RUNX1-Regulated Pathways and Biomarkers in Acute Myeloid Leukaemia ^†

Abstract

Runt-related transcription factor 1 gene (RUNX1), also known as acute myeloid leukaemia 1 protein (AML1), plays a crucial role in the pathogenesis of AML. RUNX1/AML1 is one of the most frequently mutated leukaemias associated with a poor prognosis in AML. Researchers and clinicians can develop personalized medicines and improve diagnosis by identifying the biomarkers associated with mutations. In the current study, we used the genome and transcriptome data from The Cancer Genome Atlas-Acute Myeloid Leukemia (TCGA-AML) cohort. We analysed RUNX1 mutant AML patients compared to non-mutant patients using an integrated multi-omics, multi-database analysis of exome, and transcriptomics data. Finally, we identified the gene signature associated with RUNX1 mutations, including prognostic genes that significantly influenced the overexpression of RUNX1 mutation-associated genes in AML patients. Our results can help to diagnose AML patients with RUNX1 mutations at an early stage.

1. Introduction

Acute myeloid leukaemia (AML) is a malignant hematological disease affecting the blood and bone marrow. RUNX1 (also known as AML1) is a transcription factor that plays an important role in blood cell development and function [1]. Mutations in the RUNX1 gene have been linked to several blood disorders, including AML, and are associated with a poor prognosis. RUNX1 mutations can lead to the cause of familial platelet disorder (FPD). Researchers are actively studying RUNX1 and its role in blood disorders with the aim of developing more effective treatments [2]. This includes the development of targeted therapies that specifically target abnormal blood cells produced by RUNX1 mutations, as well as the development of new strategies to restore the RUNX1 gene to normal function. In this study, we used TCGA-AML [3] data in which RUNX1 is mutated in 9% of patients and identified the prognostic biomarkers specific to the RUNX1 mutation.

2. Methods

2.1. Identification of Mutational Landscape of RUNX1 in TCGA-AML

The cBioPortal for Cancer Genomics website was used to identify the RUNX1 mutational landscape in AML patients from the TCGA study (n = 200) [3,4,5].

2.2. Analysis of Differentially Expressed Genes (DEGs) in RUNX1-Mutated TCGA-AML

Out of 200 TCGA-AML patients, only 173 patients’ RNA-Seq data were available. Based on the RUNX1 mutations of TCGA-AML, we stratified the total number of patients into two groups and designated them as RUNX1-mutated (n = 17) (Table S1) and wild-type (n = 156) (without RUNX1 mutations), respectively. The mRNA expression profiles (RNA Seq V2 RSEM) were checked to identify the DEGs in these two groups. From the list of DEGs, we can conclude that they are the driving genes behind tumorigenesis and cancer progression.

2.3. Functional Annotation and Survival Analysis

The functional annotation of the DEGs from RUNX1-mutated patients was performed by a web tool named DAVID [6]. This analysis provides the Gene Ontology (GO) and the Kyoto Encyclopedia of Genes and Genomes (KEGG) Pathway information for genes.

The GEPIA2 tool [7] was used to evaluate the prognostic value of the DEGs identified in the patients with the RUNX1 mutation in the TCGA-AML cohort. Briefly, for the TCGA-AML cohort, the patient samples are divided into two risk groups such as low-risk and high-risk groups, and the log-rank test, also known as the Mantel–Cox test, was performed to construct the overall survival plots. A p-value cut-off of <0.05 was used as the significance threshold in the search for prognostic biomarkers.

3. Results and Discussions

In TCGA-AML, RUNX1 mutations occurred in 9% of patients (n = 17) out of a total of 200 patients (Figure 1A and Table S1). We then performed the DEGs analysis between the RUNX1-mutated vs. wild-type patients using cBioPortal. As a result, we obtained a total of 210 DEGs containing 155 upregulated and 55 downregulated genes in RUNX1-mutated patients with a fold change (FC) threshold of >2 and a p-value and q-value of <0.05 (Figure 1B and Table S2).

Next, we selected DEGs, and then performed functional annotation analysis with DAVID separately for up- and downregulated genes. Interestingly, the KEGG analysis of upregulated genes obtained 10 pathways with a stringent p-value cut off of <0.05, in which the majority of genes are involved in pathways in cancer, focal adhesion, apelin signalling pathway, ABC transporters, small cell lung cancer, JAK-STAT signalling pathway, toxoplasmosis, proteoglycans in cancer, endocrine resistance, hematopoietic cell lineage and amoebiasis (Figure 1C and Table S3).

Next, we focused on the KEGG pathway analysis (p-value < 0.05) of downregulated genes, from which we obtained only one pathway identifying the genes involved in Staphylococcus aureus infection (Table S3). Overall, the functional annotation analysis showed that the genes that are upregulated in the RUNX1-mutated patients are involved in a variety of signalling pathways that drive AML.

Our next goal was to examine whether these upregulated genes in RUNX1-mutated patients play a role in the prognosis of AML patients. Using the GEPIA2 web tool, we identified seven poor prognostic biomarkers whose significantly higher expression (p-value < 0.05) results in poor overall survival in TCGA-AML patients (Figure 2A). The genes with poor prognosis identified in our study are EGFEM1P, DOCK1, HTR1F, CALCRL, HOPX, TRIM9 and MYLK. These results clearly indicate that the increased expression of genes associated with RUNX1 mutations acts as a biomarker in AML patients. We considered these seven genes to be RUNX1 mutation-associated gene signatures (RMAGS) in AML. Notably, the higher expression of two downregulated genes such as KCNE5 and ROPN1L showed a good prognosis in AML patients (Figure 2B,C).

4. Conclusions

Taken together, our results identified a list of genes that are associated with the RUNX1 alterations in AML patients, and we named them RMAGS. The increased expression of RMAGS predicts poor survival in AML patients. These seven genes may act as prognostic biomarkers individually and combinedly and can be used to identify the RUNX1 mutation status in AML patients. In summary, our identified RMAGS could be possible targets in the treatment of AML, in that the development of combined inhibitors for this gene signature, along with RUNX1, could pave the way for the development of personalized/precision medicine to suppress RUNX1-mediated tumour growth and drug resistance.