MicroRNAs and their involvement in T-ALL: A brief overview.

T-cell acute lymphoblastic leukemia (T-ALL) is an aggressive malignancy in which the transformed clone is arrested during T-cell development. Several genetic and epigenetic events have been implicated in this transformation. MicroRNAs (miRNAs) are small, non-coding RNAs that primarily function as endogenous translational repressors of protein-coding genes. The involvement of miRNAs in the regulation of cancer progression is well-established, namely by down-regulating the expression of key oncogenes or tumor suppressors and thereby preventing or promoting tumorigenesis, respectively. Similar to other cancers, several miRNA genes have been identified and implicated in the context of T-ALL. In this review we focused on the most studied microRNAs associated with T-ALL pathogenesis.

Acute lymphoblastic leukemia (ALL) is a hematological cancer originated from the malignant transformation of a lymphoid precursor that is blocked at an immature stage of differentiation. ALL can arise in T- or B-cell precursors, hence being classified as T-ALL or B-ALL, respectively, which can be further subdivided according to specific genetic abnormalities (Pui and Evans, 2013; Pui et al., 2004).

T-ALL is characterized by the recurrent over-expression of a specific set of transcription factors, namely TLX1, TLX3, TAL1, LMO, HOXA, LYL1 and NKX family members, each defining a distinctive gene expression signature that identifies molecular subtypes in T-ALL (Belver and Ferrando, 2016; Ferrando et al., 2002; Homminga et al., 2011). Translocations leading to the activation of a small number of oncogenes occur in 25–50% of T-ALL cases. Other genetic abnormalities include frequent micro-deletions leading to the loss of tumor suppressor genes (Clappier et al., 2006; Graux et al., 2006; Hebert, 1994). The events driving a full malignant phenotype in T-ALL include also over-expression of oncogenes, defects in the cell cycle control, gain of function mutations in cytokine receptors (Oliveira et al., 2019) and tyrosine kinases (Bains et al., 2012; Oliveira et al., 2017), and activating mutations of NOTCH1(Atak et al., 2013; Weng et al., 2004).

MicroRNAs and leukemia

MicroRNAs (miRNAs) - the most studied class of non-coding RNAs – are small, 19–22 nt long, single stranded RNAs that provide post-transcriptional control of gene expression. In mammals, miRNAs function mostly as endogenous translational repressors of protein-coding genes through sequence-specific binding to the 3′-untranslated region (3′UTR) of a target messenger RNA (Bartel, 2004; Lim et al., 2005). The differential expression of microRNAs in hematopoiesis suggested early on that their deregulation could have a role in leukemogenesis (Chen et al., 2004). Initial reports of microRNA expression profiles revealed that miRNA gene expression signatures could be used to distinguish cancer types, including leukemia (Fulci et al., 2009; Lu et al., 2005; Mi et al., 2007; Schotte et al., 2009). In fact, the initial evidence for the involvement of miRNAs in cancer came from the molecular characterization of the 30 kb deletion 13q14 in human chronic lymphocytic leukemia (CLL). This genomic region harbors the miR15 and miR16 genes that were found to be deleted or down-regulated in the majority of CLL patients (Calin et al., 2002). Both microRNAs from this cluster negatively regulate BCL2, inducing apoptosis in leukemic cells (reviewed in (Pekarsky and Croce, 2019).

Importantly, leukemia signatures should be framed in light of the microRNA variations that occur during normal hematological development (Fabbri et al., 2009). Currently, the involvement of miRNAs in leukemogenesis is well established and different miRNAs have been identified as oncogenes or tumor suppressors in human leukemia. Moreover, microRNA expression signatures can be used not only to classify human acute lymphoblastic leukemia but also to predict prognosis, namely CNS involvement (Zhang et al., 2009), risk of relapse (Zhang et al., 2009) and relapse-free survival (Han et al., 2011; Schotte et al., 2011).

MicroRNAs and T-ALL

The participation of miRNA genes, individually or as part of a network, has been implicated in the pathogenesis of T-ALL.

Initial studies concluded that, contrary to B-ALL subtypes (Fulci et al., 2009), hierarchical clustering and principal component analysis of the expression levels of 430 miRNAs in 50 clinical T-ALL specimens did not distinguish between the major cytogenetic groups (HOXA, TAL or LMO and TLX1 or TLX3), which differ by few miRNAs (Mavrakis et al., 2010). Nevertheless, in the high-risk subgroup of Early T-cell precursor ALL (ETP-ALL), the microRNAs miR-221 and miR-222 were found significantly up-regulated when compared to non-ETP-ALL cases (Coskun et al., 2013). Moreover, it has been proposed that miR-222 may, to some extent, contribute to the myeloid character of ETP-ALL by down-modulating ETS1 expression. The authors hypothesized also that the fact that ETP-ALL cases failed to respond to standard intensive chemotherapy and displayed poor prognosis in initial studies, could be due to the actions of miR-222. This is because miR-222, by significantly inhibiting proliferation and causing cell cycle arrest, might not only partially explain the stem-like character of ETP-ALL but also counteract the efficacy of standard chemotherapy directed to actively cycling cells (Coskun et al., 2013). In addition, miR-221 associates with poor prognosis: increased expression correlates significantly with lower 5-year overall survival rates (Gimenes-Teixeira et al., 2013).

More recent publications involving more patients and deeper sequencing techniques, showed that molecular genetic subtypes of human T-ALL display also unique microRNA expression signatures (Wallaert et al., 2017). Moreover, distinct molecular signatures on a transcriptomic and epigenetic level (microRNAs) can differentiate infant from childhood T-ALL (Doerrenberg et al., 2017).

Many miRNAs have been found over-expressed in different tumors, functioning as oncogenes, and for that reason called oncomiRs (Calin et al., 2004; He et al., 2005; Valencia-Sanchez et al., 2006). OncomiRs generally promote tumor development by negatively inhibiting tumor suppressor genes and/or genes that control cell differentiation or apoptosis. In fact, the ablation of Dicer1 – an essential component of the MicroRNA biogenesis machinery (Cobb et al., 2006) – prevents the development and maintenance of Notch-driven T-ALL (Junker et al., 2015). Deletion of Dicer promoted apoptosis in T-ALL cells which is, in part, mediated by miR-21 and its target Pdcd4 (programmed cell death 4) (Junker et al., 2015). Notably, the most highly expressed set of miRNAs in human T-ALL was defined (miR-223, -19b, −20a, −92, -142-3p, −150, −93, −26a, −16 and miR-342) and tested in a mouse model of Notch1-induced T-ALL. The conclusion was that highly expressed miRNAs behave as oncomiRs and cooperate in regulating key tumor suppressor genes in human T-ALL, namely PTEN, BIM, NF1, FBXW7, IKZF1 and PHF6 (Mavrakis et al., 2011).

Conversely, miRNAs can act as tumor suppressors (Calin et al., 2002; Lim et al., 2005) by negatively regulating proto-oncogenes. For example, Li and colleagues (Li et al., 2011) found that down-regulation of miR-451 and miR-709 – direct repressors of Myc – is a key event during intracellular Notch1 (ICN1)-induced T-ALL in mice. ICN1 promotes the degradation of E2a, a transcriptional activator of miR-451 and miR-709, hence leading to the down-regulation of these miRNAs. In agreement, human T-ALL cells with NOTCH1 activating mutations have decreased miR-451 (miR-709 is not conserved in humans) and increased MYC levels. Later, Sanghvi et al. used differential expression analysis of miRNAs against normal thymocytes, analysis of anti-proliferative effects in vitro and tumor suppressive function in vivo, to identify a network of miRNAs (miR-29, miR-31, miR-150, miR-155, and miR-200) with interconnected tumor suppressive effects in T-ALL (Sanghvi et al., 2014). This work showed that there is a widespread mechanism of oncogene activation in T-ALL via elimination of tumor suppressor miRNA actions.

With no intension of being an exhaustive summary of all microRNAs implicated in T-ALL pathogenesis, this review will focus on some of the most studied ones (Table 1).

Table 1

Main microRNAs referred in this review, their role in the disease and their validated or predicted targets.

microRNA	Expression level	Role	Target	Reference
miR-142-3p	OE in T-ALL	OncomiR	GRa	Lv et al. (2012)
miR-146b-5p	DR in T-ALL	Tumor Supressor miR		Correia et al. (2016b)
miR-19	OE in T-ALL	OncomiR	Bim, Pkraa1, Pten, PP2A	Mavrakis et al. (2010)
miR-196b	OE in HOXA T-ALL subtypeDR in pediatric T-ALL	Dependent on the genetic context	ERG c-Myc	(Bhatia et al., 2011; Coskun et al., 2011)
miR-21		OncomiR in Notch-driven T-ALL	Pdcd4	Junker et al. (2015)
miR-222	OE in ETP-ALL	OncomiR	ETS1	Coskun et al. (2013)
miR-223	OE in T-ALL	OncomiR	FBXW7	(Correia et al., 2013; Mansour et al., 2013)
miR-26b	DR in T-ALL	Tumor Supressor miR	PIK3CD	Yuan et al. (2017)
miR-30a		Tumor Supressor miR	NOTCH1 NOTCH2	Ortega et al. (2015)
miR-451	DR in NOTCH1mut T-ALL	Tumor Supressor miR	MYC	Sanghvi et al. (2014)

OE – overexpressed.

DR – down-regulated.

The miR-17-92 microRNA cluster

The human polycistron encoding the miR-17-92 microRNA cluster is located at the chromosome 13q31, a genomic region that is recurrently amplified in lymphomas and other cancers (reviewed in (Dal Bo et al., 2015)). The human miR-17−92 transcript can be processed into seven mature miRNAs (miR-17−5p, −17−3p, −18a, −19a, −20a, −19b, and miR-92). The miRNAs encoded from this cluster are highly expressed in murine lymphocytes, embryonic stem cells and precursors. The expression of theses miRNAs decreases upon maturation during lymphocyte development (Xiao et al., 2008). The genomic region encoding the miR-17-92 cluster is often amplified in human B-cell lymphomas and cooperates with up-regulated c-Myc expression to accelerate the formation of B cell lymphomas in mice (He et al., 2005). Furthermore, increased expression of miR-17−92 in mouse lymphocytes results in the development of a lymphoproliferative disease and autoimmunity, causing the premature death of the mice (Xiao et al., 2008). In T-ALL, this cluster is involved in a genomic rearrangement, the translocation t(13; 14)(q32; q11) with the TCRA/D locus (Mavrakis et al., 2010). MiR-19, the cluster component with highest expression in T-ALL, enhances lymphocyte survival and cooperates to promote leukemogenesis in a mouse model of Notch1-induced T-ALL. In this model, miR-19 targets the pro-apoptotic protein Bim, AMP-activated kinase (Prkaa1), and the tumor suppressors Pten and PP2A, resulting in the overall activation of PI3K signaling. In this way, miR-19 directs a coordinated action to control PI3K signaling to promote lymphocyte survival and leukemogenesis (Mavrakis et al., 2010). The mechanisms of pri-miR-17-92 activation in T-ALL extend beyond the translocation mentioned above and remain to be fully understood. It has been proposed that NK-like homeodomain proteins could stimulate the expression of this polycistron in T-ALL (Nagel et al., 2009).

MiR-223

The miR-223, reported as a ‘myeloid’ gene (Fukao et al., 2007), is highly up regulated in T-ALL. It was shown to promote Notch1-driven leukemia at least in part by controlling the E3 ligase FBXW7 (Mavrakis et al., 2011). Moreover, miR-223 is a direct target of NOTCH in human T-ALL cells. The binding of NOTCH to miR-223 promoter region requires NF-kB activation (Kumar et al., 2014). We and others (Correia et al., 2013; Mansour et al., 2013) have shown that TAL1 – a major oncogenic transcription factor involved in T-ALL pathogenesis (reviewed in (Correia et al., 2016a)) - binds upstream of miR-223. This binding occurs in a previously described region containing a conserved proximal genomic element with binding sites for the transcription factor C/EBPα and PU.1 (Fukao et al., 2007). MiR-223 is activated by TAL1 and follows the same pattern of expression along thymocyte development as TAL1 (Pike-Overzet et al., 2007), with high levels in progenitor cells and sharp down-regulation in more differentiated thymocyte subsets (Correia et al., 2013). Enforced expression of miR-223 was able to partially rescue the proliferation inhibition caused by TAL1 knockdown, establishing this miRNA as a functional downstream effector of TAL1 in T-ALL cells (Mansour et al., 2013). In this study the link between TAL1 and FBXW7 down-regulation, mediated by miR-223, was firmly established as a mechanism leading to down-regulation of FBXW7 in the majority of T-ALL cases that lack gene-specific FBXW7 inactivating mutations or deletions (Mansour et al., 2013). FBXW7 targets to degradation other oncogenic proteins, such as c-MYC, MYB, cyclin E, mTOR, HIF-1 and MCL-1 (Wang et al., 2012). The oncogene MYB, known to be involved in malignant hematopoiesis (Lahortiga et al., 2007), is also a direct target of TAL1 (Sanda et al., 2012). Given that FBXW7 targets MYB to degradation and that TAL1 might down-regulate FBXW7 through miR-223, it is tempting to speculate that TAL1 may regulate MYB expression levels not only transcriptionally (Sanda et al., 2012) but also at the level of protein stability through a miR-223/FBXW7 axis. In addition, the oncogenic role of miR-223 in T-ALL may extend beyond FBXW7 and involve the potential down-regulation of targets such as E2F1, FOXO1, RHOB or EPB41L3, which have been associated with induction of apoptosis and/or have tumor suppressor roles (Correia et al., 2013).

MiR-142-3p

miR-142-3p was initially identified as a specific microRNA for hematopoietic cells (Chen et al., 2004; Landgraf et al., 2007). It is highly expressed in pediatric ALL samples, particularly in T-ALL cells as compared with healthy donor T-cells (Schotte et al., 2009) and especially in T-ALL cells from patients with poor prognosis ((Schotte et al., 2011) Importantly, miR-142-3p stands out as an example of the essential role that a single miRNA can play in leukemia progression, chemotherapeutic resistance and prognosis. In T-ALL, miR-142-3p was shown to promote leukemic cell growth and to induce resistance to glucocorticoid (GC) treatment (Lv et al., 2012). In addition, ectopic expression of miR-142-3p results in the increased proliferation of T-ALL cells without affecting apoptosis. This miRNA specifically targets the cAMP/PKA pathway and glucocorticoid receptor alpha (GRa) and, importantly, T-ALL cells from patients with poor response to prednisolone are sensitized to cell death induced by dexamethasone after down-modulation of miR-142-3p (Lv et al., 2012).

MiR-196b

The role of some miRNAs in leukemia is still controversial since it may depend on the cellular context. For instance, miR-196b is highly expressed in T-ALL patient samples as compared to B-ALL (Schotte et al., 2009). In pediatric patients, miR-196b is highly co-expressed with genes from the HOXA cluster, namely in T-ALL cases characterized by the activation of HOXA genes, suggesting co-transcriptional activation (Schotte et al., 2010). The over-expression of this miRNA in mouse bone marrow cells leads to increased proliferative capacity and survival, pointing to a role in leukemogenesis (Coskun et al., 2011). In contrast, it was reported that miR-196b can down-regulate the oncogenic transcription factor ERG in adult AML and T-ALL patients. This hints on the possibility that, instead of promoting leukemogenesis, miR-196b may inhibit this process in some circumstances. In line with this, miR-196b was found down-modulated in pediatric T-ALL patients with respect to normal cells, which suggests a possible tumor suppressor function for this miRNA also in childhood T-ALL (Bhatia et al., 2011). In contrast to what was previously found in B-ALL, c-Myc expression is not down-regulated by miR-196b in T-ALL. Interestingly, miR-196b loses its ability to down-regulate c-Myc expression in T-ALL as a result of mutations in the target 3′UTR of c-Myc, pointing again to a possible tumor suppressive role in this disease. This microRNA may therefore have a dual role in leukemia, depending on the genetic context (Bhatia et al., 2011).

Other miRs with impact in T-ALL

Other microRNAs implicated in T-ALL include for instance miR-146b-5p, a tumor suppressor miRNA downregulated by TAL1 (Correia et al., 2016c) and involved in T-cell migration (Correia et al., 2016b; Tu et al., 2019); the miR-30a - transcriptionally suppressed by MYC and a direct inhibitor of NOTCH1 and NOTCH2 expression – mediates a regulatory circuitry that modulate the oncogenic signals of these T-ALL drivers (Ortega et al., 2015); miR-125b causes the malignant transformation of different hematopoietic lineages, leading to B-ALL, T-ALL, or myeloproliferative neoplasms (Bousquet et al., 2010). Recently this microRNA was shown to be activated by TLX3 in the TLX3-subtype of T-ALL (Renou et al., 2017); miR-26b, another tumor suppressor, is an inhibitor of the PI3K/AKT pathway via the targeting of PIK3CD. MiR-26 expression is promoted by PTEN – another inhibitor of the pathway - by an unknown mechanism that involves the differential regulation of isoforms of the transcription factor Ikaros (Yuan et al., 2017); miR-181a – encoded by the miR-181ab1 gene – contributes to the maintenance of Notch oncogenic signals in T-ALL by diminishing Notch negative feedbacks (mediated for instance by Nrarp) and potentiating pre-TCR signals (Fragoso et al., 2012). The deletion of miR-181ab1 gene has no significant impact on normal development, making miR-181a an interesting therapeutic target to inhibit oncogenic signals as alternative to the targeting of oncogenes themselves (Fragoso et al., 2012).

Conclusion

Since microRNAs were discovered, around 25 years ago, we have understood their biogenesis, the mechanisms how they regulate mRNA stability and protein translation and their importance in normal development and physiology. We have also started to appreciate the importance of their deregulation in the genesis and development of malignancy. Many, if not most, protein-coding transcripts are potential targets for miRNA regulation (Lewis et al., 2005). An additional layer of complexity is added by the fact that one miRNA is able to target many transcripts of genes often with related functions. Furthermore, several individual transcripts can be targeted by numerous miRNAs that are often regulated together. Integration of data from the miRnome with the transcriptome and proteome will be fundamental for a more comprehensive analysis of the consequences of disrupting the cellular architecture sustained by microRNAs. This obviously requires broad and large cohorts of patients and appropriate healthy tissue, which is a challenge for a rare disease like T-ALL. Nevertheless, miRNAs are not only very interesting biomarkers for various diseases, including T-ALL, but also their tissue specificity makes them attractive candidates for targeted therapy.

Conflicts of interest

The authors have not conflict of interest to disclose.