Blood plasma samples from 42 patients with AML and 42 healthy controls (25 men and 17 women; mean age, 44.54 years; range, 19-76 years) were collected from The First Affiliated Hospital of Kunming Medical University (Kunming, China) between April 2021 and March 2022. The AML group consisted of 22 men and 20 women, with a mean age of 46.62 years (range, 16-80 years). The criteria for the selection of patients in this study included: i) Male or female patients with a body weight ≥45 kg; ii) no bleeding tendency; iii) no obvious abnormalities of heart, liver and kidney function; iv) expected survival, ≥3 months; v) women of childbearing age must have had a pregnancy test (serum or urine) within 7 days of enrollment, with a negative result, and be willing to use an appropriate method of contraception during the trial. Patients with other serious diseases or who started treatment within 3 months before admission were excluded, and those in the healthy control group had normal physiological function according to whole-body physiological examinations conducted at the aforementioned hospital. The diagnosis of AML was made according to the 2016 World Health Organization criteria (41). The basic information for all patients is shown in Table I. The present study was approved by the Ethics Committee of The First Affiliated Hospital of Kunming Medical University (approval no. L-10), and all individuals provided written informed consent for participation.

The following cell lines were used in the present study: A normal bone marrow stromal cell line (HS-5) and human AML cell lines (U937, THP-1, HL-60 and NB4), which were purchased from Shenzhen Haodi Huatuo Biotechnology Co., Ltd. All cells were cultured in DMEM (MilliporeSigma) containing 10% FBS (Gibco; Thermo Fisher Scientific, Inc.) and 1% penicillin/streptomycin (MilliporeSigma), and were incubated at 37°C in a 5% CO₂ incubator. The cells were subcultured when the cell density reached 90%. The cell lines used in the present study were authenticated by short tandem repeat profiling.

The circPVT1 promoter region was identified by searching the UCSC database (http://genome.ucsc.edu/). Specifically, GRCh38/hg38 was used as the human reference genome. CircPVT1 (hsa_circ_0085536) was located at chromosome chr8:128806778-128903244, and the sequence 2,000 bp upstream of its transcription start site was selected as the promoter region. MethPrimer (http://www.urogene.org/methprimer/) (42) was used to analyze the CpG island regions in the circPVT1 promoter sequence. Specifically, the minimum CpG island length was set at 200 bp, the CpG percentage threshold was 50%, and the expected CpG island ratio threshold was 0.6.

The circPVT1 promoter sequence was identified as aforementioned. JASPAR (https://jaspar.genereg.net/) was used to analyze the transcription factors to which the circPVT1 promoter sequence might bind. The relative profile score threshold was set to 80%.

NB4 or HL-60 cells were seeded at a density of 2×10⁵ cells/well, and were transfected with 50 nM miR-455-3p mimic, miR-455-3p inhibitor, small interfering RNA (siRNA/si)-PVT1, si-MCL1 and 2 μg pcDNA3.1-MEF2A vector (oe-MEF2A) using Lipofectamine^® 2000 (Invitrogen; Thermo Fisher Scientific, Inc.). Negative control (NC) mimic, NC inhibitor, NC-small interfering siRNA (NC-si) and pcDNA3.1 empty vector (NC-oe) were used as NCs. The cells were cultured at 37°C for 48 h. All oligonucleotides (Table SI) were designed and synthesized by Shanghai GenePharma Co., Ltd. The transfection efficiency was measured 48 h after transfection.

The biological information database StarBase (http://starbase.sysu.edu.cn/) was used to analyze the binding sites of circPVT1 and miR-455-3p or miR-455-3p and MCL1.

The circPVT1 and MCL1 3′UTRs containing the miR-455-3p binding site were cloned and inserted into the pGL3-Basic vector (Promega Corporation) to construct the PVT1-wild-type (WT) vector and the MCL1-WT vector, respectively. The 3′UTRs of circPVT1 and MCL1 containing mutant (MUT) binding sequences of miR-455-3p were also amplified and inserted into the pGL3-Basic vector downstream of the firefly luciferase gene to generate the PVT1-MUT vector and the MCL1-MUT vector. Subsequently, 293 T cells (Thermo Fisher Scientific, Inc.) were seeded in 24-well plates at a density of 1.0×10⁵/ml per well. Subsequently, the WT or MUT vectors were cotransfected with the miR-455-3p mimic/inhibitor or NC mimic/inhibitor into 293T cells using Lipofectamine 2000. A total of 48 h after transfection, luciferase activity was measured using the dual-luciferase reporter gene assay system (Beyotime Institute of Biotechnology). In addition, pGL3-Basic-circPVT1 and pGL3-Basic-MEF2A vectors were transfected into NB4 cells using Lipofectamine 2000, followed by assessment using the aforementioned luciferase assay kit at 48 h post-transfection. Firefly luciferase activity was normalized to Renilla luciferase activity.

A TRIzol^® RNA extraction kit (Invitrogen; Thermo Fisher Scientific, Inc.) was used to extract total RNA from AML cells or the blood plasma of patients with AML. First-strand cDNA was subsequently assembled using total RNA as a template with the PrimeScript RT Reagent kit (Takara Bio, Inc.) according to the manufacturer's protocol, and the cDNA obtained was used as a template for qPCR amplification. qPCR was performed according to the instructions of the SYBR Green qPCR kit (Thermo Fisher Scientific, Inc.), and the qPCR thermocycling conditions were 94°C for 2 min; followed by 45 cycles at 94°C for 15 sec and 60°C for 30 sec. GAPDH, β-actin and U6 were used as internal controls for circRNAs, mRNAs and miRNAs, respectively. Moreover, sterilized deionized water was used as a negative control for PCR in place of a nucleic acid template. The expression levels of genes were calculated using the 2^−ΔΔCq method (43). The primer sequences are displayed in Table II.

Agarose powder was dissolved in TBE buffer in a flask to prepare 2.5% agarose gel. After the agarose powder was completely dissolved, ethidium bromide was added and then poured into the gel-making plate. The PCR product of circPVT1 was then mixed with the loading buffer and applied to the gel, followed by electrophoresis; finally, the results were observed using a gel imaging system (Tanon Science & Technology Co., Ltd.).

NB4 cells were transfected with a NC mimic or inhibitor, and miR-455-3p mimic or inhibitor for 24 h, and a RIP experiment was performed using the Magna RIP RNA-Binding Protein Immunoprecipitation Kit (cat. no. 17-700; MilliporeSigma) according to previously described methods (44). RT-qPCR was performed as aforementioned to assess circPVT1 and MCL1 levels in Ago2 or IgG (negative control) immunoprecipitates.

RNase R (10 U/μl; Guangzhou Geneseed Biotech. Co., Ltd.) was added to the experimental samples (2.5 μg total RNA extracted from NB4 cells) at 37°C for 30 min. RT-qPCR was subsequently performed as aforementioned to verify the stability of circPVT1.

NB4 and HL-60 cell viability was assessed using a CCK-8 kit (Beyotime Institute of Biotechnology). The cells were added to 96-well plates at a concentration of 3×10³ cells/well and an equal amount of medium was added to each well. Subsequently, 10 μl CCK-8 solution was added to each well and the plates were incubated at 37°C and 5% CO₂ for 2 h. The absorbance was measured at a wavelength of 450 nm using a microplate reader. As a blank control, CCK-8 solution was added to cell-free medium and the absorbance was measured at 450 nm.

The cells from each group were collected and washed twice with precooled PBS. According to the manufacturer's protocol, the percentage of apoptotic cells was detected using an Annexin-V-FITC/PI apoptosis kit (Absin Bioscience, Inc.). Cells that were not stained with FITC or PI were used as negative controls. Cells were analyzed using a flow cytometer (FACSCalibur; BD Biosciences) and FlowJo-V10 software (FlowJo, LLC) was used to process the experimental results.

Total proteins were extracted from NB4 and HL-60 cells using RIPA lysis buffer (Beijing Solarbio Science & Technology Co., Ltd.), and the protein concentration was determined using a BCA kit (Beijing Solarbio Science & Technology Co., Ltd.). The proteins (40 μg/lane) were subsequently separated by SDS-PAGE on 10% gels and were transferred to PVDF membranes. The membranes were blocked with 5% skim milk for 2 h at room temperature, and were then incubated with primary antibodies (Abcam) against MEF2A (1:1,000; cat. no. ab76063), caspase-3 (1:2,000; cat. no. ab184787), Bax (1:1,000; cat. no. ab32503), Bcl-2 (1:2,000; cat. no. ab182858), E-cadherin (1:1,000; cat. no. ab231303), N-cadherin (1:1,000; cat. no. ab245117), vimentin (1:1,000; cat. no. ab92547), fibronectin (1:1,000; cat. no. ab268020), and MCL1 (1:500; cat. no. ab243136) overnight at 4°C. GAPDH (1:1,000; cat. no. ab181602) was used as the internal reference protein. The membrane was subsequently incubated with an HRP-conjugated secondary antibody (1:2,000; cat. no. ab288151; Abcam) for 1 h at room temperature. The protein bands were visualized with an enhanced chemiluminescence detection kit (BD Biosciences), and ImageJ 1.8.0.345 software (National Institutes of Health) was used for protein band gray value analysis. In addition, nuclear and cytoplasmic proteins were isolated from NB4 and HL-60 cells using a Nuclear and Cytoplasmic Protein Extraction Kit (Beyotime Institute of Biotechnology) according to the manufacturer's instructions. H3 (1:1,000; cat. no. ab1791; Abcam) was used as the internal reference for nuclear proteins and western blotting was performed as aforementioned.

The cell experiments were repeated three times. All experimental data are presented as the mean ± standard deviation. GraphPad Prism 7 (Dotmatics) was used to analyze and plot the data. For normally distributed and homogeneous variance variables, unpaired Student's t-test was used for comparisons between two groups, and one-way ANOVA was used for comparisons between multiple groups, followed by Tukey's multiple comparisons test. The Mann-Whitney U test was used for nonparametric data. The correlations between circPVT1 and miR-455-3p, miR-455-3p and MCL1, or circPVT1 and MCL1 were assessed by Pearson correlation analysis. P<0.05 was considered to indicate a statistically significant difference.

To explore the role of MEF2A as a transcription factor of circPVT1 in the malignant development of AML, the promoter region of circPVT1 was first analyzed. As shown in Fig. 1A, the promoter region of circPVT1 was enriched in CpG islands, forming an obvious CG enrichment phenomenon, and three CpG islands were identified via Meth Primer analysis. The sizes of CpG islands 1, 2 and 3 were 217, 183 and 368 bp, respectively. CpG islands often appear in the promoter region and can affect the affinity of the promoter for transcription factors. The binding sites of the circPVT1 promoter were then analyzed, and it was revealed that 34 sequences could bind to the circPVT1 promoter (Table III). It was revealed that the transcription factor MEF2A can bind to the circPVT1 promoter region with high affinity. Subsequently, the pGL3-Basic-circPVT1 and pGL3-Basic-MEF2A eukaryotic expression vectors were cotransfected into NB4 cells. The empty vector pGL3-Basic was used as a control, and the luciferase activity was detected after transfection for 24 h. Compared with the pGL3-Basic-circPVT1 group, the cells cotransfected with pGL3-Basic-circPVT1 and pGL3-Basic-MEF2A group showed a significant increase in luciferase activity (Fig. 1B). These results indicated that the MEF2A gene can significantly increase circPVT1 promoter activity.

Evaluation by RT-qPCR revealed that the expression levels of circPVT1 were markedly elevated in the AML group compared with those in the healthy control group (Fig. 2A). These findings suggested that circPVT1 may be highly expressed in the blood plasma of patients with AML. Moreover, the expression levels of circPVT1 were evaluated in different AML cell lines and it was demonstrated that circPVT1 expression was greater in AML cell lines (U937, THP-1, HL-60 and NB4) compared with that in HS-5 cells; this effect was particularly pronounced in NB4 and HL-60 cells (Fig. 2B). Therefore, NB4 and HL-60 cells were selected for subsequent experiments. In addition, data analysis on circPVT1 we performed using the UCSC Genome Browser and it was revealed that circPVT1 was circularized from the PVT1 gene and located on chr8:128806778-128903244 (Fig. 2C). The length of the spliced mature sequence of circPVT1 was 777 bp according to agarose gel electrophoresis (Fig. 2D). Finally, RNase R, a 3′ exonuclease that has no effect on circRNAs (45), was added to the experimental samples to verify the stability of circPVT1. The RT-qPCR results revealed that RNase R had no significant effect on the expression levels of circPVT1. However, linear PVT1 was digested by RNase R (Fig. 2E). These results demonstrated that circPVT1 has greater stability.

To investigate the effect of MEF2A on AML cells, NB4 and HL-60 cells were transfected with the MEF2A overexpression vector (oe-MEF2A), with NC-oe as the control. The results of western blotting revealed that the expression levels of MEF2A showed a clear upward trend in cells post-transfection with oe-MEF2A, indicating successful overexpression of MEF2A (Fig. 3A). After successful transfection, the results of RT-qPCR revealed that, compared with in the NC group, the overexpression of MEF2A promoted the expression of circPVT1 in the cells (Fig. 3B). Furthermore, the results of the CCK-8 assay revealed that overexpression of MEF2A significantly increased the viability of NB4 and HL-60 cells (Fig. 3C). Next, the apoptosis of NB4 and HL-60 cells after oe-MEF2A transfection was detected by flow cytometry. Compared with that in the NC group, apoptosis was significantly inhibited after MEF2A overexpression (Fig. 3D). Moreover, western blotting was used to detect the expression of apoptosis-related proteins, and MEF2A overexpression inhibited the expression of the proapoptotic proteins caspase-3 and Bax, and promoted the expression levels of the antiapoptotic protein Bcl-2 in AML cells (Fig. 3E), which was consistent with the results of flow cytometry. In summary, MEF2A overexpression markedly inhibited the apoptosis of AML cells. In addition, EMT serves a key role in tumor progression by allowing cells to metastasize and invade (46). Western blotting was used to detect the expression levels of EMT marker proteins. It was revealed that the expression levels of the epithelial marker E-cadherin were downregulated, whereas the expression levels of the mesenchymal markers N-cadherin, vimentin and fibronectin were upregulated after MEF2A was overexpressed in NB4 and HL-60 cell lines, indicating that MEF2A overexpression promoted the EMT process in NB4 and HL-60 cells (Fig. 3F). Taken together, these findings suggested that MEF2A may promote the malignant progression of AML cells through circPVT1.

To explore the role of circPVT1 in AML, the biological information database StarBase was used to analyze circPVT1 target genes, and it was revealed that circPVT1 could bind to miR-455-3p (Fig. 4A). RT-qPCR was used to detect the transfection efficiency of the miR-455-3p mimic. The results showed that transfection with the miR-455-3p mimic significantly increased the expression levels of miR-455-3p in cells compared with those in the NC mimic group, indicating successful transfection (Fig. 4B). Moreover, a dual-luciferase reporter gene assay was used to verify the relationship between circPVT1 and miR-455-3p. The results revealed that the miR-455-3p mimic or the miR-455-3p inhibitor significantly affected the luciferase activity of the PVT1-WT vector, whereas there was no significant effect on the luciferase activity of the PVT1-MUT vector (Fig. 4C). Ago2-RIP assay was further used to verify the binding relationship between circPVT1 and miR-455-3p, and the IgG group was used as a negative control. The results revealed that circPVT1 was enriched by the miR-455-3p mimic, whereas the miR-455-3p inhibitor did not enrich circPVT1 (Fig. 4D). These results indicated that circPVT1 may have a targeted binding relationship with miR-455-3p. In addition, after evaluation by RT-qPCR, it was revealed that the expression levels of miR-455-3p showed a clear downward trend in the AML group compared with those in the healthy control group (Fig. 4E). Pearson correlation analysis revealed that miR-455-3p and circPVT1 were negatively correlated; however, the R-value was <-0.3, indicating that the relationship was weak (Fig. 4F). Thus, these results identified an interaction between miR-455-3p and circPVT1, and indicated that circPVT1 may negatively regulate miR-455-3p.

The present study next explored how circPVT1 affects the viability and EMT of AML cells through miR-455-3p. Cells were transfected with the miR-455-3p inhibitor and/or si-circPVT1, and the transfection efficiency was detected by RT-qPCR. The results revealed that transfection with the miR-455-3p inhibitor significantly decreased the expression levels of miR-455-3p in cells, and knockdown of circPVT1 significantly reduced circPVT1 expression levels compared with those in the NC inhibitor or NC-si groups, thus indicating successful transfection (Fig. 5A). The NC group in Fig. 5B-E refers to NB4 and HL-60 cells without any treatment. Next, cell viability was detected by a CCK-8 assay. The results revealed that circPVT1 knockdown significantly inhibited the viability of NB4 and HL-60 cells, whereas miR-455-3p inhibitor transfection weakened the effect of circPVT1 knockdown and promoted cell viability (Fig. 5B). Moreover, flow cytometry revealed that si-circPVT1 promoted apoptosis, whereas the miR-455-3p inhibitor weakened the effect of si-circPVT1 and inhibited apoptosis (Fig. 5C). Western blot analysis of apoptosis-related proteins in NB4 and HL-60 cells revealed that cotransfection with si-circPVT1 and the miR-455-3p inhibitor reversed the promoting effect of si-PVT1 on caspase-3 and Bax, and its inhibitory effect on Bcl-2 (Fig. 5D). In addition, the expression of EMT-related proteins was detected by western blotting, and it was demonstrated that the expression levels of E-cadherin were increased, whereas those of N-cadherin, vimentin and fibronectin were decreased after circPVT1 was knocked down. However, after transfection with the miR-455-3p inhibitor, the opposite results were observed and the effects of si-circPVT1 were markedly reversed (Fig. 5E). These data suggested that circPVT1 knockdown may promote apoptosis and inhibit the progression of EMT, whereas miR-455-3p knockdown could alleviate the effects of circPVT1 knockdown, thereby inhibiting apoptosis and promoting the progression of EMT. In summary, miR-455-3p is associated with the regulatory effects of circPVT1 on AML cell viability and EMT.

MCL1 has been shown to be upregulated in patients with AML, and it has been suggested that MCL1 is associated with poor prognosis in patients with AML and could be used for disease surveillance (47). In addition, Lu et al (48) revealed that miR-181b could improve the drug resistance of AML cells by decreasing the levels of MCL1. Therefore, it was hypothesized that MCL1 may be a target of miR-455-3p, and that miR-455-3p may regulate the progression of AML by targeting MCL1. First, the expression of MCL1 in the nucleus and cytoplasm of NB4 and HL-60 cells was detected by western blotting. The results indicated that MCL1 was mainly expressed in the nucleus (Fig. 6A). StarBase predicted that MCL1 would be targeted by miR-455-3p (Fig. 6B). Moreover, dual-luciferase reporter gene and Ago2-RIP assays were used to verify the targeting relationship between miR-455-3p and MCL1. The results of the dual-luciferase reporter gene assay revealed that the miR-455-3p mimic reduced the luciferase activity of the MCL1-WT vector but had no significant effect on the luciferase activity of the MCL1-MUT vector (Fig. 6C). The Ago2-RIP assay results revealed that the miR-455-3p mimic significantly enriched MCL1; the IgG group was used as a negative control (Fig. 6D). These results indicated that miR-455-3p may interact with MCL1. In addition, after evaluation by RT-qPCR, the results revealed that the expression levels of MCL1 exhibited a clear upward trend in the plasma of patients with AML compared with in those from healthy controls (Fig. 6E). The relationship between miR-455-3p and MCL1 was then analyzed by correlation analysis, which revealed that miR-455-3p was negatively correlated with MCL1 (Fig. 6F). In addition, MCL1 was positively correlated with circPVT1 (Fig. 6G). These results suggested that MCL1 may be a target gene of miR-455-3p and could serve an important role in circPVT1/miR-455-3p-mediated AML.

The present study further explored how circPVT1 influences the malignant progression of AML through the miR-455-3p/MCL1 axis. Following the observation that knockdown of circPVT1 and miR-455-3p affected the viability and EMT of NB4 and HL-60 cells, the effects of MCL1 knockdown on these cells were further explored. First, si-MCL1 was transfected into the cells, and the transfection efficiency was detected by western blotting. The expression levels of MCL1 in the si-MCL1 group were significantly lower than those in the NC-si group, indicating successful transfection (Fig. 7A and B). The NC group in Fig. 7C-F refers to NB4 and HL-60 cells without any treatment. The CCK-8 assay was used to assess the effects on cell viability, and it was revealed that knockdown of MCL1 weakened the effect of the miR-455-3p inhibitor, and inhibited the viability of NB4 and HL-60 cells (Fig. 7C). Moreover, flow cytometry was used to detect apoptosis, and the results showed that the knockdown of MCL1 promoted apoptosis compared with that in the si-circPVT1 + miR-455-3p inhibitor group (Fig. 7D). Furthermore, western blotting was used to detect the expression levels of apoptosis-related proteins. Notably, si-MCL1 partially reversed the effect of the miR-455-3p inhibitor, and promoted the expression of Caspase-3 and Bax, and inhibited the expression of Bcl-2 (Fig. 7E). In addition, the expression levels of EMT-related markers were detected by western blotting. The results showed that si-MCL1 promoted the expression levels of E-cadherin, and inhibited the expression levels of N-cadherin, vimentin and fibronectin compared with those in the si-circPVT1 + miR-455-3p inhibitor group (Fig. 7F). These data indicated that further knockdown of MCL1 may weaken the effects of the miR-455-3p inhibitor, promote the apoptosis of AML cells, and inhibit cell viability and the EMT process. These findings suggested that circPVT1 could influence the malignant progression of AML through the miR-455-3p/MCL1 axis.