A total of 43 patients with skin melanoma (mean age, 52 years; age range, 39–61 years; 27 male patients and 16 female patients) were enrolled from January 2014 to December 2018 at The Affiliated Changzhou No. 2 People's Hospital with Nanjing Medical University (Changzhou, China) or Weihai Central Hospital (Weihai, China). All patients were diagnosed pathologically by at least 3 experienced and independent pathologists from Weihai Central Hospital or the Affiliated Changzhou No. 2 People's Hospital with Nanjing Medical University. Patients who had radiotherapy or chemotherapy before surgery were excluded. Patients that had not received radiotherapy or chemotherapy prior to sample collection were included. The cases included 19 patients at stage I+II and 24 patients at stage III+IV. Tissue samples excised during the operation were immediately frozen in liquid nitrogen until RNA extraction. Adjacent normal tissue was used as a control. The distance between the resected tumor tissue and the healthy tissues was 2 cm. This study was approved by the ethics committee of The Affiliated Changzhou No. 2 People's Hospital with Nanjing Medical University. The patients signed informed consent forms.

The human epidermal melanocyte HEMa-LP cell lines and the malignant melanoma A375, A2058 and SK-MEL-2 cell lines were purchased from the Cell Bank of Type Culture Collection of the Chinese Academy of Sciences. Cells were cultured in DMEM medium containing 10% fetal bovine serum (Gibco; Thermo Fisher Scientific Inc.) and 1% penicillin-streptomycin. The cells were incubated at 37°C in an incubator with 5% CO₂.

The eukaryotic expression plasmid vector pSilencerTM3.1-H1neo containing DUXAP8-shRNA (5′-GGAACTTCCCAAACCTCCATGATTT-3′) and control shRNA (5′-TTCTCCGAACGTGTCACGTTT-3′) were constructed by Shanghai GenePharma Co., Ltd. The miR-3182 mimic (5′-GCUUCUGUAGUGUAGUC3′-), NC mimic (5′-UUCUCCGAACGUGUCACGUTT-3′), miR-3182 inhibitor (5′-GACUACACUACAGAAGC-3′), NC inhibitor (5′-CAGUACUUUUGUGUAGUACAA-3′) and NUPR1 overexpression vector (oeNUPR1; pcDNA3-NUPR1) were designed and synthesized by Shanghai GenePharma Co., Ltd. Cells were seeded in 6-well plates. pcDNA3 (Invitrogen; Thermo Fisher Scientific Inc.) empty vector was used as the negative control for oeNUPR1. Transfection was performed using Lipofectamine^® 2000 reagent (Invitrogen, Thermo Fisher Scientific, Inc.) at 37°C for 48 h with a 100 nM of the aforementioned plasmids, negative controls, mimics and inhibitors, according to the manufacturer's protocol. Transfection efficiency was measured by RT-qPCR and after 48 h of transfection, subsequent experimentation was performed.

Total RNA was extracted from tissues and cells using TRIzol^® (Invitrogen; Thermo Fisher Scientific, Inc.). Total RNA (500 ng) was reverse transcribed into cDNA using SuperScript III (Invitrogen; Thermo Fisher Scientific, Inc.) at 37°C for 15 min. Subsequently, qPCR was performed using an ABI PRISM 7500 Sequence Detection system (Thermo Fisher Scientific, Inc.) and SYBR Select Master mix (Thermo Fisher Scientific, Inc.). U6 or GAPDH were used as the internal reference genes. The following thermocycling conditions were used for qPCR: i) Initial denaturation at 95°C for 5 min; ii) 40 cycles at 95°C for 20 sec, 58°C for 30 sec and 74°C for 30 sec; and iii) final extension step at 72°C for 5 min. The primer sequences were as follows: i) DUXAP8 forward, 5′-ACCCAAACACTAATTGTAGACT-3′ and reverse, 5′-TGTCTGGGAGACTGCTTACA-3′; miR-3182 forward, 5′-CACTCAGCTGGCTTCTGTAGTG-3′ and reverse, 5′-CTGGTGTCGTGGAGTCG-3′; NUPR1 forward, 5′-AGGACTTATTCCCGCTGACTGA-3′ and reverse, 5′-TGCCGTGCGTGTCTATTTATTG-3′; GAPDH forward, 5′-GTCGATGGCTAGTCGTAGCATCGAT-3′ and reverse, 5′-TGCTAGCTGGCATGCCCGATCGATC-3′; and U6 forward, 5′-CTCGCTTCGGCAGCACATATACT-3′ and reverse, 5′-ACGCTTCACGAATTTGCGTGTC-3′. miRNA and mRNA expression levels were quantified using the 2^−ΔΔCq method (17).

After cell transfection for 48 h, A375 and A2058 cells were seeded in a 96-well cell culture plate at a density of 2×10³ cells/well. After incubation for 24, 48 or 72 h, 10 µl of CCK-8 reagent (Beyotime Institute of Biotechnology) was added to each well and incubated for 4 h. The absorbance at 450 nm was measured using a microplate reader (Bio-Rad Laboratories, Inc.). The average value of each well was taken to evaluate the cell proliferation.

Cell migration and invasion were assessed by a Transwell assay. For the migration assay, 2×10⁴ A375 or A2058 cells were diluted in FBS-free medium and seeded into the upper chamber (BD Biosciences). For the invasion assay, cells were added onto a Matrigel^®-pre-coated (Sigma-Aldrich; Merck KGaA) upper chamber. Subsequently, 500 µl of culture medium containing 10% FBS was inoculated into the lower chamber. After 24 h, non-migratory/-invasive cells were detached while the migrated/invaded cells in the lower chamber were fixed in 4% paraformaldehyde for 30 min at room temperature and stained using 0.5% crystal violet for 30 min at room temperature. The numbers of migratory/invasive cells were counted in five randomly selected fields using an inverted light microscope (Olympus Corporation; magnification, ×200).

Biotin-labeled wild-type (WT) or mutant miR-3182 (Shanghai GenePharma Co., Ltd.) were synthesized. A375 and A2058 cells were lysed using specific lysis buffer (Sigma-Aldrich; Merck KGaA) containing RNase inhibitor (Beyotime Institute of Biotechnology). The lysate was subjected to centrifugation at 12,000 × g for 12 min at 4°C and the supernatant (500 ml per reaction) were incubated with biotin-labeled wild-type or mutant miR-3182, followed by incubation with M-280 streptavidin beads (100 µl; Sigma-Aldrich; Merck KGaA) at 4°C overnight. The beads were washed twice with cold lysis buffer and precipitated RNAs were eluted using elution buffer (Sigma-Aldrich; Merck KGaA) according to the manufacturer's protocol, followed by RT-qPCR analysis as described above. Transfection conditions were the same as described above.

The bioinformatics online tools miRDB (http://mirdb.org/miRDB/index.html) and TargetScan 7 (http://www.targetscan.org/vert_71/) predicted that miR-3182 had targeted binding sites with lncRNA DUXAP8 and NUPR1. The 3′UTR of DUXAP8 and NUPR1 containing binding site sequence was amplified and constructed into psiCHECK2 luciferase reporter vector (Promega Corporation), referred to as DUXAP8-WT and NUPR1-WT reporter plasmids, respectively. The mutant (MUT) sequence was constructed on the reporter gene vector and recorded as DUXAP8-MUT and NUPR1-MUT reporter plasmids. A375 and A2058 cells were transfected with indicated luciferase reporter and miR-3182 mimic (5′-GCUUCUGUAGUGUAGUC3′-) or NC mimic (5′-UUCUCCGAACGUGUCACGUTT-3′) (Shanghai GenePharma Co., Ltd.) using Lipofectamine^® 2000 reagent (Invitrogen, Thermo Fisher Scientific, Inc.) for 48 h and then luciferase reporter assay was performed using a Dual Luciferase Reporter Assay kit (Promega Corporation) according to according to the manufacturer's protocol. Dual luciferase activity was measured separately, and the relative luciferase activity of each cell was calculated based on the fluorescence intensity and normalized to Renilla luciferase activity.

Statistical analysis was carried out using SPSS 22.0 software (IBM Corp.). Continuous data ware expressed as the mean ± standard deviation. Differences between groups were compared using an unpaired Student's t-test or one-way ANOVA followed by Tukey's post hoc test. The prognosis of patients was analyzed using the Kaplan-Meier method and log-rank tests. The relationship between DUXAP8 expression and miR-3182 or NUPR1 levels was analyzed using Pearson's correlation coefficient. P<0.05 was considered to indicate a statistically significant difference.

The expression levels of lncRNA DUXAP8 were examined in melanoma tissue samples and cell lines. DUXAP8 was upregulated in melanoma tissue compared with adjacent normal tissue (Fig. 1A). Moreover, patients with stage III+IV melanoma displayed higher levels of DUXAP8 than patients in stage I–II (Fig. 1B). Similar results were also observed in cell lines. The expression of DUXAP8 in malignant melanoma cell lines (A375, A2058 and SK-MEL-2) was higher than that in HEMa-LP cells (Fig. 1C). A375 and A2058 cell lines displayed the highest levels of DUXAP8. Therefore, these two cell lines were used in subsequent experiments. In addition, the patients were divided into two groups based on the median value of DUXAP8 expression, and patients with higher expression of DUXAP8 experienced shorter survival time (Fig. 1D). High DUXAP8 levels were associated with poor prognosis.

The expression of DUXAP8 was significantly reduced following transfection with sh-DUXAP8, which confirmed the transfection was successful (Fig. 2A). In order to examine the effect of sh-DUXAP8 on proliferation, CCK-8 assays were carried out. After transfection for 48 h and further culturing for 72 h, the proliferation of A375 and A2058 cells significantly decreased in the sh-DUXAP8 group compared with that of the sh-NC group (Fig. 2B). In addition, DUXAP8 knockdown significantly decreased migration and invasion of melanoma cells (Fig. 2C and D). Therefore, DUXAP8 knockdown inhibited the proliferation, migration and invasion of melanoma cells.

To identify the potential targets of DUXAP8, bioinformatics analysis was carried out using miRDB. miR-3182 was identified as the most likely target according to the prediction results (Fig. 3A). To validate the interaction between DUXAP8 and miR-3182, a dual luciferase reporter assay was carried out. Both in A375 and A2058 cell lines, transfection of miR-3182 reduced the luciferase activity in the DUXAP8-WT group. However, NC miRNA did not significantly reduce the luciferase activity of the vector. In the DUXAP8-MUT group, luciferase activity remained unchanged. These results indicated that miR-3182 significantly downregulated luciferase activity (Fig. 3B and C). Pull-down assays also demonstrated that miR-3182-WT precipitated DUXAP8, proving their binding (Fig. 3D). Following DUXAP8 knockdown, the expression of miR-3182 significantly increased (Fig. 3E). To identify the targets of miR-3182, bioinformatics analysis was performed using TargetScan7. The results indicated that NUPR1 was the most likely candidate. The predicted binding sites of miR-3182 and NUPR1 are shown in Fig. 3F. Moreover, luciferase reporter assays confirmed this interaction (Fig. 3G and H). Transfection with miR-3182 downregulated the expression of NUPR1 in both A375 and A2058 cells (Fig. 3I). More importantly, sh-DUXAP8 decreased the expression of NUPR1, while the miR-3182 inhibitor reversed this effect (Fig. 3J and K). Thus, DUXAP8 promotes NUPR1 expression by inhibiting miR-3182.

For rescue assays, NUPR1 was overexpressed in A375 cells and A2058 cells (Fig. 1A). CCK8 and transwell assays demonstrated that the cell proliferation, migration and invasion were inhibited in the sh-DUXAP8+miR-NC+pcDNA3-vector group compared with the sh-NC+miR-NC+pcDNA3-vector group (Fig. 4A-C). However, proliferation, migration and invasion were rescued by NUPR1 overexpression in sh-DUXAP8+miR-NC+oeNUPR1 group compared with that in the sh-DUXAP8+miR-NC+pcDNA3-vector group (Fig. 4A-C). Proliferation, migration and invasion were suppressed by miR-3182 in the sh-NC+miR-3182+pcDNA3-vector group compared with the sh-NC+miR-NC+pcDNA3-vector group (Fig. 4A-C). Proliferation, migration and invasion were also rescued by NUPR1 overexpression in sh-NC+miR-3182+oeNUPR1 group compared with the sh-NC+miR-3182+pcDNA3-vector group (Fig. 4A-C). In melanoma tissue samples, DUXAP8 expression negatively correlated with that of miR-3182, but positively correlated with that of NUPR1 (Fig. 4D and E). Altogether, these results confirmed the regulatory relationship between DUXAP8, miR-3182 and NUPR1.