Bioactive peptide inhibits acute myeloid leukemia cell proliferation by downregulating ALKBH5-mediated m⁶A demethylation of EIF4EBP1 and MLST8 mRNA

MV4-11, HL-60 and HEK293T cell lines were obtained from the American Type Culture Collection (ATCC). All cells were cultured according to the instructions from the ATCC. The leukemia cell lines were grown in RPMI-1640 medium supplemented with 10% FBS (Gibco), 100 U/ml penicillin/streptomycin (Invitrogen) and 2 mM L-glutamine (Invitrogen). HEK293T cells were grown in DMEM (Invitrogen) supplemented with 10% FBS (Gibco) and 100 U/ml penicillin/streptomycin (Invitrogen).

All animal procedures and studies were approved by the ethics committee of the Inner Mongolia Medical University. Five- to six-week-old female CB17 SCID mice were purchased from Beijing Weitonglihua Company (Beijing, China). HL-60 cells (1 × 10⁷) suspended in 0.1 ml PBS containing 50% Matrigel were subcutaneously injected into the flanks of the mice. When tumor sizes reached 200 mm³, the mice were randomly distributed into four groups with the indicated dosages of saline, cytarabine and BP alone or in combination. For BP injections, the solution was delivered intraperitoneally at 106 µg/kg body weight for the first 8 consecutive days. For cytarabine injections, the solution was delivered intraperitoneally at 100 mg/kg body weight three times (once every three days). The combination group was administered intraperitoneally three times (once every three days) with the same dosages as described above. The control group was treated with an equivalent amount of saline. All xenograft tumors were allowed to reach an endpoint volume of 1200 mm³.

Wild-type ALKBH5, H204A mutant ALKBH5, MLST8 and EIF4EBP1 expression plasmids were generated by PCR subcloning of the human ALKBH5 coding sequence into pcDNA3.1 plasmids. DNA fragments of MLST8-3’UTR and EIF4EBP1-3’UTR containing wild-type m⁶A motifs were directly synthesized and inserted downstream of firefly luciferase of a pMIR-Report vector. HEK293T cells were transfected with Lipofectamine 3000 (Thermo Scientific). Ingenio® Electroporation Kits and Solutions were used for electroporation of DNA into MV4-11 and HL-60 cells using Bio-Rad Gene Pulser® electroporation instruments according to the manufacturer’s instructions.

Cells were collected and seeded at the requested concentrations and then treated with the corresponding drugs for a period of 24 h, after which apoptosis and cell viability assays were carried out using an ApoLive-Glo Multiplex Assay Kit (Promega) according to the manufacturer’s instructions. For the proliferation assay, 5000–10,000 cells/well were seeded in 96-well plates in triplicate and then treated with the corresponding drugs for 24, 48, 72 and 96 h. Cell proliferation was assessed by MTT (Promega) following the manufacturer’s instructions. Optical densities (OD) were detected at 570 nm using a SpectraMax i3X Multimode Detection Platform of Molecular Devices once per day for 5 days.

Proteins were separated on denaturing SDS-PAGE gels and transferred onto PVDF membranes. Next, the membranes were blocked with 5% non-fat milk for 1 h at room temperature and incubated overnight at 4 °C with anti-ALKBH5 (ab195377, Abcam), anti-EIF4EBP1 (ab32024, Abcam), anti-MLST8 (ab228832, Abcam), anti-Cleaved caspase 8 (9496, CST), anti-Pro caspase 8 (ab108333, Abcam) and anti-P84 (10920-1-AP, Proteintech) antibodies. Then, the membranes were washed and incubated for 1 h with an HRP-conjugated secondary antibody. An ECL chemiluminescence system (Tanon-4800 Multi) was employed for the detection of proteins.

Total RNA was isolated using TRIzol (Invitrogen) following the manufacturer’s protocols. First-strand cDNA was synthesized using the GoScript™ Reverse Transcription System (Promega). RT-qPCR analysis was performed using SYBR Green GoTaq® qPCR Master Mix (Promega) in an AB 7500 Real-time PCR Instrument (Applied Biosystems). All reactions were performed in triplicate, and Gapdh or Actin were used as endogenous controls. The following RT-PCR primers were used:

ALKBH5: 5’- CGGCGAAGGCTACACTTACG − 3’; 5’- CCACCAGCTTTTGGATCACCA − 3’.

MLST8: 5’- ATCCGCATGTATGATCTCAACTC − 3’; 5’- CCACAGACGCGATGTTCTTG − 3’.

EIF4EBP1: 5’- CCTGATGGAGTGTCGGAACT − 3’; 5’- CCGCTTATCTTCTGGGCTATTG − 3’.

m⁶A-seq analysis was performed according to a standard method previously described [23, 28, 31]. Polyadenylated RNA was fragmented and incubated with a specific anti-m⁶A antibody (Synaptic Systems) for 2 h at 4 °C. Captured RNA was washed and then eluted and purified using an RNeasy Mini Kit (QIAGEN). Library preparation was performed using a SMARTer Stranded Total RNA-Seq Kit v2 - Pico Input Mammalian (TaKaRa-Clontech). The samples were sequenced using a NextSeq 500 High Output Mode 75 cycles kit (Illumina).

Changes in global m⁶A levels in mRNAs were measured using an EpiQuik m⁶A RNA Methylation Quantification Kit (Epigentek) according to the manufacturer’s instructions. Briefly, total RNA was bound to strip wells using RNA high-binding solution. m⁶A was detected using capture and detection antibodies. The detected signals were enhanced and then quantified colorimetrically by reading the absorbance using a microplate spectrophotometer. The amount of m⁶A is proportional to the OD intensity measured.

MeRIP was performed according to a previously reported protocol [30]. Polyadenylated RNA was incubated with an anti-m⁶A antibody (Synaptic Systems) in 500 µl IP buffer (150 mM NaCl, 0.1% NP-40, 10 mM Tris, pH 7.4, 100 U RNase inhibitor) for 2 h at 4 °C. Next, Dynabeads® Protein A (Thermo Fisher Scientific) was added to the mixture and rotated for another 2 h at 4 °C. The captured RNA was washed 4 times, eluted twice with 50 µl m⁶A-elution buffer (IP buffer, 6.7 mM m⁶A, 30 U RNase inhibitor), incubated and shaken at 4 °C for 1 h and precipitated with 5 mg glycogen and one-tenth volumes of 3 M sodium acetate in 2.5 volumes of 100% ethanol at -80 °C overnight. Gene-specific m⁶A enrichment was assessed by qPCR.

Cells seeded in 12-well plates were cotransfected with 200 ng pcDNA3.1- ALKBH5-WT/H204A, 20 ng pRL-TK (Renilla luciferase control reporter vector) and 300 ng pMIR-Report luciferase vector (Promega) fused with the wild-type EIF4EBP1-3’UTR and MLST8-3’UTR. 24 h after transfection, cells were treated with BP for a period of 24 h. Then, the relative luciferase activities were calculated using a Dual-luciferase Reporter Assay System (Promega). Each experiment was repeated in trice.

MV4-11 and HL-60 cells seeded in 6-well plates were treated with BP for a period of 48 h, after which 5 µg/ml actinomycin D was added at 6 h, 3 h and 0 h before cell scraping and collection. Total RNA was isolated using an RNeasy kit (QIAGEN) under conditions recommended by the manufacturer. The relative levels of target mRNA were determined by RT-PCR. The degradation rate of target mRNA was estimated according to a previously published method[19].

After deparaffinization and rehydration, tumor sections were immunostained using an anti-Ki67 monoclonal antibody. All images were captured at 20x magnification. The procedure was carried out according a previously published method [31].

All statistical analyses were performed using the SPSS 20.0 software package (IBM, Chicago, IL, USA). The significance of the differences was assessed using the two-tailed Student’s t-test or a chi-squared test, as appropriate. Data are presented as the mean ± SEM derived from three independent experiments. Significant p-values are indicated as follows: not significant (ns); p-value < 0.05 (*); p-value < 0.01 (**); p-value < 0.001 (***).

The IC50 of BP on AML cells was evaluated using a CCK-8 assay to determine the proper concentration required for 50% inhibition (IC50). The IC50 values of BP were 389.04 ng/ml and 66.83 ng/ml for MV4-11 and HL-60 cells, respectively (Fig. 1a). To examine the biological effects of BP, we monitored cell proliferation, viability, and apoptosis under BP treatment in different human leukemia cell lines, including MV4-11 and HL-60. First, we employed a MTT assay to investigate the effect of BP treatment on AML cell proliferation. As shown in Fig. 1b, we found that the BP group, cytarabine group and BP and cytarabine combination group exhibited lower proliferation rates than the control group. Significantly, the BP and cytarabine combination group showed the lowest growth rate (Fig. 1b). For further investigation, we performed an ApoLive-Glo™ Multiplex assay to assess the viability and caspase activity of AML cells under different treatment conditions. Similarly, we found that compared to the control group, treatment with BP or cytarabine decreased cell viability and promoted apoptosis. As expected, the combination of BP and cytarabine led to the lowest cell viability (Fig. 1c, d). Taken together, these results suggest that BP suppresses cell proliferation and promote apoptosis in vitro.

We next employed a xenograft model of HL-60 cells to examine the effect of BP on tumor growth in vivo. We found that BP treatment significantly (p < 0.05; log-rank test) slowed down tumor growth in tumor-bearing mice (Fig. 2a). Across multiple time points, a significantly lower (26.4% lower at the endpoint) tumor volume was detected in mice treated with BP compared to that in the control group (Fig. 2b). Consistently, at the experimental endpoint, mice treated with BP carried tumors with a lower weight than those formed in the control group (Fig. 2c). Furthermore, we explored the benefit of combining BP with cytarabine in the HL-60 xenograft model. As shown in Fig. 2a, c, we found that tumors in the combination group were significantly smaller and weighted less those in the single BP or cytarabine treatment group. At each time point, there was no significant body weight loss, which indicates that all treatments were well tolerated (Fig. 2d). Then, we employed H&E staining to assess the histologies of the respective spleens (Fig. 2e). We found that compared to the control group, all treatments relieved the disruption of the spleen structure, including white pulp atrophy and boundaries separating white and red pulp blur. In addition, the histologies of the subcutaneous tumors were confirmed by H&E staining, with many tumor cells observed in the tumor mass (Fig. 2f). By utilizing immunohistochemistry, we found that the proportion of Ki67-positive cells, a marker of cell proliferation, in the BP group was significantly lower than that in the control group, but slightly greater than that in the cytarabine group or the combination of BP and cytarabine group (Fig. 2 g). Collectively, these data indicate that BP can significantly inhibit AML tumor growth in vivo.

Emerging evidence indicates that m⁶A modification is involved in the progression of various cancers. Therefore, we speculate that BP may inhibit leukaemic cell proliferation and induce apoptosis, leading to suppression of leukemia tumorigenesis by regulating mRNA m⁶A enrichment and expression. Using m⁶A sequencing (m⁶A-seq), we mapped the m⁶A methylomes of AML cells under different treatment conditions. We found that the GGACC motif was highly enriched within m⁶A sites of BP-treated AML cells (Fig. 3a). In total, we identified 10,478 m⁶A peaks representing the transcripts of 7684 genes in control cells, 15,474 m⁶A peaks representing the transcripts of 10,369 genes in BP-treated cells, 9403 m⁶A peaks representing the transcripts of 6901 genes in cytarabine-treated cells, and 16,508 m⁶A peaks representing the transcripts of 10,829 genes in combination-treated cells (Fig. 3b; Fig. S1a). Furthermore, m⁶A-seq revealed that the expression of regulators related to RNA methylation changed under different treatment conditions compared with the control samples. Compared to those in the control group, methyltransferases, such as METTL3, METTL14, METTL16, WTAP, RBM15 and ZC3H13, in the BP treatment group were markedly upregulated, while the demethylase ALKBH5 was significantly downregulated. The combination treatment group shared similar changes, while there was no significant difference in the cytarabine group (Fig. 3c; Table S1). Given the significant changes in m⁶A modification, we further explored the target genes affected by BP treatment. A total of 399 genes showed differential methylation modification in the BP group, in which methylation modification of 273 genes was upregulated, while that in 126 genes was downregulated. Similar to the BP group, 261 genes showed upregulated methylation modifications, and 139 genes showed downregulated methylation modifications in the BP and cytarabine combination group. Of the 139 genes with changes in methylation modification in the cytarabine group, 64 exhibited upregulated changes, while 75 were linked with downregulated changes (Fig. 3d). Four quadrant graph analysis revealed that the mRNA m⁶A level increased in the BP treatment group and indicated that genes with upregulated methylation are expected to encompass genuine targets of BP treatment (Fig. S1b). Among genes with upregulated methylation in the BP group, we found that the expression of 5 genes was increased and that of 36 genes was decreased (FC > 2 & p < 0.05). We employed the Molecular Signatures Database (MSigDB) to analyse the genes filtered above and found that 26 out of 36 downregulated genes and 3 out of 5 upregulated genes were included in the leukemia-related gene set (Fig. 3e; Table S2). We next investigated the pathways of the differentially expressed genes that were shared by ClueGO (plugin for Cytoscape software) and found that the significantly enriched pathways included mTORC1-mediated signalling, mTOR signalling and TOR signalling (Fig. 3f). Therapeutic targeting of the mTOR pathway as an anticancer strategy for leukemia has been under extensive investigation. mTOR signalling plays a central role in the proliferation and survival of AML and is aberrantly activated in 60% of AML patients [32, 33]. mTOR, which is a serine/threonine kinase, is a major component of different protein complexes, including mTOR complex 1 (mTORC1) and mTOR complex 2 (mTORC2) [34]. mTORC1 controls mRNA translation and protein synthesis through phosphorylation of S6K and 4E-BP [35], whereas mTORC2 mainly mediates signal transduction by regulating AKT kinase [36, 37]. In the BP group, the expression of EIF4EBP1 and MLST8 was significantly decreased. Next, we focused on these two potential ALKBH5 targets for further study.

Next, we investigated whether BP treatment increases the m⁶A RNA modification levels in cells. Consistent with the previous m⁶A-seq results, we found that BP treatment significantly increased global m⁶A levels in leukaemic cells (Fig. 4a). Since MLST8/EIF4EBP1 may act directly downstream of ALKBH5 and drive the altered proliferative effects, we set out to investigate the underlying mechanism. First, we confirmed their protein and mRNA levels by Western blot and RT-qPCR analyses, respectively. Compared to control cells, BP-treated MV4-11 and HL-60 cells showed both decreased protein and mRNA levels of ALKBH5, MLST8 and EIF4EBP1 (Fig. 4b, c). Next, we employed methylated RNA immunoprecipitation (Me-RIP) combined with RT-qPCR to determine the levels of MLST8 and EIF4EBP1 m⁶A methylation following BP treatment. We found that the m⁶A levels of MLST8 and EIF4EBP1 in BP-treated cells were significantly increased compared to those in control cells (Fig. 4d). Together, these results indicate that MLST8 and EIF4EBP1 act as direct downstream targets of ALKBH5 in AML.

To investigate the importance of the demethylation activity for ALKBH5-induced downregulation of MLST8/EIF4EBP1 expression, we constructed an ALKBH5 mutant H204A that has completely lost its demethylation activity. Next, we employed Western blotting to examine protein expression of MLST8 and EIF4EBP1. We found that overexpression of wild-type ALKBH5, but not mutant H204A, could rescue the expression of MLST8 and EIF4EBP1 in BP-treated cells (Fig. 5a). Sequence analysis revealed that there were three and two matches to the 5′-RRACU-3′ (methylated adenosine residue is underscored) m⁶A consensus sequences within the MLST8 and EIF4EBP1 3′-UTRs, respectively. Next, we constructed m⁶A consensus sequences within the MLST8 and EIF4EBP1 3′-UTRs in the luciferase plasmid and found that compared with the control, BP treatment resulted in constitutively increased luciferase activities of MLST8 and EIF4EBP1. Overexpression of wild-type ALKBH5, but not inactive mutant H204A, could rescue BP-induced enhancement of the luciferase activities of MLST8 and EIF4EBP1(Fig. 5b). We utilized actinomycin D to block de novo RNA synthesis to assess the RNA stability of MLST8 and EIF4EBP1. BP treatment significantly decreased the half-life of MLST8 and EIF4EBP1 mRNA transcripts in AML cells (Fig. 5c). These results indicate that BP may suppress the ALKBH5-mediated reduction in the RNA stability of MLST8/EIF4EBP1 by catalysing m⁶A demethylation.

To further determine whether the inhibitory effect of BP on AML cell proliferation relies on the ALKBH5/MLST8/EIF4EBP1 axis, we performed a rescue experiment. A CCK-8 assay showed that BP treatment resulted in a lower proliferation rate compared to that in the control group. Notably, compared with BP treatment, overexpression of MLST8 and EIF4EBP1 drastically increased cell proliferation. Moreover, overexpression of wild-type ALKBH5, but not mutant H204A, significantly rescued BP-induced inhibition of cell proliferation (Fig. 6a). Next, we performed an ApoLive-Glo™ Multiplex assay to assess the viability and caspase activity of these stable AML cells. In accordance with the increased cell proliferation, we found that re-expression of ALKBH5/MLST8/EIF4EBP1 led to rescue of the reduced viability, while mutant ALKBH5 H204A had lost this regulatory function (Fig. 6b). Interestingly, we found that BP treatment led to increased apoptosis, while reintroduction of ALKBH5-WT/ALKBH5-H204A/MLST8/EIF4EBP1 upregulated the BP-induced promotion of apoptosis (Fig. 6c). We also assessed the expression levels of apoptosis related genes including Pro-caspase 8 and Cleaved-caspase 8 and found that BP treatment resulted in upregulation of Cleaved-caspase 8 p43/p41compared to the control group. Overexpression of ALKBH5-WT/MLST8/EIF4EBP1 could reverse the increased Cleaved-caspase 8 level rather than mutant ALKBH5 H204A (Fig. 6d). These data suggest that BP-mediated inhibition of cell proliferation depends on the ALKBH5/MLST8/EIF4EBP1 axis.

To investigate whether the BP-regulated ALKBH5/MLST8/EIF4EBP1 axis has clinical relevance in AML patients, we analysed ALKBH5/MLST8/EIF4EBP1 expression profiles of AML patients for whom overall survival (OS) data were available in the TCGA dataset. We found that a lower expression was positively correlated with better OS among these patients. The median survival times of patients with a lower expression of ALKBH5, MLST8 and EIF4EBP1 were 950 days, 650 days and 700 days, respectively, whereas thoseof patients expressing higher levels were 334 days, 400 days and 350 days, respectively (Fig. 7a). Next, we examined expression correlation among ALKBH5, MLST8 and EIF4EBP1 in the TCGA dataset of AML patient samples. We found a statistically positive correlation between the expression levels of ALKBH5 and EIF4EBP1 or MLST8 (Pearson’s r = 0.249, p = 0.002; Pearson’s r = 0.224, p = 0.006) (Fig. 7b). These data suggest a critical role of the ALKBH5/MLST8/EIF4EBP1 axis in the prediction of clinical outcome and highlight the potential of the ALKBH5/MLST8/EIF4EBP1 axis as a molecular target in AML. Overall, our data indicate that BP-mediated downregulation of ALKBH5 leads to decreased m⁶A demethylation of EIF4EBP1 and MLST8 mRNA and, thereby, reduced EIF4EBP1 and MLST8 RNA stability, contributing to inhibition of AML cell proliferation (Fig. 7c).