37,38,40,42,78,84,88,89, and 90) OSCC patient samples, respectively, when compared to their matched normal oral tissue samples ( Figure 5 ). We observed that miR-617 and DDX27 levels were concomitantly downregulated in 10/36 (viz., patient no. 3,11,36,39,62,71,73,75,81, and 86) and upregulated in 6/36 (viz., patient no. 12,25,41,43,72, and 77) OSCC samples compared to their matched normal oral tissue samples ( Figure 5 ). Taken together, RT-qPCR analyses showed that in the OSCC tissues, miR-617 and DDX27 levels concomitantly increased or decreased in 16/36 (44.44%) samples compared to their matched normal tissue samples (viz., patient no. 3,11,36,39,62,71,73,75,81,86,12,25,41,43,72, and 77), exemplifying the biological relevance of the interaction between miR-617 and the DDX27 promoter ( Figure 5 ). However, we did not observe a concomitant increase or decrease of miR-617 and DDX27 levels in the remaining 20/36 (55.56%) OSCC samples compared to their matched normal oral tissue samples.

Interestingly, 11/16 (68.75%) OSCC samples which showed the biological relevance between miR-617 and DDX27 levels are T4 stage tumor samples; 10/16 (62.5%) OSCC samples originated from OSCC of the buccal mucosa ( Figure 5 ; Supplementary Table S8 ). Additionally, it is noteworthy that DDX27 levels were downregulated in 10/11 (90.9%) OSCC samples compared to their matched normal oral tissue samples (viz., patient no. 3,11,36,39,62,71,73,75,81, and 86), wherein miR-617 levels were also downregulated ( Figure 5 ). DDX27 levels were upregulated in 6/18 (30%) OSCC samples compared to their matched normal oral tissue samples (viz., patient no. 12,25,41,43,72, and 77), wherein miR-617 levels were also upregulated.

The positive correlation in the transcript levels of miR-617 and DDX27 served as an impetus to study the role of DDX27 in cell viability and proliferation. We transfected pDDX27 and the vector control, pcDNA3.1(+), separately in OSCC cells and found a significant reduction in total viable cell count of cells transfected with pDDX27 compared to those transfected with the vector control ( Figure 6 ). Furthermore, BrdU cell proliferation assay in SCC131 cells showed that there was a significant decrease in proliferation of cells transfected with pDDX27 compared to those transfected with the vector control ( Supplementary Figure S10 ). This suggested that DDX27 acts as a negative regulator of cell viability and proliferation in OSCC cells contrary to its oncogenic role in other cancers.

Next, to explore the potential molecular effectors of the miR-617/DDX27 axis, we chose to examine the PI3K/AKT/MTOR pathway as it is one of the most frequently activated pathways in cancers, including OSCC. We transfected OSCC cells with pmiR-617 or pDDX27 separately and assessed the levels of phospho-S6K1 (p-S6K1), the molecular readout of this pathway. The results showed that miR-617 and DDX27 suppress the PI3K/AKT/MTOR pathway activation by reducing p-S6K1 levels, indicating that miR-617 modulates this pathway, in part, via regulating DDX27 levels. This further correlates the reduction in cell proliferation by miR-617 to the decreased activation of the pro-survival PI3K/AKT/MTOR pathway ( Figure 7 ).

We co-transfected OSCC cells with different combinations of pmiR-617, pmiR-617-M and DDX27 constructs (viz., pPr-null-DDX27 and pPr-DDX27) to investigate the effect of miR-617 on the protein levels of DDX27. The results showed that the endogenous DDX27 level was upregulated in OSCC cells co-transfected with pPr-null-DDX27 and pmiR617 compared to those co-transfected with pPr-null-DDX27 and pcDNA3-EGFP ( Supplementary Figure S11 ). Moreover, there was a further increase in DDX27 level in cells co-transfected with pPr-DDX27 and pmiR-617 in comparison to those co-transfected with pPr-DDX27 and pcDNA3-EGFP by virtue of the interaction between miR-617 and the endogenous and exogenous promoters of DDX27 ( Supplementary Figure S11 ). As expected, in OSCC cells co-transfected with pPr-DDX27 and pmiR-617-M neither the endogenous DDX27 promoter nor the DDX27 promoter in pPr-DDX27 was induced to cause DDX27 upregulation in comparison to those co-transfected with pPr-DDX27 and pcDNA3-EGFP, due to the absence of a functional miR-617. This data further underscores the exclusivity of DDX27 upregulation by miR-617 ( Supplementary Figure S11 ).

To investigate the effect of miR-617-mediated upregulation of DDX27 on cancer hallmarks, we co-transfected the same set of constructs in SCC131 and SCC084 cells. We demonstrated that miR-617 and DDX27 are negative regulators of cell viability and proliferation ( Supplementary Figures S1A, B, 6, S10 ). Thus, as expected, cell viability and anchorage-independent growth was significantly reduced in SCC131 cells co-transfected with pPr-null-DDX27 and pmiR-617 or pPr-DDX27 and pcDNA3-EGFP compared to those co-transfected with pPr-null-DDX27 and pcDNA3-EGFP ( Figures 8A, B ). Moreover, the interaction of miR-617 with the endogenous and exogenous DDX27 promoters further reduced cell viability and anchorage-independent growth in SCC131 cells co-transfected with pPr-DDX27 and pmiR-617 compared to cells co-transfected with pPr-DDX27 and pcDNA3-EGFP ( Figures 8A, B ). This interaction was abolished in those co-transfected with pPr-DDX27 and pmiR-617-M, indicating that miR-617 regulates cell viability and anchorage-independent growth by targeting the DDX27 promoter in SCC131 cells ( Figures 8A, B ). Similarly, we also showed that miR-617 positively regulates apoptotic induction of SCC131 cells, in part, via targeting the promoter of DDX27 ( Figure 8C ). Thus, our data in SCC131 cells suggested that the significant reduction in OSCC cell viability that occurs upon overexpression of miR-617 is, in part, a combined effect of reduced cell proliferation and increased apoptosis brought about by upregulation of DDX27. miR-617-mediated upregulation of DDX27 showed a similar effect on cancer hallmarks in SCC084 cells ( Supplementary Figure S12 ).

We also showed that there is a functional synergism between miR-617 and DDX27 by calculating their CI (co-efficient of interaction) in SCC131 cells. We observed that the CI in cells co-transfected with pPr-DDX27 and pmiR-617 was 0.83, indicating marked functional synergism ( Supplementary Figure S13 ). As expected, CI was 1 (actual value 0.99) in cells co-transfected with pPr-M-DDX27 and pmiR-617. Since there is functional synergism between miR-617 and DDX27, we decided to test if there is a positive feedback loop between them as well. Thus, we transfected SCC131 cells with an increasing dosage of pDDX27 and observed that the endogenous miR-617 levels were significantly upregulated compared to those transfected with the vector control ( Supplementary Figure S14 ). This suggested that DDX27, in turn, can also upregulate miR-617 at the transcriptional level and/or in its biogenesis.

To explore the therapeutic potential of miR-617-mediated upregulation of DDX27 on OSCC xenografts in nude mice, we transfected SCC131 cells with different quantities of a synthetic miR-617 mimic and a relevant mimic control separately to optimize the dosage. The results showed that both 200 nM and 400 nM dosages of miR-617 mimic effectively increased the levels of DDX27 in SCC131 cells in a dose-dependent manner ( Supplementary Figure S15A ). As expected, the RT-qPCR analysis showed a significantly robust expression of miR-617 and a concomitant increase in DDX27 levels in mimic-transfected cells compared to those transfected with the mimic control, confirming its specificity ( Supplementary Figure S15A ). To further ascertain the exact dosage to perform OSCC xenograft studies, we checked for the effect of miR-617 mimic on viability of SCC131 cells. We observed that the cell viability is significantly reduced in cells transfected with 200 nM and 400 nM of miR-617 mimic in comparison to those transfected with the mimic control ( Supplementary Figure S15B ). However, a maximum reduction in cell viability was observed in cells transfected with 400 nM of miR-617 mimic, suggesting that 400 nM of miR-617 mimic is optimum to perform OSCC xenograft study in nude mice ( Supplementary Figure S15B ). We then transfected 400 nM of miR-617 mimic or mimic control in SCC131 cells for the nude mice experiments. The results showed that nude mice xenografts derived from SCC131 cells pre-transfected with 400 nM of the miR-617 mimic showed a significant reduction in tumor volume and weight compared to those derived from SCC131 cells pre-transfected with 400 nM of the mimic control ( Figures 9A–C ). As expected, subsequent RT-qPCR and Western blot analysis showed that DDX27 was significantly upregulated in OSCC xenografts with the miR-617 mimic in comparison to those with the mimic control ( Figure 9D ). Therefore, these results indicated that miR-617 acts as a tumor suppressor in OSCC, in part, by inducing DDX27 levels.