Non-Mutational Changes of Autophagy Marker LC3A in Patients with Acute Myeloid Leukemia; Effect of DNA Methylation and Expression Level of LncRNA-GAS5 and miRNA-155-5p, A Case Control Study

Abstract

Clinical translation of autophagy modulators is tied to thoroughly acquainted with the precise state of this process and its regulators in a particular cancer. LC3Av1 is a marker of autophagosome membrane that has been contributed with pathobiology of myriad of human cancers. In the present study, we examined the effect of promoter methylation and miR-155 and LncRNA-GAS5 (GAS5) expression levels on transcription of LC3Av1 in AML patients. The study included 60 patients with de novo AML and 20 subjects with normal bone marrow cellular composition. Methylation-Sensitive high resolution melting (MS-HRM) was performed for analysis of LC3Av1 CpG island methylation and reverse transcription-quantitative polymerase chain reaction (RT-qPCR) for assessing LC3Av1, GAS5 and miR-155 expression levels. There was a significant elevation in the expression level of miR-155 and repression of LC3Av1 in AML samples. We found that LC3Av1 downregulation was negatively associated with its CpG island hypermethylation and miR-155 expression. Aging leads to overexpression of LC3Av1. GAS5 neither was differently expressed in AML patients compared to control samples nor has been related to LC3Av1 expression. The present study revealed that epigenetic changes like DNA methylation and alteration of miR-155 have a pivotal role in repression of autophagy marker LC3Av1, which potentially could provide the important clues of prognostic and therapeutic targets. The optimal strategies for clinical implementation of autophagy in AML is yet to be fully achieved and deserve further studies.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12288-024-01765-3.

Introduction

Acute myeloid leukemia (AML) is a serious health concern with highest rate of death per incidence among leukemias [1]. In addition to the well-established role of oncogene and tumor suppressor gene (TSG) mutations in AML, it is now abundantly obvious that epigenetic mechanisms have a more widespread ability in perturbation of protooncogenes and TSGs [2]. For example, altered transcriptional regulatory mechanisms give rise to a substantially upregulation of protooncogene, resembling the level seen in cancer cells with mutational defects that alter copy number or the structure of the protooncogene [3]. Conversely, gene-silencing mechanisms can inflict dramatic effects on the expression of certain TSGs in cancer cells, principally rendering the genes functionally inactive in the absence of gene mutations [2, 3]. Regarding the reversibility of epigenetic alterations, it seems AML cells may selectively rely on this mechanism for making change in some cellular functions like autophagy which may be need to restored during cancer progression [2, 4].

The term macroautophagy (autophagy) means “eating (phagy) of part of the cell itself (auto),” through a lysosomal mediated mechanism [5]. Autophagy dysfunction has been linked to a great number of pathophysiologic conditions, including cancer as well as neurodegenerative, cardiovascular, infectious and renal disorders [6]. But even so, two allelic mutational defects of autophagy genes are uncommon in human disorders may be because of its pivotal role in maintaining cellular hemostasis [5]. As aforementioned, it seems that non-mutational defects like abnormal signaling pathways and epigenetic changes play a significant role in autophagy malfunction in the AML [7]. Studies of signaling pathways revealed that genes that are significantly perturbated in AML, like mTORC1 and AMPK serve as the main autophagy regulators in the cell cytoplasm [8]. Activation of these pathways leads to phosphorylation of autophagy related genes such as ULK, BECN1 and ATG13, which finally affect the nucleation step of autophagy process [7]. Beyond the cytoplasmic mechanisms, processes that determined the genes expression level, including regulators of chromatin state and non-coding RNAs, have a fundamental role in the baseline autophagy level of each cell [4, 8]. Analyses of the gene expression profile of ATGs in leukemic samples demonstrated dramatic downregulation compared with normal subjects [9–11]. Conversely, some studies reported an increase autophagy level in AML [12]. These contrary results make autophagy as a dilemma in pathobiology of diseases. Whether the baseline activity of autophagy is elevated or reduced, means to restore its function are available and could be used as therapeutic options. For clinical application of autophagy regulators, we must first know the precise state of autophagy and its regulators in a particular cancer. As our colleagues showed dramatic downregulation of LC3A in AML samples, we were eager to seek non-mutational regulators of this gene in AML. LC3Av1 is a marker of autophagosome membrane that has been contributed with pathobiology of myriad of human cancers [5]. In the present study, we examined the effect of promoter methylation and miR-155 and LncRNA-GAS5 expression levels on transcription of LC3Av1 in AML samples.

Materials and Methods

Subjects’ Characteristics

Bone marrow (BM) samples were collected from 60 patients with AML and from 20 subjects normal bone marrow cells as control. Samples were collected randomly. Each individual signed informed consent for their participation in this study. The Ethics committees of Aja University of Medical Sciences approved the study (IR.AJAUMS.REC.1400.191), and all patients provided their full consent. AML was diagnosed using examination of bone marrow samples and preformation of ancillary tests like flow cytometry and molecular methods. All patients were diagnosed with one of AML-M0-M5, patients with therapy related AML and secondary AML, and whom that started chemotherapy were excluded from study. Control subjects consist of people with normal BM compartment who have an indication for BM aspiration because of ITP susception or for grading of lymphoma. These subjects didn’t receive treatment before that samples were obtained. Table 1 presents the summarized information of the patient.

Table 1.

Summary of patient's demographic data(n = 60)

Age median, y (Range, y)	32.5(3–89)
Sex
Male	37
Female	23
CD34 (with threshold 20%)
Positive	36
Negative	24
Blast percent (Range %)	80(26–98)
Immunological Classification (%)
AML M0-M2	28(46.7)
AML M3	15(25)
AML M4,M5	17(28.3)

Immunophenotypic Analysis

100µL of BM samples were labeled with 10µL antibodies against CD45, CD34, CD117, HLA-DR, cMPO, CD-11b, CD-11c, CD13, CD33, CD14, CD64, CD4, CD3, CD7, CD19 and isotype control (Dako, Denmark). Antibodies were conjugated with FITC, PE, PerCP and APC. Stain-lyse-wash and Lyse-wash-stain-wash procedures were applied for surface and cytoplasmic markers, respectively. Finally, CD-marker of cells were determined by a flow cytometry instrument (Attune, USA). Based on the cytomorphology and status of CD-markers expression patients were categorized as AML with the myeloblastic phenotype (AML-M0, -M1, and -M2), AML with the promyelocytic phenotype (AML-M3), and AML with myelomonocytic differentiation (AML-M4 and − M5) (supplementary Fig. 1).

RNA Extraction and cDNA Synthesis

Total RNA from mononuclear cells was extracted using RNeasy Kit (Qiagen, Germany). Nanodrop ND-1000 (NanoDrop Technologies) was used for the assessment of RNA concentration and purity. Approximately 1 µg RNA was applied for synthesize cDNA using an ABI kit according to the manufacturer’s protocol. A specific stem-loop sequence was used for conversion of miR-155 to cDNA. ABL-1 expression level was checked using RQ-PCR on all samples as a control gene and as a determinator of cDNA quality. Expression levels of target genes were assessed using specific primers (Table 2) via Applied Biosystems 7500 Fast (ABI) real-time reverse transcription polymerase chain reaction (RT-qPCR) System. For each gene, assay optimization and data interpretation were performed by the instruction of Tania Nolan and colleagues [13].

Table 2.

Oligonucleotide primers

Genes	Primers	Sequence	TM	Amplicon size(bp)
LC3Av1	F	CGTCCTGGACAAGACCAAGT	59.6	181
	R	ACCAGGTCTACCACAGTCG	59.5
miR-155	F	GCTCAGTTAATGCTAATCGTG	55.2	67
	R	GAGGAAGAAGACGGAAGAAT	54.5
ABL-1	F	TGGAGATAACACTCTAAGCATAACTAAAGGT	62	124
	R	GATGTAGTTGCTTGGGACCCA	60
LC3Av1 Methylation	F	TGGGGTTGTAGTTTGGTT	53.3	270
	R	TCAAATCACAACATTCCTTAAAAA	53.9
LncRNA-GAS5	F	GTGTGGCTCTGGATAGCAC	57.9	214
	R	ACCCAAGCAAGTCATCCATG	58.2

DNA Extraction and Examination of LC3A Methylation

DNA extraction was performed using a QIAamp DNA Mini Kit (Qiagen, Hilden, Germany). DNA purity and concentration were measured by Nanodrop ND-1000 Spectrophotometer (NanoDrop Technologies). Finally, the beta actin gene was amplified on all samples for determination of the quality of extracted DNA. After confirmation of DNA quality, a methylation-sensitive high-resolution melting (MS-HRM) was performed to analyze the CpG island of LC3Av1 gene. The selected region for designing primer was previously proved to have control on gene expression level of LC3Av1. Primers designing (supplementary Table 1), set-up of PCR condition and interpretation of the results were performed under a protocol provided by Tomasz K Wojdacz and colleagues [14].

Translocation Analysis

Most frequent fusion genes in AML including RUNX1- RUNX1T1, PML-RARA, MLL rearrangements, CBF-bMYH11 and BCR-ABL1 were investigated using RQ-PCR. The procedure and data analysis were performed based on protocols published by J. Gabert and colleagues [15].

Statistical Analysis

Statistical analyses and graph designing were performed using the SPSS software (version 16.0) and the GraphPad Prism 6 software. Differences among multi-state variables were examined by One-Way ANOVA or Kruskal − Wallis. For two-state variables, t-test or Mann − Whitney U test were chosen. The selection of these tests was based on the distribution of data. Spearman rank test was performed for analysis of correlation. A two-tailed P value of less than 0.05 was considered statistically significant.

Results

Characteristics of Patient and Control Subjects

60 patients with de novo AML, including 37 male and 23 female, were analysed for LC3Av1 gene expression and its possible nuclear regulators. Data are reported in comparing with 20 subjects with normal hematopoietic system aged 4 years to 87 years (median = 41.5) including 12 male and 8 female cases. Fusion genes including PML-RARA, RUNX1- RUNX1T1, MLL rearrangements, CBF-bMYH11 and BCR-ABL1 were detected in 17, 4, 2, 2 and 0 patients, respectively. Overall, 25 patients (41.6%) were diagnosed with one of AML related chromosomal translocations. There were no statistically significant difference among gene expression of LC3Av1, GAS5, miR-155 and methylation intensity of LC3Av1 among genetically subtypes of AML. Difference and correlation analyses also depicted that gender, blast percent, leukocyte count, platelet count, hemoglobin level, CD34 and HLA-DR expression levels and immunophenotype subtypes had no relation with gene expression and methylation in the present study. Demographic characteristics of patients are summarized in Table 1.

LC3Av1 and miR-155 Were Aberrantly Expressed in AML Patients

Epigenetic changes accompanied by alterations of regulatory RNAs dramatically changed the expression of oncogenes and tumor suppressor genes (TSGs) in AML. In this regard, we assess the gene expression level of LC3Av1, GAS5 and miR-155 in a case–control study. When the 95% confidence interval of the control group was considered as threshold of normal gene expression, AML patients demonstrated 68.5% downregulation for LC3Av1 and 78% overexpression for miR-155 (Fig. 1). Fold change analysis depicted LC3Av1 expression levels decrease by 0.44 fold (P < 0.001) and miR-155 increased by 5.7 fold (P < 0.001) relative to control subjects. ROC curve analysis showed that miR-155 and LC3Av1 can discriminate AML from control subjects (miR-155, AUC = 0.86 at cut of point 0.82 and LC3Av1 AUC = 0.72 at cut off point 1.7). GAS5 expression level was comparable between the two groups. It is important to note that FAB subtypes of AML patients did not influence the gene expression level (Fig. 2) and methylation status of LC3Av1.

Fig. 1.

Gene expression pattern and the ROC curve analysis of assessed genes. LC3Av1 expression was significantly repressed, while miR-155 was significantly upregulated in AML patients. These genes have acceptable AUC for distinguishing AML from normal subjects. GAS5 expression was comparable to normal subjects

Fig. 2.

Expression pattern of selected genes among FAB subtypes. ANOVA test showed not significant differences for miR-155, LC3A and GAS5 among FAB subtypes while miR-155, LC3A were differently expressed in all FAB subtypes when comparing with control subjects( *** P < 0.001, NS not significant)

miR-155 and Aging Influence the Gene Expression Level of LC3Av1 in AML

A substantial body of literature now exists regarding the effect of aging and regulatory RNAs of autophagy level [16, 17]. Most investigators believe that the activity of autophagic decreases with aging [17]. But the role of aging in the expression of miR-155 is unravel. In the present work, correlation analysis revealed aging influences LC3Av1 in a significant negative manner while had a weak but even statistically significant positive effect on miR-155 gene expression (Fig. 3). On the other hand, miR-155 and LC3Av1 expression was opposite in AML samples. GAS5 expression level, had no relation with age and with the expression levels of other genes in AML patients.

Fig. 3.

Evaluation of the correlation of different studied parameters. LC3Av1 expression had a positive correlation with the age of subjects and negative correlation with the miR-155 gene expression. miR-155 expression had a negative correlation with the age of patients. No correlation was found between the GAS5 expression and LC3Av1 expression

CpG Island Hypermethylation Dramatically Decreased LC3Av1 Gene Expression

CpG islands are present at the promoter or even coding region of Up to 60% of protein-coding genes, and can modulate the transcription level base on its methylation status [14]. Many methylations’ patterns have been found to be tumor specific and in relation to disease prognosis [18]. Our results showed that 74% of AML patients have a methylated region in LC3Av1 gene, while just 15% of control subjects showed this pattern (Fig. 4, P < 0.001). Patients and control subjects with methylated CpG island depicted dramatically decreased in LC3Av1 (P < 0.001).

Fig. 4.

LC3Av1 methylation pattern. MS-HRM showed above 10% methylation in the CpG island of LC3Av1 in 74% and 15% of case and control subjects respectively (left part). LC3Av1 expression was significantly lower among both case and control subjects with hypermethylated CpG island

Discussion

The chance of AML treatment decreases drastically with age [19–21]. In AML, we reached a median survival of about 1 year for patients aged 65–69 years [22]. However, this seems like an amazing progress in the perspective of an experienced oncologist who knows that these patients had no treatment choice years ago. But in real life, this is a disaster for patients, especially when we noticed that the median age of AML patients is about 69 years at diagnosis. Even though the prognosis has improved for younger patients thanks to the intensive chemotherapy, the majority of those patients are also going to end up relapsing. In this regard, knowing the basic biology of AML may provide an opportunity for better risk stratification, outcome prediction, and introduction of novel therapeutic approaches. Genetic lesions per AML cells are of the lowest among human malignancies, but they mainly influence epigenetic regulators and transcription factors whereupon dramatically changed the cell transcriptome [2]. In this line, perturbed autophagy related genes expression was reported in AML patients and multiple studies have suggested a pathogenic role for this system and it was even introduced as a novel targetable process in AML treatment.

This initial success in AML and other malignancies led to the search for additional compounds that target autophagy or its regulators. What we have to know for a clinical translation of these achievements is the precise status of autophagy in a given cancer, molecules that are involved in autophagy regulation and the combination effect of autophagy modulators with routine treatments of the patients. In this line, we found that the expression level of autophagy marker LC3Av1 was significantly reduced in AML samples. In agreement, accumulating evidences argue in favor of tumor suppressor role of autophagy with evidence of monoallelicaly loss of Beclin1 in 75% of ovarian cancer cases and in 50% of breast cancer cases [23]. Zhu Han and colleagues showed that genetic mutations that altering selective autophagy represents a common mechanism for the pathogenesis of several malignancies. Furthermore, study of 244 cell lines revealed that transcription of LC3Av1 was inactivated in about 45.5% of human cancers [24]. However, genetic loss or gene mutations of ATGs are not among frequent aberrancies in AML, decreased expression of these genes is documented in some recent studies [25]. Moreover, in AML patients with unfavorable outcome, downregulated beclin 1 and p62 was accompanied with worse overall survival compared with counterparts with high expression values. Adding to this, a survey of BECN1 in 128 de novo AML patients showed reduced expression level in intermediate and unfavorable patients compared with favorable subgroup and healthy control [10]. Besides, an integrated bioinformatics analysis introduced a 12-autophagy gene signature that would provide prognostic significance and potential therapeutic targets in AML [26]. Finally, the prognostic application of LC3A expression level was also delineated in solid tumors, including gastric and colorectal cancers [27]. In contrast, there are also studies that represented an appreciably upregulation of ATGs mRNA expression in newly diagnosed or relapsed patients with AML compared with control group or patients in complete remission [12, 28]. For example, Parisa Tandel et al. noted that autophagy level reduced in complete remission [12], which is in sharp contrast with the study of Jing Jin et al., which suggests that restoring autophagic activity plays an important role in AML differentiation [9]. The synoptic conclusion from these studies is that selection of different ATGs, different AML subtypes and control groups, or the genetic background of different population may be involved in bias of these results. It’s also important to note that autophagy has a versatile nature that is in line with is goal of maintaining cellular hemostasis, and therefore its level may be varied in different stages of a malignancy [9]. On the other hand, it merits to note that we cannot directly translate gene expression level as autophagy activity, because autophagy activity relies upon cytoplasmic stimuli and inhibitors [5]. Therefore, in this context, an elevation in gene expression level of one of multiple ATGs should not be taken as increased autophagy function, but it must notice that downregulation of even one crucial ATGs could disrupt the entire performance of autophagy machinery [16].

It worth repeating that autophagy regulation is sophisticated, because it plays role in the adaptation of cells to a continuously changing milieu. Regarding the important role of this machinery in both cell death and cell survival, a better understanding of autophagy regulators may provide molecular targets that could be relevant to cancer therapy. In between, epigenetics which determine chromatin configuration and regulating its accessibility to transcription factors and modulating effect of regulatory RNAs act as fundamental nuclear signals in autophagy regulation [4]. On the other hand, in the cell cytoplasm, AMPK and mTORC1 are among the key regulatory signaling pathways [7]. Considering the scarcity of studies assessing the role of epigenetic and regulatory RNAs in alteration of autophagy, it is certainly worth much further investigation. In this context, we found that LC3Av1 CpG Island was appreciably methylated in AML patients and this phenotype was accompanied by a significant reduction in LC3Av1 gene expression. Recently, two review articles assess the role of epigenetic mechanisms in regulation of ATGs including LC3s. In summary, CpG island hypermethylation of a plethora of ATGs including LC3Av1, ATG16L2, GABARAPL1, ATG2B, ATG4D, ATG9A, ATG9B and so on, were pertinent to progression or prognosis of multiple cancers [8]. Meanwhile, H Bai and colleagues showed a decreased in LC3Av1 expression, in esophageal squamous cell carcinoma cell lines and primary tumors [24]. They attributed this phenomenon to DNA hypermethylation and revealed that reversing this condition led to inhibition of tumor growth. Regarding rare occurrence of genetic mutation in ATGs it seems that alteration of epigenetic and regulatory RNAs, as opposed to direct genetic alteration, may be an alternative molecular mechanism in dysregulation of autophagy and many of leukemia related genes in AML patients.

Besides difference in methylation of LC3Av1, our results demonstrated a dramatically increased in miR-155 gene expression that was negatively correlated with LC3Av1 expression level. miR-155 overexpression has been proven to be widely associated with prognosis in various hematological disorders. In this line, Tarek Elgohary et al. represented that miR-155 overexpression was significantly associated with the patient’s outcomes, including poor cytogenetics, treatment resistance and shorter survival [29]. These findings were corroborated in the study of L–H Xu mir-155 83 pediatric with AML [30]. Alteration of autophagy appears to be one of miR-155 tools in the progression of cancers, inflammation, and infectious diseases. In Acute pancreatitis, miR-155 enhanced inflammatory responses through impairing autophagy machinery [31]. Mycobacterium tuberculosis inhibits autophagy via repression of ATG3 mediated by the induction of miR-155 [32]. A great deal of evidence is already in hand that explores the role of miR-155 in cancer progression or treatment response by affecting autophagy machinery [33–36]. Regulatory effect of miR-155 on ATGs was reported by Emam, A. A. et al. These investigators examined the plausible anti-cancer mechanisms of exopolysaccharides (EPSs) on hepatocellular carcinoma cell line (HEPG II), breast cancer cell line (MCF-7), and colorectal carcinoma (Caco2). They concluded that EPSs dramatically decreased miR-155 expression level and by this means, influence autophagy marker LC3A and autophagy regulator mTOR expression level [37]. Adding these, we found that the expression level of miR-155 significantly decreases with the age in AML patients, while LC3Av1 was expressed in opposite manner. Overexpression of LC3Av1 in the elderly patients with AML is in contrary to the decreased autophagic activity that has been seen in almost all cells and tissues as organisms age [38]. However, our data are explainable with study of Kang C and Elledge SJ which revealed that autophagy can both trigger and impede cellular senescence [39]. Regarding the analysis of correlation needs larger study population, further investigations of autophagy using functional analysis will be required to establish the exact contribution of aging and autophagy in the AML patients.

We also assess the expression level of GAS5, since it was contributed to both AML pathogenesis and autophagy [40]. Our results demonstrate that GAS5 expression level in AML was comparable with control subjects. Furthermore, we didn’t see any correlation between GAS5 expression and autophagy marker LC3Av1. Given discrepancies between studies that examined the role of GAS5 in AML pathogenesis and prognosis, the final conclusion about this gene could be explained only by larger studies.

Conclusion

The present study revealed that epigenetic changes like DNA methylation and alteration of miR-155 have a pivotal role in repression of autophagy marker LC3Av1, which potentially could provide the important clues of prognostic and therapeutic targets. The optimal strategies for clinical implementation of autophagy in AML is yet to be fully achieved and require further studies.

Supplementary Information

Below is the link to the electronic supplementary material.

Declarations

Competing Interests

No funding was received for conducting this study.