Copy Number Alterations in CDKN2A/2B and MTAP Genes Are Associated With Low MEF2C Expression in T-cell Acute Lymphoblastic Leukemia

The molecular heterogeneity of T-cell acute lymphoblastic leukemia (T-ALL) makes this disease complex. Early T-cell precursor ALL (ETP-ALL) is a recognized subtype of T-ALL associated with a high probability of induction failure with conventional therapy. Higher expression of myocyte enhancer factor 2C (MEF2C) and the absence of a biallelic deletion (ABD) are the designated markers for the ETP-ALL. Co-deletion of the contiguous genes cyclin-dependent kinase inhibitor 2A/2B (CDKN2A/2B) and the methylthioadenosine phosphorylase (MTAP) cluster, located at 9p21.3, is another common alteration in T-ALL and confers poor response to treatment. We used real-time polymerase chain reaction (PCR) analysis to assess MEF2C mRNA expression and ABD status. Copy number alterations (CNAs) in key genes previously reported to be altered in T-ALL were assessed using multiple ligation probe amplification (MLPA). We observed that CNAs in this co-deletion cluster of CDKN2A/B and MTAP genes exhibited low MEF2C expression while ABD was associated with CNA in the Abelson murine leukemia 1 (ABL1) gene. Assessment of MEF2C expression based on immunophenotype revealed that its association with CDKN2A/2B alteration is present in non-immature immunophenotype. Additionally, ABD was associated with copy number alterations of T-cell acute lymphocytic leukemia protein 1 (TAL1), myeloblastosis (MYB), and LIM domain only 2 (LMO2) genes in immature immunophenotypes. Further, STIL::TAL1 fusion was associated with low expression of MEF2C. These associations may help explain the difficulties in assessing disease heterogeneity and the prognostic importance of 9p21.3 alterations in T-ALL.


Introduction
T-cell acute lymphoblastic leukemia (T-ALL) is an aggressive malignancy that can be stratified into various heterogeneous molecular subgroups [1,2]. These subgroups can be recognized by their characteristic immunophenotypes and by distinctive gene expression profiles. Myocyte enhancer factor 2C (MEF2C) is an oncogene belonging to the MADS-box family of transcription factors, which consists of four members: MEF2A, B, C, and D, and plays a critical role in embryonic development [3,4]. MEF2C is expressed at the onset of cardiac-skeletal muscle lineage differentiation, and MEF2C knockout mice are embryonically lethal at the E9.5 stage of development [5]. Further, MEF2C also plays a crucial role in the normal hematopoietic system, particularly in the production of immature and mature lymphocytes [4]. In the context of its role in pathophysiology, it has been widely reported to be abnormally expressed in immature subsets of T-ALL, particularly in early thymocyte precursor (ETP) ALL, an entity that has been associated with adverse risk characteristics and poor outcomes in T-ALL [6][7][8][9]. Studies have shown that MEF2C acts as a transcriptional regulator for genes that are expressed in the immature subgroup of T-ALL, including lymphoblastic leukemia-derived sequence 1 (Lyl1 ) and LIM Domain Only 2 (LMO2) [10]. One group characterized another identifier for the immature subgroup, the absence of biallelic deletion (ABD) of the T-cell receptor (TCR) gamma gene locus, representing an early maturation arrest before the onset of T-cell receptor rearrangements [11]. Previously, our study suggested the association between ABD and high MEF2C expression [12]. Further, high MEF2C expression was associated with poor overall and event-free survival while ABD exhibited no significant association with patient overall and event-free survival. In addition to transcriptional regulators, there are other factors, such as somatic genetic abnormalities, which can function as initiators and/or drivers of leukemia progression [13].
The genetic landscape of T-ALL and its clinical utility along with the established factors is poorly understood 1 and needs to be studied more comprehensively. Targeting transcription factors pharmacologically has proven to be extremely difficult. Therefore, it is important to integrate other ways to target key signaling pathways, such as neurogenic locus notch homolog protein 1 (NOTCH1), anti-apoptotic signaling pathways, and genetic aberrations and clinical risk factors, to improve prognostication in T-cell precursor acute lymphoblastic leukemia [10,14]. In the current study, we attempted to integrate both transcriptional and genetic aspects of disease and established an association between these factors. Copy number alterations (CNAs), which are somatic modifications to the chromosome structure that cause the addition or deletion of copies of certain DNA segments, are common in many cancer types [15,16]. CNAs are prevalent in all forms of leukemia, including T-ALL, particularly the genes involved in transcription, cell cycle control, and T-cell differentiation [15,17]. According to earlier research, the T-ALL genome frequently demonstrates CNAs in NOTCH1 and cyclin-dependent kinase inhibitor 2A/2B (CDKN2A/2B) cell cycle regulators. It has been proposed that CNAs might be useful in improving the T-ALL risk stratification. We previously determined the frequency of key CNAs in a cohort of Indian T-ALL patients using Multiplex Ligation-dependent Probe Amplification (MLPA) [18]. Additionally, we determined the association of CNAs with patient prognosis with a comparative analysis of the existing literature. In order to define the clinical utility of CNAs, it is important to determine their association with other molecular alterations and prognostic markers such as immunophenotype and gene expression profile. In the current study, we analyzed data from 88 T-ALL patients to determine the association of key CNAs with poor-risk indicators ABD and MEF2C mRNA expression.

Patient samples and treatment protocol
We reassessed the data from our previously published studies [12,15] and included patients for whom both MEF2C expression level and status of copy number alteration were available. Briefly, MEF2C expression and copy number alteration were determined in 88 patients (age two months to 68 years) who were diagnosed with T-lineage ALL. Berlin-Frankfurt-Muenster98 (BFM98) protocol was used to treat adult T-ALL patients (age ≥18 years), and the Indian Childhood Collaborative Leukemia group protocol (ICiCLe) was used to treat pediatric T-ALL patients (age <18 years) [19]. After giving these individuals steroid prophase for the first seven days, the absolute blast count was assessed on day eight to assess their response to prednisolone. Based on the absolute blast count in peripheral blood (PB) on day eight of medication beginning, the patients were divided into two groups: prednisolone good responders (PGR) and prednisolone poor responders (PPR). As stated in our prior paper, postinduction bone marrow (BM) samples were used to determine the minimal residual disease (MRD). Patients were classified as either MRD positive (MRD >0.01%) or MRD negative (MRD <0.01%) [18]. BM and/or PB samples were collected at the Department of Medical Oncology, Dr. BR Ambedkar Institute Rotary Cancer Hospital (BRA-IRCH), All India Institute of Medical Sciences, New Delhi, India. The current study was approved by the institutional ethics committee, following the Declaration of Helsinki. All patients or guardians/parents gave written informed consent to the use of leftover diagnostic material for research purposes.

Immunophenotyping
All patient samples were collected in ethylenediamine tetraacetic acid (EDTA) vials and immunophenotyping was performed using the stain-lyse-wash method. The immature group was defined by the presence of cytoplasmic CD3, CD2, and CD5+/-and the absence of both CD4 and CD8; the cortical group was defined by the presence of a CD1a marker and the mature group was defined by the presence of a CD3 surface expression and positivity for one of the CD4 or CD8 markers. The diagnosis of ETP was based on previously defined criteria: CD5 weak/negative, CD1a negative, CD8 negative, and expression of one or more myeloid and stem cell markers [9].

Patient characteristics
A total of 88 newly diagnosed T-ALL patients were included in the study. Of these 88 newly diagnosed T-ALL samples, MEF2C expression was examined at the mRNA level while copy number change and ABD in DNAs from newly diagnosed patients were examined. There were 76 males and 12 females, which is in agreement with previous reports [20].

Measurement of MEF2C expression
Total RNA was isolated from the T-ALL-BM/PB samples using TRIZol (Thermo Fisher Scientific, Waltham, Massachusetts) reagent according to the recommendation of the manufacturer and reverse transcribed to cDNA using RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific). MEF2C expression was measured by real-time PCR using the same primer and probe as used previously ( Table 1) [12]. β-actin, ABL1, and GPI genes were used as housekeeping genes to normalize the real-time

ABD assessment of TCR gamma chain
Total DNA was extracted using a DNA extraction kit from Thermo Fisher Scientific. The TCR deletion was determined as described previously [11]. The expression level of TCR was quantified using real-time PCR as a fold change of the anillin, actin binding protein (ANLN) gene compared to T cell receptor gamma-V (variable) and J (joining) gene segments (TCRG-VJ ) expression. Sequences of primers have been given in Table 1 [12]. For ABD assessment, patients were assigned to the ABD group based on the ratio of TCR: ANLN, which was > 0.5 and a high blast count of approximately >85%, and patients with a change of <0.35 fold were assigned to the non-ABD group [11].

CNA assessment using MLPA reaction and analysis
MLPA reaction method has been previously published [18]. Briefly, copy number changes were detected using the SALSA MLPA probe mix P383-A2 (MRC Holland, The Netherlands). This kit was used to detect changes in signaling (PTEN, NF1, and PTPN2), cell cycle (CDKN2A, CDKN2B, and CASP8AP2), transcription factors (LEF1 and MYB), and epigenetic regulator genes ( EZH2, SUZ12, and PHF6) and in addition to the identification of STIL::TAL1 and NUP214::ABL1 gene fusions. All reactions were performed in a thermal cycler with a preheated lid. 100 ng of DNA was diluted in 5 µl of water and heated to 98°C for 5 min, followed by the addition of probe mix, hybridization, and amplification. The amplified product was then subjected to capillary gel electrophoresis on an ABI Genetic Analyzer (Thermo Fisher Scientific). Data analysis was performed using the Coffalyser.Net software (MRC Holland, The Netherlands) according to the manufacturer's recommended protocol. Patients were divided into five groups for each probe set, as no copy number changes (0.80 to 1.20), heterozygous or homozygous deletions (0.40 to 0.65 and 0.00, respectively), and hetero-and homozygous duplication (0.30 to 1.65 or 1.75 to 2.15, respectively) were noted based on the final probe ratio obtained by normalizing the sample signal to the control signal.

Statistical analysis
The nonparametric Mann-Witney U test for continuous variables and Fisher's exact test for categorical variables were used to examine the associations. All analyses were performed using STATA version 20 (StataCorp LLC, College Station, TX), SPSS (IBM Corp., Armonk, NY), and Graphpad version 8 statistical software (Graphstats Technologies Private Limited, Bengaluru, Karnataka, India). The mean of continuous variables among clinicopathological features was described as the mean ± standard deviation (SD). A p-value of 0.05 (two-tailed) was considered statistically significant.

Association of MEF2C expression with copy number alterations
To determine the association of MEF2C expression and ABD with CNA in our patient cohort, we divided patients into two groups, MEF2C high and MEF2C low, based on the median expression. Higher MEF2C expression was associated with lower frequency of CNAs in CDKN2A (36.36% vs 84.09%, p<0.0001), CDKN2B (25% vs 72.73%, p<0.0001), and MTAP (13.64% altered vs 47.73% unaltered, p=0.001) ( Figure 1,

FIGURE 1: Comparison of MEF2C mRNA expression between T-ALL patient's groups based on copy number alterations of studied genes
The box plot shows Tukey plots without outliers. The Mann-Whitney U test was applied for comparing respective groups based on the alteration status of genes. The level of significance is denoted as ***p<0.001, **p<0.01, *p<0.05.

TABLE 2: Association of MEF2C expression with the key copy number alterations in total T-ALL cases
We further assessed the association of MEF2C expression with CNAs in the immature group separately. This suggested that the MEF2C high group was not associated with alterations in the copy number of any of the genes assessed (Figure 2A). In non-immature cases, i.e., cortical and mature, low MEF2C expression was observed in cases with CNA in CDKN2A (92.59% vs. 55.56%, p=0.004) and CDKN2B (81.48% vs. 37.04%, p=0.002) while no association was observed with CNAs of other assessed genes ( Figure 2B, Table 3). The box plot shows Tukey plots without outliers. The Mann-Whitney U test was applied for comparing respective groups based on the alteration status of genes. The level of significance is denoted as ***p<0.001, **p<0.01, *p<0.05.

Association of ABD with copy number alterations
The results of Fisher's exact test showed that the ABD group was significantly associated with less ABL1 CNA (24% vs 6.35%, p=0.028, Table 4). ABD was not associated with the copy number alterations of genes involved in signaling (PTEN, NF1, and PTPN2, p=1.     The box plot shows Tukey plots. The Mann-Whitney U test was applied with the level of significance denoted as ns, not significant, and *p<0.05.

Discussion
Despite the diversity of genetic abnormalities and gene expression profiles in T-ALL, a limited number of subtypes have been proposed for risk stratification, which is primarily based on distinct expression levels of a few transcription factors. These expression profiles include the TAL/LMO, TLX1, TLX3, and HOXA clusters as well as the recently reported proliferative and immature T-ALL subtypes, which are mediated by the overexpression of NKX2-1, NKX2-2, and MEF2C. MEF2C functions as a transcriptional regulator, triggering an exhaustive transcriptional program characteristic of early T-cell progenitor (ETP) ALL. Furthermore, the immunophenotypically characterized ETP-ALL overlaps with the immature MEF2C-expressing cluster, suggesting that both may represent a single disease entity. The absence of biallelic deletion (ABD) also represents ETP-ALL and is associated with adverse risk features and patients outcome [12,21]. In a previous study, we showed that ABD and high MEF2C expression are associated with the immature immunophenotype [12]. Further, we also observed an association of MEF2C overexpression with a poor response to prednisolone and MRD positivity after induction, suggesting the use of MEF2C expression as a surrogate for true transcriptionally defined ETP-ALL [12]. In this study, we investigated whether ABD and MEF2C expression are associated with the common copy number alterations in T-ALL. We observed that copy number alterations in CDKN2A/B and MTAP, which are otherwise widely established to be associated with cortical/mature immunophenotype, are associated with a low MEF2C expression. While we considered it as an obvious association, keeping in mind that MEF2C expression is indeed enriched in immature phenotype, our subgroup-specific analysis revealed that the abovementioned association was not present in the immature group but in the non-immature group, thus making a possibility of patient stratification specifically in non-immature T-ALL cases.
Cyclin-dependent kinase inhibitor 2A/B (CDKN2A/B) genes at chromosomal arm 9p21 are frequently altered in patients with acute lymphoblastic leukemia (ALL) [22]. Inactivation of these genes by deletion, mutation, and promoter methylation can lead to the malignant transformation of tumor cells and induces chemoresistance in a variety of malignancies [23]. CDKN2A/2B and MTAP deletion frequency were the highest among the genes analyzed and their codeletion was also previously reported in solid cancers and hematological malignancies [24,25]. In T-ALL, CDKN2A/2B frequently exhibits deletion and promoter hypermethylation leading to their downregulation, which is further associated with poor clinical outcomes [26,27]. We and others have previously shown that the frequency of CDKN2A deletion was the highest among all gene deletion/duplication evaluated and has variable co-deletion of contiguous genes, including CDKN2A/B and MTAP present at 9p21.3 cluster [24].
Inactivation of CDKN2A/B and other translocations affecting the T-cell receptor genes provide initial insights for genetic defects present in T-ALL. In the context of the biomarker potential, common copy number alterations, such as CDKN2A/2B, are easily detectable in clinics by MLPA. 9p21 deletion and alteration of genes localized on this locus lead to a malignant proliferation of tumor cells and play a crucial role in the pathogenesis and drug resistance of ALL [24,25]. Interestingly, our recent findings also suggested that CDKN2B and MTAP copy number changes were associated with poor overall patient survival and CDKN2B emerged as an independent prognostic factor for poor overall survival and event-free survival [18]. CDKN2A alteration has been associated with aberrant MYC oncogene activation and novel therapies are under trials to target MYC-driven T-ALL [28]. Therefore, considering our observation that patients with CDKN2A alteration exhibit low MEF2C expression, it will be interesting to investigate whether this association will impact the response to targeted therapy in the future. If any such effect is observed, MEF2C expression may also serve as an additional prognosticator. Further, we reported earlier that MTAP copy number alterations also act as an independent prognostic factor for patient survival [18]. MTAP gene codes for methylthioadenosine phosphorylase, which is involved in purine and methionine utilization metabolism. MTAP alterations lead to the deficiency of the salvage pathway for purine synthesis. Some studies suggest that drugs that interfere with purine metabolism, such as methotrexate, might prove beneficial in the treatment of MTAP alteration associated with T-ALL. It has been recently shown that MEF2C opposes NOTCH-mediated T cell maturation program in the thymus [29]. Therefore, Salt-inducible kinase (SIK) inhibitors, which impair MEF2C activity have been proposed to target MEF2C expressing ETP-ALL [29]. While SIK inhibitors have been shown to increase prednisolone sensitivity in ETP-ALL cell models, it will be interesting to assess in the future whether this inhibitor can also target a subgroup of patients with non-immature immunophenotype who express high MEF2C and does not possess alterations in CDKN2A/B.
Interestingly, we also observed the association of ABD with the alteration in the ABL1 gene, but no association was found with its fusion NUP214::ABL1. Activation of the ABL1 gene along with the NUP214::ABL1 fusion is associated with the TLX1 and TLX3 subgroup [30] while ABD is an established marker of the ETP-ALL subgroup of T-ALL [11]. As we showed in our previous report, ABD and MEF2C expression exhibited no association with age, sex, leukocyte count, CNS involvement, and mediastinal mass but with the immature subgroup [12]. ABD shows no association with ETP immunophenotype and prednisolone response while these clinical features were associated with high MEF2C expression. In addition to the association of MEF2C expression with common copy number alterations, its lower expression in the STIL::TAL1 fusions was also observed.
This work implies that evaluating MEF2C expression along with CNAs may be useful in identifying the patient subgroups with a poor prognosis. However, no positive association of MEF2C expression with CNAs was observed, which may limit the translation of this association into the clinics. The fact that this study was conducted in the Indian population may limit its applicability to other populations because there might be variations in the molecular profiles depending on ethnicity. In order to use genetic changes in T-ALL as a risk stratification factor in the clinics, additional research is therefore urgently required.

Conclusions
The current study demonstrates that low MEF2C expression is associated with the alterations of the CDKN2A/2B and MTAP genes. Taken together, our study demonstrated the association of ABD and MEF2C expression with copy number alterations in the CDKN2A/2B and MTAP cluster in T-ALL. Considering the potential impact of MTAP and CDKN2A/2B deletion and high MEF2C expression on chemosensitivity, it would be interesting to explore the detailed functional role of these associations, which might help address the difficulties in predicting prognosis in this malignancy.