A Yin-Yang 1/miR-30a regulatory circuit modulates autophagy in pancreatic cancer cells

Six human pancreatic cancer cell lines (BxPC-3, CFPAC-1, Colo-357, MiaPaCa-2, PANC-1, and SW1990) were purchased from the Shanghai Cell Bank (Shanghai, China). The normal human pancreatic ductal cell line, HPNE, was purchased from the American Type Culture Collection (ATCC, USA). Cell lines that stably overexpress YY1 (BxPC-3 and PANC-1), cell lines in which YY1 is knocked down, and control cell lines were prepared and cultured as previously described [20]. Human pancreatic cancer cell lines were cultured in DMEM supplemented with 10% FBS, penicillin (100 U/mL), and streptomycin (100 μg/mL). The HPNE cell line was cultured according to the recommendations of the ATCC.

Total RNA was extracted from different cell lines using Trizol reagent and reverse-transcribed into cDNA using the PrimeScript RT Master Mix according to the manufacturer’s instructions. For miRNA quantifications, miRNAs were extracted using the miRNeasy Mini Kit (Qiagen, China). Reverse-transcription was performed using the Mir-XTM miRNA First-Strand Synthesis Kit (Clontech, China) according to the manufacturer’s instructions. The mRNA expression levels were determined by qRT-PCR using the Step One Plus Real-Time PCR System (Applied Biosystems, USA) with the FastStart Universal SYBR Green Master according to the manufacturer’s instructions. The expression levels of the YY1 mRNAs and miR-30a were normalized to GAPDH and U6, respectively, using the 2^−ΔΔCT method. Each qRT-PCR experiment was performed in triplicate and independently repeated three times.

Pancreatic cancer cells or xenograft tumor tissues were lysed in ice-cold lysis buffer containing the following reagents: 50 mM Tris–HCl (pH 7.4), 1% NP-40, 150 mM NaCl, 1 mM EDTA, 1 mM PMSF, and a complete protease inhibitor cocktail (1 tablet per 10 mL, Roche Diagnostics GmbH, Mannheim, Germany). The total protein concentration was determined using a BCA Protein Assay kit (Beyotime Biotechnology, China). Western blotting was performed using standard methods. The following dilutions were used for each antibody: anti-YY1 (ab109228, Abcam, 1:3000), anti-Beclin 1 (#669922, R&D, 1:1000), anti-ATG5 (ab108327, Abcam, 1:3000), anti-P62 (ab109012, Abcam, 1:10,000), anti-LC3A/B (#12741, Cell Signaling, 1:1000), and anti-GAPDH (SAM1003, Sun Shine Bio, 1:5000). Each blot was independently repeated three times.

All reagents for fixation, washing, and blocking were purchased from Beyotime (Beyotime Biotechnology, China). Pancreatic cancer cells were washed with PBS three times for 5 min each and then fixed for 10 min. They were permeabilized with 0.1% Triton X-100 in PBS for 10 min. They were then blocked for 1 h at room temperature, followed by incubation with anti-LC3A/B antibody (#12741, Cell Signaling, 1:1000) at 4 °C overnight. After washing with PBS extensively the following day, cells were incubated with Cy3-Labeled Goat Anti-Rabbit IgG (Beyotime Biotechnology, China) for 3 h at room temperature. Chromatin was stained with DAPI for 10 min, and cells were observed under a fluorescence microscope (Nikon, Eclipse 80i, Japan).

Pancreatic cancer cells were washed with PBS twice and fixed in 2.5% glutaraldehyde in PBS at 4 °C overnight. Fixed cells were washed with PBS three times and post-fixed with 1% osmium acid in PBS for 3 h. After washing the cells three times with PBS, the samples were dehydrated in a graded ethanol series (30, 50, 70, 80, 90, 95, and 100%) for about 15–20 min for each incubation. Samples were then transferred to absolute acetone for 20 min. The samples were placed in a 1:1 mixture of absolute acetone, followed by incubation in a Spurr resin mixture for 1 h at room temperature. Samples were incubated in a 1:3 mixture of absolute acetone, followed by the final resin mixture for 3 h and a final Spurr resin mixture overnight. Finally, the specimens were placed in capsules containing embedding medium and heated to 70 °C for about 9 h. Sections 70 nm in thickness were generated using an ultramicrotome (Leica Ultracut R, Germany). Specimen sections were stained with uranyl acetate and lead citrate for 10 min each and observed using a JEM1230 (JEOL, Japan).

A miR-30a mimic, miR-30a inhibitor, and negative control mimic were purchased from Ribobio (Guangzhou, China). Lentiviruses used to knockdown or overexpress human miR-30a, in addition to negative control lentiviruses, were purchased from Hanbio (Shanghai, China). All transfections were carried out according to the manufacturer’s instructions.

The luciferase assay was carried out as previously described [20]. Briefly, dual-luciferase reporter plasmids were constructed by iGeneBio (Guangzhou, China). Transfection of all plasmids was performed using the Lipofectamine 3000 Reagent. Luciferase activity was measured 48 h after transfection using the Dual-Luciferase Reporter Assay Kit (Promega,USA). The assay was repeated at least three times in independent experiments.

ChIP assays were performed as previously described [20] using the EZ ChIP Kit (Merck Millipore, Germany) according to the manufacturer’s instructions. Briefly, cross-linked chromatin was sonicated into fragments ranging from 100 to 1000 base pairs. The chromatin was then immunoprecipitated using a ChIP-grade anti-YY1 antibody (ab12132, Abcam).

All animal research procedures adhered to the guidelines of The Care and Use of Laboratory Animals published by the National Institutes of Health and were approved by the Laboratory of Animal Care and Use Committee of Nanjing Medical University. Four-week-old female nude mice (BALB/cA-nu) were purchased from the Model Animal Research Center of Nanjing University. Twenty mice were randomly divided into four groups. To establish a xenograft subcutaneous implant tumor model, PANC-1-YY1-knockdown and vector control cells, PANC-1-miR-30a-knockdown and vector control cells (1.5 × 10⁶ cells/100 μL per mouse) were injected subcutaneously. The mice were euthanized after 4 weeks, and tumor xenografts were resected.

All data were representative of at least three independent experiments. Statistical analysis was performed using the SPSS software (Version 22.0). Quantitative data are presented as mean ± SD. All statistical tests were two-tailed exact tests with a p < 0.05 considered significant.

During autophagosome formation, the cytosolic microtubule-associated protein 1 light chain 3 (LC3)-I is converted into LC3-II, it is phosphatidylethanolamine- conjugated form. LC3II then targets autophagic membranes [21]. To determine the state of autophagy in pancreatic cancer cells, Western blotting was performed for LC3 II in six pancreatic cancer cell lines, and results were compared to those of the control cell line, HPNE (Fig. 1a). Autophagy was clearly elevated in pancreatic cancer cells, which was reflected by the increased in LC3II. Immunofluorescence further confirmed it (Fig. 1b). Next, to investigate the potential role of YY1 in autophagy, we either overexpressed or silenced YY1 in BxPC-3 and PANC-1 cells. We found that YY1 overexpression increased LC3II levels, while YY1 knockdown decreased LC3II levels. Furthermore, other autophagy-related genes, ATG5 and Beclin-1 (the mammalian homologue of yeast ATG6), displayed similar expression dynamics as LC3II (Fig. 2a). In addition, Western blotting and immunofluorescence showed that, compared to the YY1-treated group, autophagy was significantly reduced following treatment with a combination of YY1 and the autophagy inhibitor, 3-methyladenine (3-MA), suggesting that YY1-induced autophagy is inhibited by 3-MA (Fig. 2b, c). To better assess autophagy, we observed autophagosome formation and double-membrane autophagic vacuoles by transmission electron microscopy in both BxPC-3 and PANC-1 cells (Fig. 3a, b). From these micrographs, we found that overexpression of YY1 increased the number of autophagosomes, while YY1 knockdown decreased the abundance of autophagosomes. Taken together, YY1 likely plays roles in autophagy in pancreatic cancer cells.

Synergies between YY1 and miRNAs have recently been reported in various cellular processes [5, 12, 22]. Therefore, it is possible that autophagy-related miRNAs might be directly regulated by YY1. Previously, we found that YY1 regulates the expression of ATG5 and Beclin 1 proteins (Fig. 2a). Since ATG5 and Beclin 1 are both known to be direct targets of miR-30a, [19, 23] we hypothesized that YY1 and miR-30a might form a functional regulatory circuit that modulates autophagy in pancreatic cancer cells. Using an RT-PCR assay, miR-30a levels were found to decrease when YY1 was overexpressed and increase when YY1 was knocked down in both BxPC-3 and PANC-1 cells (Fig. 4a, b). This suggested that miR-30a might negatively be regulated by YY1. A bioinformatics prediction algorithm (http://​jaspar.​binf.​ku.​dk) revealed three putative YY1 binding sites upstream of the miR-30a genomic loci (Fig. 4c). Using the predicted site sequence that displayed the highest score, we constructed luciferase reporter vectors and performed a dual-luciferase reporter assay. The results showed that overexpression of YY1 significantly decreased luciferase activity for the miR-30a 3′UTR compared to the negative control. In contrast, a mutated version of the miR-30a 3′UTR showed unchanged luciferase activity, demonstrating direct regulatory control of miR-30a (Fig. 4d). To map the YY1 binding sites in the miR-30a promoter under physiological conditions, a ChIP assay was performed. The results revealed that YY1 directly bound the promoter of miR-30a and repressed its expression in BxPC-3 and PANC-1 cells (Fig. 4e). Taken together, these experiments verify miR-30a as a direct target of YY1.

Since miR-30a was found to be a direct target of YY1, we next investigated whether miR-30a mediated the autophagy-related effects of YY1. We first manipulated the expression levels of miR-30a in both BxPC-3 and PANC-1 cells using a miR-30a mimic, a miR-30a inhibitor, and a control miRNA. Using Western blotting and immunofluorescence, we found that miR-30a negatively regulated autophagy in pancreatic cancer cells (Fig. 5a, b). Furthermore, we performed a rescue assay in BxPC-3 and PANC-1 cells consisting of a treatment that combined YY1-overexpressing lentiviruses and miR-30a mimics. As expected, overexpression of miR-30a attenuated the pro-autophagic effects of YY1 (Fig. 5c, d), suggesting that miR-30a is indeed a functional target of YY1.

Since YY1 appears to promote autophagy in pancreatic cancer cells by directly targeting miR-30a, we next explored whether miR-30a modulates YY1 expression, since this type of feedback regulation is commonly observed between transcription factors and miRNAs (such as the Pitx3/miR-133b feedback circuit in midbrain dopamine neurons) [24, 25]. Using various bioinformatics tools, such as Targetscan [26] and RNA22-HAS [27], we found a putative miR-30a binding site in the 3′UTR of YY1 (Fig. 6a). To test whether the predicted YY1 3′UTR binding sites reflect direct regulation by miR-30a, we inserted these sequences downstream of a luciferase reporter gene. We found that a decrease in luciferase activity occurred in the presence of wild type YY1. In contrast, this luciferase inhibition was rescued by mutating the YY1 3′UTR sequences that were predicted to be the miR-30a binding sites (Fig. 6b). We further tested the effect of miR-30a on YY1 mRNA and protein expression using RT-PCR and Western blotting, respectively. This revealed that the expression levels of YY1 were elevated in cells transfected with the miR-30a inhibitor lentivirus and decreased in cells transfected with the miR-30a overexpressing lentivirus (Fig. 6c). Cumulatively, these results suggest that miR-30a acts in a feedback loop to inhibit YY1, thereby regulating autophagy in pancreatic cancer cells.

To assess the effects of YY1-induced autophagy in vivo, we implanted xenograft tumors with alternatively varying levels of both YY1 and miR-30a. The mice were euthanized after 4 weeks, and tumor xenografts were resected and photographed (Fig. 7a). The results showed that knockdown of YY1 increased the expression of miR30a (Fig. 7b). We also quantified ATG5 and Beclin-1 expression levels and overall autophagy activity in the xenograft samples (Fig. 7c). Consistent with the in vitro results, knockdown of YY1 inhibited autophagy through miR30a. On the other hand, when nude mice were subcutaneously injected with PANC-1 cells that had been transfected with either a lentiviral vector that inhibited miR30a or a control lentiviral vector, we found that knockdown of miR30a could also increased the expression of YY1 and autophagy activity (Fig. 7d).