Acute myeloid leukemia (AML) is characterized by the accumulation of immature blast cells due to the inhibition of normal hematopoiesis, and aberrant proliferation and differentiation of immature myeloid hematopoietic cells in the bone marrow (BM) (Xin et al., 2022; Cheng et al., 2020). It is the most prevalent variety of acute leukemia in adults and has a high morbidity and mortality rate (Cheng et al., 2020; Juliusson et al., 2012). Tumor formation in AML is highly complex and is associated with various gene mutations (Cheng et al., 2020). In recent years, there have been advances in genomic, transcriptomic and epigenomic studies of AML (Kirtonia et al., 2022). Although these studies have improved our understanding of AML, its pathogenesis has not yet been completely elucidated (Liu et al., 2019). AML affects numerous organs, including the lymph nodes, liver, spleen, and central nervous system, and has a poor prognosis (Cheng et al., 2020). AML can be treated with chemotherapy, hematopoietic stem cell transplantation, cell therapy, biological immunotherapy, and gene-targeted therapy, although the success rate of treatment is not that high (Xin et al., 2022). Despite the evolving treatment options, the ineffectiveness of chemotherapy and the lack of information on the molecular pathogenesis of AML of the disease still remain a problem (Kirtonia et al., 2022). The majority of patients experience a relapse after obtaining complete remission, and some even dies (Cheng et al., 2020). Accurate diagnosis, prognosis, and efficient treatment of AML depend on the identification and development of therapeutic targets (Liu et al., 2019). It is predicted that studies on the pathogenesis and therapeutic targets of AML may have important clinical application value (Cheng et al., 2020). For this purpose, revealing the potential roles of non-coding RNAs, especially in AML, will be crucial in terms of understanding the molecular pathways related to AML development, chemotherapeutic response, and relapse. Previous studies have shown the role of non-coding RNAs (ncRNAs) in the pathogenesis of cancers, including AML (Kirtonia et al., 2022).

MicroRNAs (miRNAs) are highly conserved small non-coding RNA molecules, usually 18–24 nucleotides long, acting as negative post transcriptional regulation of gene expression (He and Hannon, 2004). The abnormal expression of miRNAs is associated with a variety of diseases, including cancer, and offers crucial information on the molecular pathophysiology of the diseases (O’Brien et al., 2018). miRNAs are dysregulated levels in AML and are crucial for the development or repression of the disease (Fletcher et al., 2022). Through alterations in proximity to the oncogenic genomic area, epigenetic modifications, inappropriate targeting of miRNA promoter regions by altered transcription factors or oncoproteins, and processing of deregulated miRNAs, these molecules could be involved in the pathogenesis of AML (Liu et al., 2019). Depending on the molecular pathways of the mRNA they target, miRNAs exhibit either oncogenic or tumor-suppressive properties (Otmani and Lewalle, 2021). There are many miRNAs involved in the development of AML, one of which is miR-342 (Fletcher et al., 2022; Journal et al., 2019; Expression et al., 2022). Abnormal expression of miR-342 has been observed in various types of cancer (hepatocellular carcinoma, cervical cancer, lung cancer, etc.) and has been associated with tumor development and proliferation (Gao et al., 2017; Li X. et al., 2014; Zhao and Zhang, 2015). miR-342 functions as a tumor suppressor in AML (Journal et al., 2019). However, the molecular mechanisms involved in miR-342 underlying the pathogenesis of AML remain unclear. Identifying the potential target of miR-342 and the associated biological pathways may suggest the miR-342 is a new potential therapeutic target in AML patients.

Murine double minute (MDM) proteins are overexpressed in a various of cancer types, and they primarily exhibit their oncogenic properties by inhibiting the p53 tumor suppressor (Li and Lozano, 2013). It also modulates and responds to many signaling networks (Wade et al., 2013). Both MDM2 (murine double minute 2) and MDM4 (murine double minute 4, also know as MDMX) form a heterodimer that closely regulates p53 function (Eskandari et al., 2021). MDM4 binds p53 to induce transcriptional inactivation and thus inhibits p53 function. Experimental evidence showed that p53 is kept inactive in the cytoplasm at high levels by interactions with MDM4. Further investigation of p53 in AML cell lines OCI/AML-2 (AML2) is mostly due to cytoplasmic and MDM4 binding (Tan B. X. et al., 2014). It is crucial to disrupt the MDM-p53 relationship in order to activate p53 while developing effective therapeutic approaches for tumor therapy. Therefore, scientists have focused on the roles of MDM4 as a negative regulators of the p53 tumor suppressor (Li and Lozano, 2013).

Microarray analysis can be used to identify miRNA changes in AML (Candia et al., 2015). Based on these data, new and functional miRNAs can be identified by bioinformatic analysis. In recent years, bioinformatics analysis has been used in oncology research to identify genetic alterations and new potential cancer-related biomarkers (Sheng et al., 2021). Few studies have been conducted to analyze the prognostic lncRNA-miRNA-mRNA (Wang et al., 2019) and circRNA-miRNA-mRNA (Lv et al., 2018) ceRNA network in AML (Cheng et al., 2020).

The aim of this study is to investigate the mechanisms underlying the development of AML, potential target genes, and new potential biomarkers for AML prognosis through miRNA-associated mRNA network in silico analysis. We focus on the mir342-5p/MDM4/p53 network given these findings because it may present exciting opportunities for comprehending the underlying molecular process and for cancer therapy.

To identify DE-miRNAs, a comparison between AML and normal control groups were performed by GEO2R, and a total of 250 miRNAs were identified between AML and normal samples. 24 De-miRNAs (16 upregulated and 8 downregulated) (Table 1) were identified from the GSE51908, using adjusted p < 0.05 and llogFCl>2 for DE-miRNAs selection. Volcano plot, visually demonstrating of DE-miRNAs is shown in Figure 1A. Box data normalization (Figure 1C) and scatter plots of miR-342-5p (Figure 1B) are shown in Figure 1.

The levels of miR-342-5p were significantly different between control and AML groups (p = 0.0001). To determine the power of miR-342-5p expression levels to distinguish patients with AML from control, the ROC analysis was performed. Values for AUC, sensitivity, specificity and cut-off points for miR-342-5p shown in Figure 2A. miR-342-5p showed significantly higher AUC values (AUC = 0.795, p = 0.0001). The cut-off of miR-342-5p was 6.85.

To further evaluate the clinical relevance of AML, the overall survival of AML patients was assessed by the Kaplan-Meier test using data from the article of Bullinger et al. (2010). Interestingly, as shown in Figure 2B, AML patients with low miR-342-5p expression had significantly shorter survival than those in higher expression control group (p = 0.0001), and its AUC value to discriminate AML from control groups was 0.795 (Figure 2A).

miR-342-5p in AML, potential target genes were predicted by the mirWALK database (miRDB and miRTarBase algorithms) (Table 2). Furthermore, the target genes of hsa-mir-342-5p at HL-60 and KG-1 cells were determined using miRDB (Figures 3A, B).

Common target genes identified in HL-60 and KG-1 cells, and the mirWALK database were selected. Target gene expression values in AML were determined using GEPIA (LAML; Acute Myeloid Leukemia) (Table 3). And also, Multi-Gene comparison was performed according to AML and TCGA normal and GTEx data (Figure 4A). MDM4 expression level is increased in tumor tissue compared to healthy tissue (Figures 4B, C).

High expression of MDM4 predicts an unfavorable prognosis in AML patients. This highlights the important prognostic value of MDM4 in AML. The overall survival analysis of MDM4 is shown in Figure 5.

The PPI network of MDM4-related target genes was obtained from the STRING and GeneMANIA database (Figures 6A, B). STRING database demonstrated that the MDM4 gene primarily interacts with TP53, MDMD2, CNSK1A1, UBE2S2, UBB, UBC, USP7, CDKN2A, ATM, and CHEK2. GeneMANIA demonstrated that mdm4 interacts with MDM2, TP53, TP73,USP2,CBLC, HGF, ABL1, CD44, CDKN2A, USP7, SMARCD2, SMARCD3, SMARCD1, BECN1, MAP2K5, PELI1, GLIS2, CDKN2B, CDK4, and FAM193A. Among the genes, especially MDMD2 and p53, which are related to cell cycle and apoptosis, were selected.

The Gene Ontology and KEGG pathway analyses were performed to better understand the pathway and process affected by the identified genes. GO terms enriched by MDM4, MDM2, and p53 genes are shown in Table 4. Three terms biological process (BP), cellular component (CC), and molecular function (MF) are included in the GO ontology. The biological processes of these differentially expressed genes were primarily involved in the cellular response to hypoxia, macromolecular complex assembly, regulation of cell cycle, negative regulation of the apoptotic process, negative regulation of transcription from RNA polymerase II promoter, negative regulation of transcription, DNA-templated, positive regulation of transcription from RNA polymerase II promoter, and regulation of transcription from RNA polymerase II promoter. Meanwhile, the genes related to cellular components were mostly involved in transcriptional repressor complex, nucleoplasm, and nucleus. In terms of molecular function, these differential genes were mostly enriched in p53 binding, enzyme binding, zinc ion binding, metal ion binding, and protein binding.

In addition, the top 20 results of KEGG pathway enrichment analysis of MDM4, MDM2, and p53 genes are shown in Table 5. Among these pathways, the PI3K-Akt signaling pathway, microRNAs in cancer, proteoglycans in cancer, p53 signaling pathway, cellular senescence, and cell cycle are closely correlated with the development of AML.

Acute myeloid leukemia is one of the most prevalent hematopoietic malignancies in adults (Short et al., 2018). Although significant advancements in the clinical treatment of AML, including stem cell transplantation and chemotherapies (Deng et al., 2017), the treatment is not at the desired level (Wang et al., 2022). The 5-year survival rate for AML patients is less than 15% (Zhao et al., 2015). Early diagnosis and timely intervention in patients with AML can improve their chances of survival. Cytogenetic and molecular analysis are very important in predicting the remission and survival rates of AML patients (Article et al., 2017). Understanding the molecular mechanisms underlying the pathogenesis of AML, identifying new biomarkers for diagnosis and prognosis, exploring new therapeutic strategies and improving the clinical outcome of AML is important (Journal et al., 2019; Article et al., 2017). Therefore, an improved understanding of the pathogenesis of AML will provide new insights into the diagnosis and treatment of AML. Recent studies have revealed that miRNAs are associated with the formation and development of AML, and play a role in the pathogenesis of AML by regulating their target genes (Fletcher et al., 2022; Wang et al., 2022). In order to comprehend the molecular mechanism of AML and develop new therapeutic targets for AML, it is necessary to identify unregulated miRNAs or mRNAs.

Microarray and High-throughput RNA sequencing (RNA-Seq) are widely used in cancer research to understand gene functions, find target molecules, and identify biomarkers for disease classification and diagnosis (Tang et al., 2017). Data obtained through experimental analysis are used in bioinformatics approaches. In particular, it is seen that bioinformatics data mining is a new trend in cancer research (Çakmak, 2022). Information on gene expression variations between normal and malignant tissues, chemical interactions, and the discovery of novel potential biomarkers for diagnosis and treatment can be obtained using bioinformatics tools (Çakmak, 2022).

The goal of the study is to identify potential target genes and important miRNAs for the diagnosis and therapy of AML by bioinformatics analysis. To the best of our knowledge, this is the first study to investigate the relationship between AML and miR-342-5p and target genes by bioinformatics analysis in the literature. Microarray data of GSE51908 were obtained from the GEO database to identify DE-miRNAs between AML and normal control (healthy). It was determined that several miRNAs were significantly differentially expressed, especially miR-342-5p′s expression was decreased in AML (Table 1; Figure 1). miRNA expression profiles are associated with prognosis, highlighting the potential significance of miRNAs in this disease (Garzon et al., 2008).

The Kaplan-Meier general database showed that MDM4 was significantly associated with the prognosis of AML patients (Figure 5). In addition, Kaplan-Meier analysis was performed to determine the survival times of the AML and control groups over miR-342-5p expression levels. Low miR-342-5p strongly predicted overall survival in AML patients. Moreover, Cox regression analysis showed that miR-342-5p expression independently provided prognostic information. Similarly, studies have shown that some miRs have a lower mortality rate and a huge effect on survival. Among the target genes of miR-10a-3p, SLC14A1, ARHGAP5, PIK3CA, it has been shown to show significant negative association with it, leading to longer survival in AML (Bullinger et al., 2010; Thanh et al., 2021). Similar to these studies, it has been shown that miR-342-5p is associated with the diagnosis and prognosis of patients with AML and has a positive effect on survival.

The findings of the ROC and overall survival studies show that miR-342-5p had diagnostic and prognostic significance in AML (Figure 2). The potential use of miRNAs as disease biomarkers has been the most common clinical use to far (Wallace and Connell, 2017). ROC analysis was performed to confirm whether the miR-342-5p can be used as one of the markers for AML. The miR-342-5p levels exhibited a significantly higher AUC value (AUC = 0.795, p = 0.0001), (Figure 2A). Based on this finding, it is concluded that miR-342-5p levels can be used as a reliable marker to evaluate AML. Previous studies revealed that plasma miR-92a, miR-143, and miR-342 could be promising biomarkers for AML (Article et al., 2017). In AML, miRNAs act as tumor suppressors or oncomiRs to affect a wide variety of leukemic processes, such as self-renewal, differentiation, proliferation, and epigenetic regulation. These molecules influence leukemic development and progression by targeting directly at the mRNA level or promoting malignancy (Wallace and Connell, 2017).

In addition to being a promising biomarker of miR-342-5p, it provides information about the molecular pathways involved in AML pathology. For this purpose, the target genes of miR-342-5p were determined using the miRwalk database. In addition, target genes identified in HL-60 and KG-1 cells and genes that are match as a result of mirwalk data were selected. Using the GEPIA database, expression analyses of these genes in AML were performed. In the GEPIA analysis, TCGA normal and database based on GTEx data in 173 AML patients and 70 healthy subjects, MDM4 expression level was higher in AML patients (Table 3; Figure 4). In addition, high expression levels of MDM4 were associated with poor overall survival (Figure 5).

Mouse Double Minute 4, which is among the target genes of miR-342-5p, has been reported to increase its expression in glioma, soft tissue sarcoma, melanoma, retinoblastoma, breast cancers and hematological malignancies (Li L. et al., 2014; Touqan et al., 2013; Furgason et al., 2015; Haupt et al., 2015; Carvajal et al., 2018). It has been stated that patients with AML have very high MDM4 expression and exhibit oncogenic activity compared to many other tumor types (Carvajal et al., 2018).

Functional enrichment analysis and PPI network were performed to identify the proteins that MDM4 is associated with in the development of AML and to reveal how these proteins function. According to the match data of the STRING, and GeneMANIA database, p53 and MDMD2 proteins were determined, especially in relation to apoptosis and cell cyle (Figures 6A, B). According to the literature, MDM4 inhibit p53 transcriptional activity by interacting with the p53 transactivation domain through its N-terminal domain (Tan B. X. et al., 2014). The tumor suppressor p53 protein, which guards the genome, triggers cell DNA repair, cell cycle arrest, senescence, and/or apoptosis pathways (Vousden and Lane, 2007). Downregulate p53, MDM2 and MDM4 are mutually dependent on one another (Badciong and Haas, 2002). MDM4 forms heterodimers with MDM2 via its RING domains and enhances the degradation of p53 by stimulating MDM2 E3 ubiquitin ligase activity (Wang et al., 2011). Loss of p53 function in stem and myeloid progenitor cells promotes the onset of leukemia (Zhao et al., 2010). In AML, TP53 mutations occur in less than 10% of patients (Kadia et al., 2016). Despite the TP53 mutations found in some subtypes of AML, MDM4 or MDM2 overexpression is the primary cause of p53 inactivation. Given the critical roles of MDM4 or MDM2, disruption of MDM4/MDM2/p53 interactions may offer a new way to restore p53 activity. Despite recent efforts to develop MDM4/MDM2 inhibitors, the desired success has not been achieved as a result of lack of affinity and dose-limiting toxicity (Carvajal et al., 2018). Therefore, targeting MDM4 with miRNA replacement therapy may be an attractive therapeutic strategy in AML.

Kyoto Encyclopedia of Genes and Genomes pathway enrichment analysis revealed that MDM4, MDM2, and p53 genes are associated with PI3K-Akt signaling pathway in cancer, MicroRNAs in cancer, Proteoglycans in cancer, p53 signaling pathway, cellular senescence and cell cycle. Previous studies have reported the involvement of PI3K-Akt signaling pathway (Bertacchini et al., 2015), p53 signaling (Chen and Sun, 2022), cellular senescence (Mao et al., 2022), cell cycle (Zhou et al., 2021), microRNAs in cancer (Trino et al., 2018), and proteoglycans in AML (Wu et al., 2020).

When the studies with miR-342 in the literature are examined; It has been suggested that dysregulation of the miR-342 Naa10p axis may be involved in the aggressive progression of pediatric AML. Naa10p-siRNA was reported to significantly suppress cell proliferation and increase cell apoptosis (Journal et al., 2019). Wu et al. showed that miR-342-5p is downregulated and has a significant inhibitory effect on cell proliferation in CML. miR-342-5p was shown to target the 3′-UTR domain of CCND1 and reduce its expression, and overexpression of miR-342-5p increased imatinib-induced DNA double-strand breaks and apoptosis (Wu et al., 2021). In another study, it was determined that miR-342-3p expression was decreased in the plasma of AML patients. Further studies have shown that miR-342-3p targets SOX12 and thus its mimics suppress AML cell growth, accelerate apoptosis and induce G0/G1 cell cycle arrest (Wang et al., 2022).

Given all these data, miR-342-5p is thought to act as a tumor suppressor. Given that the expression of miR-342-5p is reduced in AML, miRNA replacement therapy may provide an opportunity to restore the previously lost contribution of miR-342-5p. MiR-342-5p mimics can suppress cell proliferation in AML. miR-342-5p negatively regulates MDM4 expression, modulates DNA replication signaling pathways by contributing to the hsa-mir342-5p/MDM4/p53 network and may promote apoptosis. miR-342-5p and MDM4 are thought to be targets for the treatment of AML.

This study has a few limitations; only one microarray profile was analyzed as the database of miRNA profiles is rare. Experimental validation was not performed for the biological functions and targets of the identified miRNA. Further studies are needed for the mechanism of miR-342-5p and its potential targets in the development of AML. We believe that it can be a potential biomarker for the early diagnosis of AML patients and that cohort studies should be conducted on this topic.

miRNA and candidate genes associated with the pathogenesis and therapeutic targets of AML were identified using integrated bioinformatics approaches. A new miRNA-mRNA regulatory was axis identified that could improve understanding of the molecular mechanisms underlying the development of AML. These target molecules can serve as prognostic biomarkers and therapeutic targets. However, further studies are needed to confirm these predictive results.