Characterization of CK2, MYC and ERG Expression in Biological Subgroups of Children with Acute Lymphoblastic Leukemia

Abstract

Despite the excellent survival rate, relapse occurs in 20% of children with ALL. Deep analyses of cell signaling pathways allow us to identify new markers and/or targets promising more effective and less toxic therapy. We analyzed 61 diagnostic samples collected from 35 patients with B- and 26 with T-ALL, respectively. The expression of CK2, MYC and ERG genes using Sybr-Green assay and the comparative 2-ΔΔCt method using 20 healthy donors (HDs) was evaluated. We observed a statistically significant difference in CK2 expression in non-HR (p = 0.010) and in HR (p = 0.0003) T-ALL cases compared to HDs. T-ALL patients with PTEN-Exon7 mutation, IKZF1 and CDKN2A deletions showed high CK2 expression. MYC expression was higher in pediatric T-ALL patients than HDs (p = 0.019). Surprisingly, we found MYC expression to be higher in non-HR than in HR T-ALL patients. TLX3 (HOX11L2)-rearranged T-ALLs (27%) in association with CRLF2 overexpression (23%) showed very high MYC expression. In B-ALLs, we detected CK2 expression higher than HDs and MYC overexpression in HR compared to non-HR patients, particularly in MLL-rearranged B-ALLs. We observed a strong difference in ERG expression between pediatric T- and B-ALL cases. In conclusion, we confirmed CK2 as a prognostic marker and a therapeutic target.

1. Introduction

Substantial advances have been made in the past five decades in the treatment of patients with acute lymphoblastic leukemia (ALL), the most common malignancy in childhood. In contemporary ALL treatment regimens, patients are stratified into different risk groups based on clinical and biologic characteristics at presentation as well as on early treatment response [1]. Improved understanding of the biologic heterogeneity of ALL and the development of sensitive minimal residual disease (MRD) response-monitoring techniques have improved modern risk stratification for childhood ALL [1]. The majority of patients with B-lineage ALL have somatic aneuploidy or recurrent chromosomal translocations, many of which have prognostic significance [2]. Conversely, the clinical significance of most recurrent genomic alterations in T-lineage ALL is less clear, and current clinical trials stratify patients with T-ALL based on their response to 8 days of prednisone administration or detection of MRD during the induction phase [3,4]. Despite excellent survival, exceeding 90%, relapse occurs in 15% to 20% of children with ALL, which remains the first cause of death in children with cancer [2]. Early relapses of both B- and T-ALL are a current challenge because of the high rate of second-line treatment failure. For this reason, analyses of significantly enriched signaling pathways might help to identify new markers for a better patient stratification and/or targets for a potential future tailored therapy. Pathways including WNT/β-Catenin, p53 and PI3K/AKT/PTEN with ERG overexpression may contribute to the dysregulation of kinase signals such as Casein Kinase 2 (CK2), which results in resistance to kinase inhibitors [5,6,7,8]. CK2 overexpression has been observed in hematological malignancies such as acute and chronic leukemias, including T- and B-ALL and acute myeloid leukemia (AML) [6,8]. Elevated levels of CK2α are highly correlated with poor clinical outcome [8]. An important CK2 target is IKAROS (IKZF1), a transcription factor that displays crucial functions in the hematopoietic system and controls the development in early B- and T-cells. Mutations that lead to reduced IKZF1 function or expression have also been found to be a major genetic feature in human B-ALL [9] and loss of IKZF1 also promotes the development of T-cell lymphoma/leukemia in mice, which suggests that IKZF1 acts as a tumor suppressor in T- and B-ALL [9,10]. Indeed, despite its low prevalence, IKZF1 loss of function is clearly a recurrent anomaly in human T-ALL [10]. Thus, IKZF1 inactivation could play a causal role in disease progression in a subset of T-ALL cases, influencing the outcome [10]. Another proto-oncogene closely involved in many cancers, including leukemia, is MYC. Aberrant expression of MYC in leukemia results in an uncontrolled rate of proliferation and, thereby, a blockade of the differentiation process [11]. MYC is not activated by mutations in the coding sequence, but its overexpression in leukemia is mainly caused by gene amplification and aberrant regulation of its transcription [11,12]. A tightly controlled increase in MYC expression is required for differentiation, but prolonged and excessive MYC activity is oncogenic and leads to increased proliferation, altered adhesion and chromatin remodeling [12,13]. Synergistic interactions between CK2 and MYC might lead to lymphocytes’ transformation; indeed, CK2 inhibitors, acting as an IKZF1 activator, suppress c-MYC in an Ikaros-dependent manner in ALL cells [14]. It has been demonstrated that in adult ALL patients, high c-MYC expression correlates with clinical high-risk factors and high proliferation markers [14]. Another useful marker in pediatric leukemia is the Ets-related gene ERG (erythroblastosis virus E26-transforming sequence family member) which has an important role in early hematopoiesis and hematopoietic stem cell (HSC) maintenance [15]. ERG is preferentially and strongly expressed in the immature B- and T-lymphoid lineages, in addition to myeloid lineage cells. Previous studies suggest that ERG overexpression is associated with inferior clinical outcome [15]. In T-ALL patients, a high level of ERG expression has been associated with poor relapse-free survival [15]. Conversely, in childhood B-ALL patients, a higher white blood cell (WBC) count, higher relapse rate and poor relapse-free survival rate were associated with low ERG expression [16]. The expression of CK2 in association with MYC and ERG has not yet been characterized in biological subgroups of pediatric ALL. The characterization of these genes would integrate new markers of disease and/or therapeutic targets in ALL. The cooperating activities of the selected genes are also supported by a recent article demonstrating that ERG and c-MYC coordinate a regulatory network in BCR::ABL1 B-ALL, controlling the expression of genes involved in several biological processes [17].

2. Results

2.1. Expression of CK2, MYC and ERG in T-ALL Cases

We preliminarily screened 26 patients with T-ALL for diagnostic molecular characterization and we found 5 cases with PTEN-Exon7 mutation (19%), 3 with PICALM/MLLT10 rearrangement (11.5%), 7 with TLX3 (HOX11L2) alteration (27%), 1 with mutated TP53 (4%), 16 with CDKN2A deletion (61.5%) and 6 with CRLF2 overexpression (23%); additionally, 3 out of 20 patients exhibited IKZF1 deletion (15%), respectively (Supplementary Table S1). Firstly, we compared CK2 expression in T-ALL cases to HDs and the CEM cell line: we observed a similarly high expression between T-ALL and CEM (Supplementary Figure S1) and a statistically significant difference in T-ALL patients (more in HR than in non-HR cases) vs. HDs (Figure 1). We point out that 10 out of 11 HR-T-ALL patients (90%) presented CK2 mRNA expression up to five times higher than that of HDs and two times higher than other T-ALL patients. Moreover, four out of five PTEN Exon7 mutated cases showed a high CK2 mRNA level (Supplementary Figure S2). Furthermore, we correlated CK2 expression among cases with CDKN2A, IKZF1 deletions and CRLF2 overexpression, respectively. We found out that children with IKZF1 and CDKN2A deletions showed high CK2 mRNA levels, although they were not statistically significant (Supplementary Figure S3A,B). We did not find any difference in CK2 expression among cases with CRLF2 overexpression compared with normal CRLF2 expression (Supplementary Figure S3C).

MYC expression in the CEM-cell line could be used to perform an exact comparison with T-ALLs. We observed a statistically significant difference in MYC expression in children with T-ALL compared to HDs (Figure 2A): 13 out of 26 T-ALL patients were identified as MYC-high (50%), and of these, 5 were HR and 8 were non-HR, respectively. Seven out of eight MYC-high non-HR patients showed a TLX3 (HOX11L2) rearrangement. Overall, we observed statistically higher MYC expression in non-HR T-ALLs than in HR T-ALLs (Figure 2B). In PTEN-Exon7-mutated patients, we observed low expression of MYC (Supplementary Figure S4). Moreover, we compared MYC expression with CDKN2A and IKZF1 status and we did not observe any statistically significant difference in MYC expression (Supplementary Figure S5A,B). By contrast, in T-ALL cases with high CRLF2 expression, we detected very high MYC expression compared to cases with normal CRLF2 expression (Supplementary Figure S5C). Five out of the six cases with CRLF2 overexpression had a TLX3 rearrangement.

ERG expression in the CEM cell line was very low, compared to HDs and T-ALL cases (Supplementary Figure S6), but statistically significant different expression was observed in children with T-ALL vs. HDs (Figure 3A); this was more prevalent in HR than in non-HR cases. However, the difference in ERG expression between the two T-ALL subgroups was not statistically significant (Figure 3B). When we analyzed the five T-ALLs with PTEN Exon7 mutations, they showed slightly higher levels of ERG mRNA than PTEN Exon7 wild type patients (Supplementary Figure S7).

Moreover, we studied thymocytes as normal internal control for CK2, MYC and ERG expression compared with HDs. CK2 expression was absolutely comparable between thymocytes and HDs. Instead, MFC MYC expression in thymocytes was comparable to T-ALL samples. This is due to an overlapping metabolic rewiring in activated T-cells and MYC-transformed lymphocytes, similar to that which occurs in tumor cells (with the activation of proto-oncogenes such as c-Myc) [16]. Erg expression was comparable in thymocytes and HDs, too (Supplementary Figure S8 and Table S3).

2.2. Expression of CK2, MYC and ERG in B-ALL Cases

We screened 35 B-ALL patients, who were stratified into four biological subgroups according to the most common molecular alterations detected at diagnosis of pediatric B-ALL (Supplementary Table S2). We performed CK2 expression analysis, observing a statistically significant difference in CK2 expression in B-ALL patients with respect to HDs (Figure 4A). The CK2 expression was homogenous among the four selected biological ALL subgroups, as shown in Figure 4B.

As for CK2 expression, we also found a statistically significant difference in MYC expression with respect to HDs (Figure 5A). Interestingly, we observed a correlation between MYC expression with a prognostic genetic subgroup: in the Ph+ cases, MYC expression was high, whereas in ETV6::RUNX1, the rate of positive B-ALL was very low. In the latter subgroup, two cases, who subsequently presented a late relapse, showed a higher value of MYC expression. Moreover, in seven out of seven HR patients in the “B-others” subgroup, we detected high MYC expression; this was also observed in four cases included in the MLL-R B-ALL subgroup (Figure 5B).

In cases with B-ALL, we also observed a high level of ERG expression compared with that of HDs (Figure 6A). These values were higher than ERG expression findings in T-ALLs. Among different biological subgroups, ERG expression was higher in ETV6::RUNX1-positive ALL patients than in BCR::ABL1, B-others and, in particular, MLL-R B-ALL patients (Figure 6B).

This inverse correlation with clinical risk was confirmed by the fact that 84% of HR B-ALL cases showed low expression of ERG, as we report in

Table 1. CK2, MYC and ERG expression according to the Final risk group in children with T- and B-ALL.

		CK2 Expression		MYC Expression		ERG Expression
		Low n (%)	High n (%)	Low n (%)	High n (%)	Low n (%)	High n (%)
T-ALL	Non-HR (n = 15)	9 (60)	6 (40)	7 (47)	8 (53)	12 (80)	3 (20)
	HR (n = 11)	1 (9)	10 (91)	6 (54)	5 (46)	2 (19)	9 (81)
B-ALL	Non-HR (n = 16)	7 (44)	9 (56)	12 (75)	4 (25)	6 (38)	10 (62)
	HR (n = 19)	7 (37)	12 (63)	2 (21)	17 (79)	16 (84)	3 (16)

, which depicts the CK2, MYC and ERG expression according to the Final risk group in children with T- and B-ALL. Instead, CK2, MYC and ERG expression according to genetic alterations is summarized in

Table 2. CK2, MYC and ERG expression according to genetic alterations in children with T- and B-ALL.

Genetic Alteration		Analyzed Patients	CK2 Expression		MYC Expression		ERG Expression
			Low n (%)	High n (%)	Low n (%)	High n (%)	Low n (%)	High n (%)
BCR/ABL1		B-ALL (n = 8/35)	3 (37)	5 (63)	1 (12)	7 (88)	6 (75)	2 (25)
ETV6/RUNX1		B-ALL (n = 16/35)	8 (50)	8 (50)	14 (88)	2 (12)	6 (38)	10 (62)
MLL rearranged		B-ALL (n = 4/35)	2 (50)	2 (50)	-	4 (100) °	4 (100) a	-
PTEN Exon7 Δ or inactivating mutations		T-ALL (n = 5/26)	2 (40)	3 (60)	5 (100)	-	1 (20)	4 (80)
PICALM/MLLT10		T-ALL (n = 3/26)	2 (67)	1 (33)	1 (33)	2 (67)	1 (33)	2 (67)
TLX3/HOX11L2		T-ALL (n = 7/26)	3 (43)	4 (57)	-	7 (100) °	5 (71.5)	2 (28.5)
O T H E R S	CDKN2A Δ	B-ALL (n = 2/7)	2 (100)	-	-	2 (100)	2 (100)	-
		T-ALL (n = 16/23)	8 (50)	8 (50)	7 (44)	9 (56)	8 (50)	8 (50)
	IKZF1 Δ	B-ALL (n = 2/7)	2 (100)	-	-	2 (100)	2 (100)	-
		T-ALL (n = 3/20)	-	3 (100)	2 (67)	1 (33)	-	3 (100)
	hi-CRLF2	B-ALL (n = 2/7)	1 (50)	1 (50)	-	2 (100)	1 (50)	1 (50)
		T-ALL (n = 6/25)	3 (50)	3 (50)	1( 17)	5 (83)	4 (67)	2 (33)

^a Very low ERG expression; ° very high MYC expression.

.

3. Discussion

Due to the high-risk disease characteristics and significant toxicity associated with chemotherapy, the outcome for ALL patients is less encouraging for defined subgroups of patients, including both standard and high-risk groups [18]. Novel targeted therapies offer the promise of effective anti-leukemic activity with reduced toxicity, but given the different molecular and genetic alterations that occur in ALL patients, it is improbable that a single agent will be effective for all patients. For this reason, it is mandatory to identify patient-specific therapy [18]. Moreover, some subtypes of this condition, such as T-ALL, despite the numerous genetic aberrations involved, lack reliable markers for precise stratification and/or potential targeted therapy [19]. Based on these considerations, we focused on studying three important genes involved in the main metabolic pathways of pediatric ALL [7]. The expression of CK2 in association with MYC and ERG overexpression has not been characterized in biological subgroups of pediatric ALL, yet. Here, we demonstrated that CK2 is overexpressed in pediatric ALL, resulting in a potential marker and, moreover, in a stable therapeutic target. Both T- and B-ALLs show higher levels of CK2 expression compared with HDs. Moreover, in T-ALLs with HR features, we showed a median CK2 expression up to five times higher than that of HDs. CK2 is a regulatory serine/threonine kinase which is ubiquitously expressed, and its activity is required for activation of pro-survival pathways [20]. Several pathways (WNT/β-Catenin, p53 and PI3K/AKT/PTEN with ERG overexpression) may contribute to the dysregulation of kinase signals such as CK2, which results in resistance to kinase inhibitors [8]. The importance of phosphorylation makes protein kinases and phosphatases promising therapeutic targets for a wide variety of human disorders [6]. Several studies demonstrated that different panels of leukemia patients and cell lines are sensitive to CK2 inhibition by TBB (tetrabromobenzotriazole) and CX-4945, a small molecule that is orally bioavailable, as shown by Song C et al. This hints that CX-4945 exerts its anti-leukemic effect via inhibition of CK2-mediated phosphorylation [21]. These data support the use of CX-4945 in a phase I clinical trial for treatment of hematologic malignancies [21,22] but since CK2 overexpression is a hallmark of ALL, the cytotoxic potential of CX-4945 in T-cell and B-cell ALL was also elucidated [13]. Overexpression of CK2 appears to impart a survival advantage in cancer cells by suppressing apoptosis through its action on a variety of cellular and nuclear substrates and favoring cell growth. Within CK2 targets, an important role is played by the transcription factor IKZF1 that acts as a tumor suppressor in T- and B-ALL; genome-wide analyses have shown that 30% of pediatric B-ALL and approximately 5% of T-ALL present a deletion or dysfunction of the IKZF1 gene [9,10]. IKZF1 haploinsufficiency is sufficient to mediate leukemogenesis. CK2 phosphorylates IKZF1 and impairs its function as a tumor suppressor in leukemia models [21]. Indeed, molecular and pharmacological inhibition of CK2 restored IKZF1 function as a tumor suppressor [13,22].

In our study, IKZF1, a prognostic factor in pediatric B-ALL [23], was also found to be mutated in children with T-ALL (3 out of 20 cases). This finding was recently confirmed by a retrospective analysis among adult and pediatric patients with T-ALL [10]. Moreover, it was recently demonstrated that CK2 and IKZF1 regulate chemoresistance to doxorubicin via repression of BCL2L1 (BCL-XL) and a combination treatment involving a CK2 inhibitor and doxorubicin has a synergistic therapeutic effect on high-risk B-ALL in vivo [24]. It is also known that MYC is an IKZF1 target gene [8]; thus, CK2 inhibitor acts as an IKZF1 activator and suppresses MYC expression in an IKZF1-dependent manner in ALL cells [8]. For this reason, we focused on MYC expression in our pediatric patients, and we demonstrate here, for the first time, a correlation between MYC overexpression, TLX3 (HOX11L2) rearrangement and CRLF2 overexpression in T-ALL patients. MYC-high expression was seen in seven TLX3 (HOX11L2) rearrangements (seven out of eight non-HR T-ALL patients); this was previously correlated with a poor prognosis [25,26]. Moreover, five out of seven non-HR T-ALL patients showed signs of CRLF2 overexpression, which has also been demonstrated as a poor prognostic marker in children with T-ALL [27]. These findings confirmed that TLX3 expression is not a prognostic indicator in pediatric T-ALL and that high levels of MYC expression are broadly present in T-ALL [28]. In patients with PTEN-Exon7 mutation, as expected, we observed a low expression of MYC. Conversely, we showed that in pediatric B-ALL patients, there is an evident correlation between a MYC-high profile and poor prognostic genetic subgroups. Interestingly, patients with 11q23-R or MLL/KMT2A-rearranged B-ALL show an extremely high level of MYC expression. Accordingly, it has recently been shown that the proliferation of MLL/KMT2A-rearranged B-ALL cells was decreased upon MYC depletion and that MYC-protein abundance in MLL/KMT2A-rearranged B-ALL cells was much higher than in non-MLL/KMT2A-rearranged B-ALL cells [29]. To date, this B-ALL subtype is correlated with a very poor prognosis, especially in infancy, and it is associated with a high rate of relapse due to a high grade of chemoresistance [30], soliciting the use of new targeted drugs. The oncogenic potential of the transcription factor ERG induced us to determine the characterization of ERG expression in pediatric ALL. Our data hint at ERG overexpression in pediatric ALL patients compared with HDs. In particular, we identified high ERG expression as an independent adverse prognostic factor in children with high-risk T-ALL and found that it was associated with an expected inferior outcome [5]. The ERG gene is involved in signal transduction pathways that regulate and promote cell differentiation, proliferation and tissue invasion [6,7]. Previous studies suggest that ERG overexpression is associated with inferior clinical outcome [15,16]. In T-ALL patients, a high level of ERG expression has been associated with poor relapse-free survival (RFS) [15]. It is also demonstrated that co-activation of the PI3K/AKT pathway and ERG overexpression collaborate with a lack of PTEN and prostate-specific androgen response in the development of prostate carcinoma [31]. Functional assays revealed that ERG may modulate kinase signaling pathways, but there are no data showing a direct correlation between CK2 and ERG expression. ERG overexpression resulted in dephosphorylation of AKT(Ser473), suggesting that ERG overexpression represses AKT activation [6,7]. So, it has been proposed that ERG-driven drug resistance overrides PI3K/AKT signaling by alternative pathways which need to be further investigated for effective drug design and adapted therapies for ERG-overexpressing high-risk leukemias [6,7]. Conversely to T-ALL, we found low ERG expression in HR patients with B-ALL, hinting that it plays the opposite role to the transcription factor within the cell. This was confirmed in a very recent article showing that ERG can act as a tumor-suppressor gene or as an oncogene with the opposite results [32].

Based on our data, we outlined a patient-specific profile based on the expression of these three biomarkers (CK2, MYC, ERG) and the mutational screening for PTEN-Exon7, IKZF1 and TLX3 genes. The CK2 overexpression associated with ERG overexpression (independently from MYC expression) related to other genetic alterations (PTEN Exon7 mutation, CDKN2A or IKZF1 deletion) could be helpful in the identification of a patient-specific profile. To be precise, we propose a “CK2-plus group” for T-ALL: CK2-high, ERG-high, PTEN Exon7 mutation, CDKN2A and IKZF1 deletion identify a subgroup of pediatric T-ALL patients with a very poor outcome, based on the occurrence of a high rate of relapse and/or death due to progressive disease. Despite the several molecular events that drive the different human leukemia subtypes, high CK2 levels appeared as a common denominator in all of them, suggesting that targeting-CK2 could represent a multi-potential therapeutic strategy [8]. Although our study presents several limitations regarding the number of children with ALL and retrospective analyses, our findings strongly suggest that CK2 and ERG represent relevant molecular markers that generate a risk-adapted treatment strategy for high-risk patients with ALL, which must be confirmed by a prospective study in a larger population. Strikingly, a recent study showed the synergistic effect of CK2 inhibitor (CX-4945) and mTOR inhibitor (rapamycin), based on the evidence that the inhibition of CK2 enhances IKAROS activity as a repressor of mTOR [33]. Thus, this dual inhibition will enhance the anti-leukemic effect in patients with HR ALL, who showed, as we demonstrated, high expression of CK2, impaired activity of IKZF1 and disruption of the PI3K/AKT/mTOR pathway [33]. Moreover, in an even more recent report, it has been confirmed that CK2 inhibitor (CX-4945) strongly suppresses the expression of WD Repeat-Containing Protein 5 (WDR5) in T-ALL by restoring Ikaros gene function. This result confirms the interaction between CK2 and Ikaros, in association with WDR5, and supports a synergistic therapeutic intervention with targeted drugs [34]. In conclusion, based on our report as a preliminary project, we propose that a prospective study should be performed (as ancillary in the upcoming international AIEOP-BFM protocol on T-ALL) with the aim of identifying the “CK2-plus group” and subsequent treatment with CK2 inhibitors in case of MRD persistence; relevant treatment should take the form of an experimental targeted therapy.

4. Materials and Methods

4.1. Patient Samples

Our study included 61 patients: 26 with T-ALL and 35 with B-ALL. All of the included patients were diagnosed in our center from September 2000 to September 2016. The clinical follow-up ranged from 72 to 216 months.

The B-ALL cohort included the following biological subgroups: 8 patients with t(9;22)/chromosome Philadelphia-positive (Ph+)/BCR::ABL1-positive; 16 with t(12;21)/ETV6::RUNX1-positive; 4 with rearranged 11q23 (-R) MLL/KMT2A-positive; and 7 without known translocations defined as “B-others”, respectively. Based on protocol response stratification criteria, patients were classified as Standard Risk (SR), Intermediate Risk (MR), or non-HR (n = 17) and High Risk (n = 18) in B-ALL, or as HR (n = 11) and non-HR (n = 15) for T-ALL patients, respectively. Patients received treatment according to the ongoing protocols [AIEOP-BFM ALL2000/R2006 and ALL2009]. The expression levels of CK2, MYC and ERG genes were retrospectively analyzed in diagnostic cDNA samples. Biological and clinical features, including adverse events and outcomes, are shown in

Table 3. Clinical and biological features of 61 children with ALL analyzed for expression of CK2-MYC-ERG.

	T-ALL n = 26 n (%)	B-ALL n = 35 n (%)	Total n = 61 n (%)
Gender
Female	6 (23)	22 (63)	28 (46)
Male	20 (77)	13 (37)	33 (54)
Age at diagnosis (years)
1–9	10 (38)	27 (77)	37 (61)
≥10	16 (62)	8 (23)	24 (39)
Presenting WBC count/μL
<10,000	4 (16)	12 (34)	16 (26)
10,000–50,000	3 (11)	11 (32)	14 (23)
50,000–100,000	8 (31)	2 (6)	10 (16)
≥100,000	11 (42)	10 (28)	21 (35)
Prednisone response ^
Good	17 (65)	27 (77)	44 (72)
Poor	9 (35)	8 (23)	17 (28)
MRD risk group *
Standard risk	3 (12)	9 (26)	12 (20)
Intermediate risk	14 (54)	19 (54)	33 (54)
High risk	4 (15)	7 (20)	11 (18)
Not performed	5 (19)	-	5 (8)
Final risk group §
SR+MR (non-HR)	15 (58)	16 (46)	30 (49)
HR	11 (42)	19 (54)	31 (51)
Event
No	18 (69)	24 (69)	42 (69)
Yes	8 (31)	11 (31)	19 (31)
Outcome
Alive	19 (73)	28 (80)	47 (77)
DOC	5 (19)	4 (11)	9 (15)
DOD	2 (8)	3 (9)	5 (8)

Table legend. ^ Good: <1000 leukemic blasts/µL after 7 days of prednisone administration; poor ≥ 1000/µL. * Minimal residual disease (MRD). Standard risk: negative both at day + 33 (end of induction or phase Ia) and day + 78 (end of consolidation therapy of phase Ib). High risk: MRD level ≥ 10⁻³ at day + 78; Intermediate risk: all others (AIEOP-BFM ALL 2000–2009). § Patients with a BCR::ABL1 or MLL/KMT2A rearrangement at diagnosis or a prednisone poor response (PPR) or ≥10% blasts (detected by cytofluorimetry) in bone marrow (BM) performed at day + 15 (for those who were enrolled at AIEOP-BFM-ALL 2009 protocol) or induction failure (≥5% blasts at day + 33 BM), were stratified into the HR final group independently from MRD results. The following events were considered: relapse or death of disease (DOD) or death of complication (DOC) during treatment.

. Institutional review board approval and an informed consent from parents were obtained, according to protocol guidelines and our hospital’s regulations.

4.2. Cell Lines and Human Normal Bone Marrows

CCRF-CEM (T-ALL bearing PTEN and CDKN2A deletions) and JURKAT (T-ALL bearing a PTEN missense mutation and CDKN2A and CREBBP deletions) cell lines were used as positive controls for T-ALL. They were cultured in RPMI-1640 medium with 10% heat-inactivated fetal bovine serum, 1% L-glutamine and 1% Pen-Strep. REH (B-ALL bearing an ETV6::RUNX1 fusion transcript) and MHH-CALL-4 (B-ALL bearing a JAK2-I682F mutation, constitutive phosphorylation of JAK2 and STAT5 and CRLF2 overexpression) were used as B-ALL positive controls (ATCC). REH was cultured as mentioned above. Instead, MHH-CALL-4 was cultured in RPMI-1640 medium with 20% heat-inactivated fetal bovine serum, 1% L-glutamine and 1% Pen-Strep. Cells were kept at 37º C in 5% CO2 and split every 3 days.

Twenty samples of bone marrow (BM) from children without a history of malignancy (healthy donors for bone marrow transplantation) were selected and served as healthy donors (HDs). We collected informed consent from their parents prior to bone marrow donation. In addition, thymocytes were also analyzed as normal internal controls for CK2, MYC and ERG expression with respect to healthy donors (Supplementary Materials).

4.3. RNA Isolation and RT-qPCR

Mononuclear cells from BM samples were isolated by Ficoll gradient centrifugation and cryopreserved. Total RNA was extracted using TRizol Reagent (Invitrogen, CA, USA) following the manufacturer’s protocols. First-strand cDNA was synthesized from total RNA with reverse transcriptase and random primers using the Superscript Reverse Transcriptase (Invitrogen, CA, USA). Samples were selected based on the quality and quantity of RNA (OD₂₆₀/OD₂₈₀ ratio: 1.8–2.0). Quantitative Real-Time RQ-PCR was performed with a 96-well optic plate using the QuantStudio 7 Flex (Applied Biosystem, Life Technologies). Each sample was tested in duplicate. Expression levels were normalized by GUS (endogenous control) and calculated using the comparative 2^−ΔΔCt method. The Comparative Cycle Threshold (CCT) method was used to determine the relative expression levels of CK2, MYC and ERG, using the median of ΔCt from HDs in two replicates and expressed as 2^−ΔΔCt (ΔCt = GUS-gene of interest). Each reaction mixture consisted of 1 μL of cDNA (from 1 mg of RNA), 7.5 μL of SYBR Green Universal Master Mix (Thermo Fisher Scientific, Waltham, MA, USA), 1 μL of each primer (250 nmol/mL) and 4.5 μL of deionized water. The PCR cycle started with an initial 95 °C, 10 min melt step, followed by 40 cycles of 15 s at 95 °C and 60 s at 60 °C. Primer sequences of CK2, MYC, ERG and GUS are shown in Supplementary Materials (Table S4). In addition, T-ALL patients were screened for the following molecular alterations: PTEN-Exon7 mutations; PICALM::AF10 fusion transcript; TP53 and pS6 mutations; TXL3 rearrangements and CRLF2 overexpression, respectively.

4.4. Data Analysis

Patients were classified as having high or low CK2 or MYC or ERG expression. We used the median value of gene expression fold changes as the cut-off. Statistical analysis was performed using GraphPad Prism 7 Software. The data are presented as the Mean ± Standard Deviations (SDs). A two-tailed p < 0.05 was indicative of a statistically significant difference between groups (Supplementary Materials).