The normal skin tissues and melanoma tissues of 50 patients with melanoma in our hospital were selected. The ethics committee of The First People’s Hospital of Lianyungang approved our study. Signed written informed consents were obtained from all participants before the study.

Melanoma cells (SKMEL1, A375, A2058 and A875) and normal human epidermal melanocytes (NHEM) were provided by Cell Bank of Type Culture Collection (Shanghai, China). Cells were cultivated in DMEM containing 10% fetal bovine serum (FBS), 100 U/mL penicillin and 100 μg/ml streptomycin at 5% CO₂, 37°C. Cell transfection was performed using Lipofectamine 3,000 as per the protocols. Cell proliferation was determined by EdU (Beyotime, Shanghai, China).

RNAs isolation was done with TRIzol and were then reversely transcribed into cDNAs. U6 and GAPDH were used as the internal controls with the method of 2^−ΔΔCt. Primers used were shown below: hsa_circ_0062270: Forward: 5′-AGG​ATG​GCT​CAG​GGA​CAG​AT-3′, reverse: 5′-AGG​CCA​TGG​TAC​AGC​TTG​TC-3′; CDC45: Forward: 5′-TTC​GTG​TCC​GAT​TTC​CGC​AAA-3′, reverse: 5′-TGG​AAC​CAG​CGT​ATA​TTG​CAC-3′; GAPDH: Forward: 5′-CGG​AGT​CAA​CGG​ATT​TGG​TCG​TAT-3′, reverse: 5′-AGC​CTT​CTC​CAT​GGT​GGT​GAA​GAC-3′: U6: Forward: 5′-GCT​GAG​GTG​ACG​GTC​TCA​AA-3′, reverse: 5′-GCC​TCC​CAG​TTT​CAT​GGA​CA-3′.

A375 cells were exposed to Actinomycin D (3 μg/ml). They were collected for isolating total RNAs. Expressions of hsa_circ_0062270 and CDC45 were detected by Quantitative real-time polymerase chain reaction. Cellular RNA (4 mg) was treated either with RNase R (10 U/μg) at 37°C for 30 min or not, followed by purification using RNeasy MinElute (Qiagen, Hilden, Germany).

Cells were seeded into the top chamber and bottom chamber. After 48-h incubation, cells in the bottom were fixed, dyed in crystal violet and captured. Migratory cells were counted in five randomly selected fields per sample. Invasion assay was conducted using transwell chamber precoated with 100 μL of Matrigel (Corning, Corning, NY, United States). In detail, Matrigel was diluted in serum-free medium at 1:3, which was coated on the top of a chamber.

Data were expressed as mean ± SD (standard deviation) and they were processed using Statistical Product and Service Solutions (SPSS) 20.0 (IBM, Armonk, NY, United States). Prognostic value of hsa_circ_0062270 in melanoma were evaluated by Kaplan-Meier and receiver operating characteristic (ROC) method, respectively. The correlation between hsa_circ_0062270 and CDC45 levels was assessed through Pearson correlation test. A significant difference was set at p < 0.05.

Firstly, we explored the expression of hsa_circ_0062270 in melanoma tissues and the normal skin tissues. Results revealed that the expression hsa_circ_0062270 in melanoma tissues was significantly higher than that in normal ones (Figure 1A). Then, receiver operating characteristic (ROC) curves depicted that the AUC of hsa_circ_0062270 was 0.6312 (Figure 1B). Kaplan-Meier analysis found that highly expressed hsa_circ_0062270 predicted a poor prognosis for melanoma patients (p = 0.0234) (Figure 1C). Hsa_circ_0062270 significantly increased in melanoma cell lines as well (Figure 1D). We furthermore examined the stability of hsa_circ_0062270 by detecting mRNA levels of hsa_circ_0062270 and CDC45 in Actinomycin D-treated A375 cells. Compared with that of CDC45 (<12 h), the half-life of hsa_circ_0062270 was over 24 h (Figure 1E). RNase R induction did not affect expression level of hsa_circ_0062270, but markedly down-regulated CDC54 (Figure 1F). Therefore, we have verified that hsa_circ_0062270 was highly stable in melanoma cells.

To investigate the effects of hsa_circ_0062270 on melanoma cell proliferation, migration and invasion, cells were treated with hsa_circ_0062270 siRNA. Results indicated that transfection of hsa_circ_0062270 siRNA markedly down-regulated hsa_circ_0062270 level in A375 and A2058 cells (Figure 2A). Knockdown of hsa_circ_0062270 reduced EdU-positive rate in melanoma cells (Figure 2B). In addition, hsa_circ_0062270 siRNA markedly reduced migratory and invasive rates (Figures 2C,D).

Then we focused on the potential target of hsa_circ_0062270 in the regulation of phenotypes of melanoma. CircRNAs are involved in pathological process via mediating expression levels of their linear transcripts. CDC45 was the linear transcript of hsa_circ_0062270 (Figure 3A) and positively correlated to hsa_circ_0062270 level (Figure 3B). Identically, CDC45 was highly expressed in melanoma cell lines (Figure 3C). Knockdown of hsa_circ_0062270 could downregulate CDC45 and as expected, CDC45 was up-regulated in A375 and A2058 cells overexpressing hsa_circ_0062270 (Figures 3D,E). It is concluded that hsa_circ_0062270 could stabilize the expression of CDC45.

To further elucidate the effects of CDC45 on melanoma phenotypes, we established the overexpression models of CDC45. Transfection of overexpressed plasmid of CDC45 effectively up-regulated CDC45 in A375 and A2058 cells (Figure 4A). In melanoma cells overexpressing CDC45, EdU-positive rate increased (Figure 4B). Moreover, migratory and invasive potentials of melanoma were promoted by overexpressed CDC45 (Figures 4C,D). Therefore, CDC45 could stimulate the malignant process of melanoma.

To uncover the co-regulation of hsa_circ_0062270 and CDC45, cells were co-transfected using si-CDC45 and overexpressed plasmid of hsa_circ_0062270 (Figure 5A). Overexpression of hsa_circ_0062270 could enhance the down-regulated CDC45 in melanoma cells transfected with si-CDC45. Knockdown of CDC45 reduced EdU-positive rate, migratory cell number and invasive cell number, which were reversed by overexpressed hsa_circ_0062270 (Figures 5B–D). All above indicated that hsa_circ_0062270 may play a carcinogenic role in melanoma by stabilizing its linear transcription CDC45.