Role of microRNA in the pathogenesis of malignant lymphoma

Abstract

MicroRNA (miRNA) are non‐coding regulatory RNA usually consisting of 20−24 nucleotides. Over the past decade, increases and decreases in miRNA expression have been shown to associate with various types of disease, including cancer. The first two known miRNA aberrations resulted from altered expression of DLEU2 and C13orf25 in hematological malignancies. DLEU2, which encodes miR‐15a and miR‐16‐1, was discovered from 13q14 deletion in chronic lymphocytic leukemia, while C13orf25, which encodes six mature miRNA (miR‐17, miR‐18, miR‐19a, miR‐19b, miR‐20a and miR‐92a), was identified from 13q31 amplification in aggressive B‐cell lymphomas. These miRNA were downregulated or upregulated in accordance with genomic deletion or amplification, which suggests that they contribute to tumorigenesis through altered regulation of target oncogenes or tumor suppressors. Consistent with that idea, miR‐15a/16‐1 is known to regulate Bcl2 in chronic lymphocytic leukemia, and miR‐17‐92 regulates the tumor suppressors p21, Pten and Bim in aggressive B‐cell lymphomas. Dysregulation of other miRNA, including miR‐21, miR‐29, miR‐150 and miR‐155, have also been shown to play crucial roles in the pathogenesis of aggressive transformed, high‐grade and refractory lymphomas. Addition of miRNA dysregulation to the original genetic events likely enhances tumorigenicity of malignant lymphoma through activation of one or more signaling pathways.

MicroRNA (miRNA) are non‐coding regulatory RNA consisting of 20−24 nucleotides.1, 2, 3 Regulatory RNA act by controlling the translation of proteins from mRNA, and by doing so play a crucial role in normal cell differentiation and proliferation.1, 2, 3 Approximately 2500 miRNA have been identified in humans, and it is known that nearly all human protein‐encoding genes can be controlled by miRNA in both healthy and malignant cells. Many miRNA alterations have also been reported in both hematological malignancies and solid tumors.2, 3 The role played by miRNA dysregulation in malignant lymphomas and other cancers is being actively studied worldwide. However, the available information remains incomplete and complex, so that the precise function of miRNA in carcinogenesis is unknown. In this review, we have summarized the actions of miRNA in distinct subtypes of malignant lymphomas, focusing in particular on their involvement in disease progression and transformation.

Discovery of MicroRNA Dysregulations in Hematological Malignancies

Abnormal expression of miRNA is now known to occur in many cancers, but it was first reported in chronic lymphocytic leukemia (CLL). In 2001, Dalla‐Favera's group found reduced expression of the gene DLEU2 through a detailed search of the minimal deleted region 13q14.4 In 2002, Croce's group reported reduced expression of miR‐15a and miR‐16‐1 from the common loss region of 13q14 in CLL.5 They also showed that miR‐15a and 16‐1 were present in the non‐coding region of DLEU2. Then, in 2003, overexpression of BIC, which encodes miR‐155, was detected in children with Burkitt lymphoma (BL),6 although miR‐155 expression is usually absent in sporadic primary BL cases.7 In 2004, Seto's group discovered the overexpression of a non‐coding gene, C13orf25, in aggressive B‐cell lymphomas.8 C13orf25 contained six mature miRNA sequences, suggesting that dysregulation mRNA plays a key role in lymphomagenesis.

On the basis of these reports, Croce's group hypothesized that miRNA instability plays a crucial role in carcinogenesis, and that aberrantly upregulated miRNA might suppress expression of tumor suppressors, while aberrantly downregulated miRNA might promote expression of oncoproteins.9 In fact, it has now been demonstrated in various cancers that miRNA alter expression of tumor suppressors and oncoproteins by inhibiting translation of their respective mRNA (canonical function). Figure 1a provides a schematic illustration of the mechanisms governing the canonical miRNA functions in cancer. Eiring et al. (10) demonstrate that a tumor‐suppressive miRNA also exhibits a “decoy” function in chronic myelogenous leukemia (CML).10 MiR‐328 negatively regulate PIM1 oncoprotein by canonical function. In CML chronic phase, miR‐328 combines to a transcription factor, hnRNP E2 (negative regulator of CEBPA/c‐EBPα), as a “decoy,” leading to differentiation from CML blast cells to granulocytes. In contrast, in CML blast crisis (BC) phase, downregulation of miR‐328 leads to differentiation block of CML blast cells via activation of hnRNP E2. In CML BC, together with the enhanced expression of PIM1, the reduced expression of miR‐328 leads to an increase in the production of undifferentiated blast cells. This finding is attracting attention as a new mechanism by which tumor‐forming miRNA act; that is, regulation of target protein expression through a “decoy” (Fig. 1b). So far, the decoy function has only been demonstrated in CML, but the function may be associated with other cancers.

Figure 1.

Schematic diagram of cancer‐related microRNA (miRNA) function. (a) Canonical function of oncogenic or tumor‐suppressive miRNA are shown. Aberrantly upregulated miRNA (e.g. miR‐17‐92) might suppress expression of tumor suppressors (e.g. Pten), while aberrantly downregulated miRNA (e.g. miR‐16‐1) might promote expression of oncoproteins (e.g. Bcl2). (b) MiRNA (e.g. miR‐328) as a decoy is shown.

MiR‐17‐92 is the First MicroRNA Known to be Dysregulated in Malignant Lymphomas

13q31‐32 amplification is a well‐known genomic alteration in diffuse large B‐cell lymphoma (DLBCL) that was recognized well before its response gene was identified. This is because there was no protein‐coding gene in that region. In 2004, Seto's group carefully examined the expression of approximately 60 expressed sequence tags (EST) and found that 1 EST, BCO40320, was strongly expressed in accordance with 13q31 amplification in DLBCL. They then identified a novel gene, C13orf25, from the EST using the 5′ or 3′ rapid amplification of cDNA end (RACE) method.8 C13orf25 contains the miR‐17‐92 polycistron, encoding six miRNA (miR‐17‐5p, miR‐18, miR‐19a, miR‐19b, miR‐20a and miR‐92‐1), which can be divided into the miR‐19 and miR‐17 families. Subsequent investigation revealed that the miR‐17‐92 is overexpressed in aggressive B‐cell lymphomas (DLBCL, mantle cell lymphoma [MCL] and BL) with genomic amplification of 13q31 (Fig. 2a).8, 11, 12 Based on those reports8, Hannon & Hammond's group investigated the polycistron's potential to act as an oncogene. They first transduced miR‐17‐19b into stem cells from the fetal liver of Eu‐Myc mice, and then inoculated the transduced cells into irradiated Eu‐Myc mice. They found that the death rate among mice inoculated with miR‐17‐19b was higher than among control mice due to the induction of B‐cell leukemia. They also showed that miR‐17‐92 could induce tumor formation by acting in concert with Myc.14 At the same time, Mendell's group demonstrated that Myc could upregulate miR‐17‐92 and that E2F is a direct target of the miR‐17 family,15 although it remains unknown whether reduction of E2F expression is associated with lymphomagenesis. Dews et al. (16) reveal that miR‐18, another member of the miR‐17‐92 polycistron, regulates thrombospondin1 and connective tissue growth factor, whose downregulation activates angiogenesis in colon cancer.16 In 2008, two groups, respectively, demonstrated miR‐17‐92 target gene(s) involved in normal B‐cell development. Using miR‐17‐92 knockout mice, Jacks' lab showed that deletion of miR‐17‐92 could induce upregulation of the pro‐apoptotic protein Bim with inhibition of differentiation from pro‐B cell to pre‐B cell transition.17 In addition, Rajewsky's lab established miR‐17‐92 transgenic mice,18 which developed lymph‐proliferative disease because the miR‐17‐92 reduced expression of Bim protein; however, they did not develop lymphomas. Nonetheless, these reports suggest the pro‐apoptotic protein Bim is a likely target of miR‐17‐92 during B‐cell lymphomagenesis.

Figure 2.

C13orf25/miR‐17‐92 is the first oncogenic microRNA (miRNA). (a) Array based CGH for B‐cell lymphoma cell lines (Karpass1718, REC1) and a primary diffuse large B‐cell lymphoma (DLBCL) sample. Green spots show genomic amplification and red spots show genomic deletion. C13orf25 is located in a common amplification region at 13q31‐q32. 13q32 (C13orf25 region) amplification commonly occurred among examined cell lines and a primary case. (b) Schematic illustration of miR‐17‐92 function in B‐cell lymphomas. Genomic amplification of 13q32 and/or Myc overexpression can induce upregulation of miR‐17‐92, which include six mature miRNA. Each miRNA family (miR‐17 family and miR‐19 family) has distinct target(s) in lymphoma genesis. (c) Genomic amplification and deletion of DLBCL subtypes. Activated B‐cell (ABC) DLBCL is showing 9p21 (CDKN2A).23 In contrast, germinal center B‐cell (GCB) DLBCL is showing either 13q31 amplification or 10q23 loss, yielding Pten loss.23

The findings summarized above strongly suggest that by enhancing anti‐apoptotic capability in B‐cell lymphomas, downregulation of Bim by miR‐17‐92 contributes to lymphomagenesis. However, in MCL, there have been several cases in which both miR‐17‐92 overexpression via 13q31 amplification and loss of BCL2L11/Bim via 2q13 homozygous deletion were observed.11, 12 This fact led us to the idea that miR‐17‐92 has an additional target, at least in some cases. Using the Jeko‐1 MCL cell line, which shows both 13q31 amplification and homozygous deletion of Bim, our group discovered that miR‐17‐92 can also regulate CDKN1A/p21. When we knocked down miR‐17 and miR‐20, cell cycling was arrested at G1/S via upregulation of CDKN1A/p21. Conversely, when we transduced miR‐17‐19b into the SUDHL4 B‐cell lymphoma cell line, which otherwise does not show upregulation of miR‐17‐92, CDKN1A/p21 was downregulated and G1/S progression was enhanced. These results demonstrate that in addition to Bim, the miR‐17‐92 also regulates p21.19

In 2009, two groups (He et al. and Ventura et al.) demonstrated that miR‐19 could regulate the tumor suppressor Pten, thereby enhancing anti‐apoptotic potential via upregulation of the AKT/mTOR pathway (Fig. 2b).20, 21 Collectively, these reports suggest that miR‐17‐92 regulates several targets in different B‐cell lymphoma subtypes (miR‐17 family: Bim and p21; miR‐19 family: Pten), and that the upregulation of miR‐17‐92 is an additional genetic event that enhances the tumorigenicity of the original cancer.

Differences in MicroRNA Expression Can Define The Lymphoma Subtype

Since the discovery of the upregulated expression of miR‐17‐92 in malignant lymphomas, the genomes of various subtypes of malignant lymphomas have been screened for miRNA expression. Details of the dysregulation of miRNA in malignant lymphoma are summarized in Table 1.

Table 1.

MicroRNA dysregulation in malignant lymphoma

MicroRNA	Subtype	Expression	Target proteins	References	Notes
let‐7f	NMZL	Up		Craig et al.60	Versus FL
let‐7a	cHL	Up	BLIMP1	Nie et al.67	In Hodgin's cells
miR‐9	BL	Down		Lenze et al.34	Versus DLBCL
	cHL	Up	BLIMP1	Nie et al.67	In Hodgkin's cells
miR‐15a	MCL	Down		Beà et al.37, Zhang et al.68	Downregulated by MYC‐HDAC3
	CTCL (SzS)	Down		Ballabio et al.69	Versus CD4+ cells
miR‐16‐1	FL	Down		Karube et al.57	t(14;18)–negative specific
	cHL	Up		Gibcus et al.70	Versus NHL
	CTCL(SzS)	Down		Ballabio et al.69	Versus CD4+ cells
	MCL	Down	BCL2, BMI1	Beà et al.37, Zhang et al.68, DiLisio et al.41	Down‐regulated by MYC, downregulated in MCL SP DiLisio et al.41
miR‐18a	DLBCL	Up		Alencar et al.28	Versus B‐cells, poor OS (R‐CHOP)
miR‐20a/b	FL	Up	p21	Oshiro et al.55	Versus CD19+ B‐cells
miR‐17‐92	MCL	Up	p21,PTEN	Zhang et al.39, Rao et al.71	Versus CD19+ B or IgD B‐cells, poor OS
	cHL	Up		Gibcus et al.70	Versus NHL
	BL	Up		Tagawa et al.13	Versus 13q31 amp(−) BL
	DLBCL(GCB)	Up	PTEN	Lenz et al.23, Fassina et al.72	Versus ABC type (Lenz et al.23) GC‐DLBCL from high‐grade FL
	ALCL‐ALK (+)	Up		Merkel et al.30	Versus ALK (−) ALCL
miR‐21	SMZL	Up		Ruiz‐Ballesteros62, Bouteloup et al.63	Versus normal spleen (aggressive SMZL) (Ruiz‐Ballesteros62), poor OS (Bouteloup et al.63)
	NK/T	Up	PDCD4, PTEN	Karube et al.46, Yamanaka et al.47	Versus CD56 cells
	cHL	Up		Navarro et al.73, Gibcus et al.70	Versus reactive lymph nodes, versus NHL
	DLBCL	Up		Malumbres et al.26, Lawrie et al.74, 75	Versus GCB DLBCL (cell line) (Malumbres et al.26 and Montes‐Moreno et al.27
miR‐23a	BL	Down		Lenze et al.34	Versus DLBCL
	cHL	Up		Navarro et al.73	Versus reactive lymph nodes
miR‐23b	BL	Down		Lenze et al.34	Versus DLBCL
miR‐26a	BL	Down		Lenze et al.34	Versus DLBCL
	MCL	Down		Zhao et al.38, Navarro et al.40	Versus CD5+ B‐cells
	FL	Down		Karube et al.57	t(14;18)–negative cases
miR‐26b	BL	Down		Lenze et al.34	Versus DLBCL
	cHL	Down		Navarro et al.73	Versus reactive lymph nodes
	SMZL	Down		Bouteloup et al.63	In HCV‐positive patient
miR‐29a	SMZL	Down		Arribas et al.61, Ruiz‐Ballesteros62	Versus FL, MCL, and CLL with splenic involvement (Arribas et al.61), Versus normal spleen (Ruiz‐Ballesteros62)
	ALCL‐ALK(+)	Down	MCL‐6	Desjobert et al.31	Versus ALK (−) ALCL
	MCL	Down	CDK6, IGF‐1R	Beà et al.37, Zhao et al.38	Epigenetically down‐regulated by MYC‐HDAC‐EZH2 (Zhao et al.38)
miR‐29b	BL	Down		Lenze et al.34	Versus DLBCL
	SMZL	Down		Arribas et al.61	Versus FL, MCL, and CLL with splenic involvement
miR‐29c	NMZL	Up		Craig et al.60	Versus lymph node with reactive lymphoid hyperplasia
	FL	Down		Karube et al.57	t(14;18)–negative cases
miR‐30a	BL	Down		Lenze et al.34	Versus DLBCL
miR‐30b	cHL	Down		Navarro et al.73	Versus reactive lymph nodes
miR‐30d	BL	Down		Lenze et al.34	Versus DLBCL
miR‐31	ATL	Down	NIK/NF‐kB	Bellon et al.53	Versus CD4+ cells
	cHL	Down		Navarro et al.73	Versus reactive lymph nodes
miR‐34a	DLBCL	Up	FOXP1	Liu59	Associated with transformation from gastric MALT to DLBCL
miR‐96	cHL EBV+	Down		Navarro et al.73	Versus EBV‐negative cHL
miR‐101	FL	Down		Karube et al.57	t(14;18)–negative cases
	ALCL	Down		Merkel et al.30	Inhibition of prolifelation in ALK(+) type (cell lines)
miR‐125b	DLBCL	Up	IRF4, BLIMP1	Malumbres et al.26	in GC‐enriched cells
miR‐128a/b	cHL EBV+	Down		Navarro et al.73	Versus EBV‐negative cHL
miR‐135a	cHL	Down	JAK2	Navarro et al.73, 76	Versus reactive pymph nodes, poor outcome
miR‐142‐3p	NK/T EBV+	Down	IL1A	Motsch et al.77	Versus NK/T EBV(−) cells
	ATL	Up		Yin et al.52	Versus HTLV1 infected cells, versus CD4+ cells
miR‐142‐5p	BL	Down		Lenze et al.34	Versus DLBCL
	MALT(gastric)	Up	TP53INP1	Saito et al.78	Poor reactivity to H. pylori eradication therapy
miR‐146a	BL	Down		Lenze et al.34	Versus DLBCL
	NK/T	—	TRAF6	Paik et al.79	Low miR‐146a patients have poor prognosis
	DLBCL(ABC)	Up		Malumbres et al.26	Versus GCB DLBCL (cell line)
miR‐146b	BL	Down		Lenze et al.34	Versus DLBCL
	DLBCL(ABC)	Up		Malumbres et al.26	Versus GCB DLBCL (cell line)
miR‐150	MCL	Down		Beà et al.37	Versus CD19+ or IgD B‐cells
	NK/T	Down	DKC1, AKT2	Yamanaka et al.47	Versus CD56 cells
	cHL	Down		Gibcus et al.70	Versus NHL
	PCMZL	Down		Monsálvez et al.80	Inferior PFS
miR‐155	MCL	Up		Beà et al.37	Versus CD19+ or IgD B‐cells
	BL	Down	AID	Kluiver et al.81, Kluiver et al.7, Dorsett et al.33, Lenze et al.34	Versus DLBCL (Lenze et al.34)
	SMZL	Up		Ruiz‐Ballesteros62	Versus normal spleen
	NK/T	Up	SHIP1	Karube et al.46, Yamanaka et al.47	Versus CD56 cells
	DLBCL(ABC)	Up		Eis et al.24, Malumbres et al.26, Lawrie et al.75	Versus GCB DLBCL
	PMBCL	Up		Kluiver et al.29	Versus GCB DLBCL
	cHL	Up		Gibcus et al.70	Versus NHL
	ALCL‐ALK(‐)	Up		Merkel et al.30	Versus ALK(+) ALCL
	MALT(gastric)	Up	TP53INP1	Saito et al.78	In cases with poor reactivity to H. pylori eradication therapy
	ATL	Up		Yin et al.52	Versus HTLV1 infected cells, versus CD4+ cells
	PCMZL	—		Monsálvez et al.80	Poor PFS
miR‐181a	DLBCL	Up		Alencar et al.28	R‐CHOP‐treated DLBCL patients have better PFS
	DLBCL(GCB)	Up		Lawrie et al.75	GCB specific (cell lines)
	ATL	Down		Yin et al.52	Versus HTLV1 infected cells, versus CD4+ cells
miR‐194	FL	Up	SOCS2	Oshiro et al.55	Versus CD10+ B‐cells
miR‐205	NK/T EBV(+)	Up	BCL6	Motsch et al.77	Versus EBV‐negative NK/T
	CTCL	Down		Ralfkiaer et al.82	Versus benign skin disease or normal skin
miR‐221	BL	Down		Lenze et al.34	Versus DLBCL
	NMZL	Up	LMO2	Craig et al.60	Versus FL and lymph node with reactive lymphoid hyperplasia
	DLBCL(ABC)	Up		Lawrie et al.74	Versus GCB DLBCL (cell line)
miR‐222	BL	Down		Lenze et al.34	Versus DLBCL
	DLBCL	Up		Malumbres et al.26, Montes‐Moreno et al.27, Alencar et al.28	R‐CHOP‐treated DLBCL patients have inferior PFS and/or OS
	DLBCL(ABC)	Up		Malumbres et al.26, Lawire (2008)	Versus GCB DLBCL (cell line)
miR‐223	NMZL	Up	LMO2	Craig et al.60	Versus FL
		Up	LMO2	Malumbres et al.26	Memory B‐cell‐enriched, but not in centroblasts
	CTCL(SzS)	Down		Ballabio et al.69	Versus MF, Versus CD4+
	MALT(gastric)	Up		Leich et al.58	In high stage and cases with poor reactivity to H. pylori eradication therapy
miR‐574‐3p	DLBCL(ABC)	Up		Malumbres et al.26	Versus GCB DLBCL (cell line)
miR‐574‐5p	DLBCL(ABC)	Up		Malumbres et al.26	Versus GCB DLBCL (cell line)
	CTCL(SzS)	Up		Ballabio et al.69	Versus CD4+ cells
miR‐768‐3p	AITL/PTCLnos	—		Valleron et al.83	pre‐miR‐768 overlaps snoRNA HBII‐239 (favourable outcome)

ABC, activated B‐cell‐like; AITL, angioImmunoblastic T‐cell lymphoma; ALCL, anaplastic large cell lymphoma; ALK, anaplastic lymphoma kinase; ATL, adult T cell leukemia; BL, Burkitt lymphoma; cHL, classical hodgkin lymphoma; CTCL, cutaneous T cell lymphoma; DLBCL, diffuse large B‐cell lymphoma; EBV, Epstein–Barr virus; FL, follicular lymphoma; GCB, germinal center B‐cell‐like; HCV, hepatitis C virus; MALT, mucosa‐associated lymphoid tissue lymphoma; MCL, mantle cell lymphoma; MF, mycosis fungoides; NHL, non‐Hodgkin lymphoma; NK/T, natural killer/T‐cell lymphoma; NMZL, nodal marginal zone lymphoma; OS, overall survival; PCMZL, primary cutaneous marginal zone B‐cell lymphoma; PFS, progression free survival; PMBCL, Primary mediastinal large B‐cell lymphoma; PTCLnos, peripheral T‐cell lymphomas not otherwise specified; R‐CHOP, Rituximab plus cyclophosphamide, doxorubicin, vincristine and prednisone chemotherapy; SMZL, splenic marginal zone lymphoma; SP, side population; SzS, Sézary syndrome.

MicroRNA may even be differentially expressed within a single tumor entity, such as DLBCL (activated B‐cell [ABC] vs germinal center B‐cell [GCB]) or anaplastic large cell lymphoma (ALCL) (anaplastic lymphoma receptor tyrosine kinase [ALK] positive [+] vs ALK negative [−]), and the difference may be crucial to the pathogenesis of the lymphoma subtype.

Diffuse large B‐cell lymphoma (DLBCL) can be divided into two distinct subtypes: ABC and GCB types.22 Lenz et al. (23) report that 13q31 amplification (C13orf25) frequently occurs in GCB but not in ABC‐type DLBCL.23 They further show that DLBCL overexpressing miR‐17‐92 also express MYC and their target genes at significantly higher levels than those without this abnormality. Interestingly, 10q23 (PTEN) deletion is frequently seen in GCB DLBCL without 13q amplification. Both 13q31 amplification and 10q23 deletion could downregulate Pten, suggesting that altered AKT‐mTOR signaling may be important in GCB DLBCL (Fig. 2c). Significantly higher levels of miR‐155 are present in the ABC type than the GCB type.24 Because AID, which is an essential regulator of class switch recombination and somatic hypermutation in germinal B‐cells,25 is expressed in the germinal center and miR‐155 can regulate this expression, it seems likely that miR‐155 is downregulated in GCB DLBCL. In ABC DLBCL, miR‐155 may act to regulate tumor suppressors, but the precise target(s) has not yet been identified. In addition, Malumbres et al. (26) provide evidence that germinal center‐enriched miR‐125b downregulates expression of IRF4 and PRDM1, and memory B‐cell‐enriched miR‐223 downregulates expression of LMO2.26 These reports also suggest that miRNA play crucial roles in different DLBCL subtypes. Furthermore, miR‐181a and miR‐222 have been shown to predict overall survival and progression‐free survival in rituximab cyclophosphamide, doxorubicin, vincristine and prednisone (R‐CHOP)‐treated DLBCL patients.26, 27, 28 In addition, miR‐155 is shown to be upregulated in primary mediastinal large B‐cell lymphoma.29

It was also recently reported that miRNA dysregulation in ALCL differs between the ALK (+) and ALK (−) subtypes. Merkel et al. (30) show that ALK (+) and ALK (−) could be distinguished based on their distinct miR‐17‐92 profiles: miR‐17‐92 was more strongly expressed in the ALK (+) subtype. Moreover, miR‐29a and miR‐101 were downregulated in ALCL and its forced expression reduced proliferation of only the ALK (+) subtype in vitro.30, 31 These miRNA can be hallmarks for distinguishing ALK (+) or ALK (−) subtypes.

Altogether, differences in miRNA expression likely contribute to the distinct behaviors of different lymphoma subtypes through regulation of specific genes and their transcripts.

MicroRNA Dysregulation Strongly Contributes the Pathogenesis of Aggressive Lymphomas

Functional analyses of miRNA have been conducted with aggressive lymphomas, including BL, MCL and NK/T‐cell lymphoma. BL is characterized by the dysregulated expression of MYC as a consequence of translocations involving the MYC (8q24) and immunoglobulin genes. It has also been shown that AID is required for the MYC translocation and development of BL.32 MiR‐155 expression is reduced in BL,7 and recent work by Dorsett et al. demonstrates that miR‐155 suppresses AID‐mediated MYC‐IGH translocation.33 This suggests that downregulation of 25miR‐155 in germinal center lymphoid tissue is a deeply associated first hit event in BL. Furthermore, several epidemiologic subtypes of BL (endemic, sporadic and HIV‐associated) share a homogeneous microRNA profile, distinct from that of DLBCL,34 which confirms the potential relevance of this signature in the diagnosis of BL.

Mantle cell lymphoma is characterized by t(11,14)(q13;q32), which results in overexpression of CCND1/CyclinD1, and is presumed to derive from naïve pre‐germinal center CD5⁺ B‐cells (subset [30%] of MCL derive from antigen experienced cells with identity of immunoglobulin heavy chain variable region gene of 92–98% with the germline).35, 36 Underlying MCL is a larger number of genetic alterations than is seen in other lymphoma subtypes.12, 37 Many miRNA aberrations and miRNA dysregulation have also been identified in MCL.38, 39, 40, 41 Essential is dysregulation of miR‐29, miR‐15a/16‐1, miR‐26 and miR‐17‐92. miR‐17‐92 is frequently upregulated in MCL, and because miR‐17‐92 appears to negatively regulate CDKN1A/p21,19 increases in its expression could enhance cell cycle progression. miR‐16‐1 is expressed normally in MCL, as compared to its expression in normal CD5⁺ B‐cells,42, 43 but Chen et al. demonstrate that the 3′UTR of CCND1 is frequently truncated or mutated in MCL, which inhibits the interaction of miR‐16‐1 with the “seed” sequence of the 3′UTR of CCND1, thereby contributing to continuous upregulation of CyclinD1.43 In addition, some miRNA have been shown to act as regulators of polycomb‐group repressive complex proteins. MiR‐16‐1 can regulate BMI1 translation, which would enhance anti‐apoptotic potential by negatively regulating pro‐apoptotic genes, such as PMAIP1/Noxa and BCL2L11/Bim.42

In aggressive B‐cell lymphomas, such as BL, DLBCL and MCL, overexpression of MYC is strongly associated with their aggressiveness. Zhang et al. (39) show that MYC, HDAC3 and EZH2 form a repressive complex tethered to miR‐29 promoter elements to epigenetically repress miR‐29 transcription in MYC‐expressing lymphoma cells. Downregulation of miR‐29 induces upregulation of CDK6 (cell cycle progression) and IGF‐1R (anti‐apoptosis). Furthermore, MYC can regulate transcription of miR‐26a, whose downregulation leads to upregulation of EZH2. This, in turn, reduces miR‐494 expression, leading to upregulation of MYC. This MYC‐miR‐26a‐EZH2‐miR‐494 positive feedback loop is observed in aggressive MCL, especially in the cases with MYC upregulation.39 A schematic illustration of these complicated signals is shown in Figure 3.

Figure 3.

Schematic illustration of microRNA (miRNA) dysregulation in mantle cell lymphoma (MCL) or aggressive B‐cell lymphoma with MYC overexpression.

NK‐cell leukemia and NK/T‐cell lymphoma are tumors derived from natural killer cells (sCD3⁻CD56⁺TCR⁻) whose onset and development are governed to a great extent by Epstein–Barr virus (EBV). Because the pathogenesis of NK/T‐cell lymphoma remained largely unknown, comparative genomic hybridization and/or gene expression profiling were conducted to detect the genes responsible.44, 45 Approximately 10 to 20% of NK/T‐cell lymphoma cases show a 6q deletion, which affects expression of the AIM1, PRDM1 and FOXO3 transcription factors.45, 46 We hypothesized that the remaining 80% of cases might show miRNA aberrations. By screening for miRNA expression, our group found that miR‐21 and miR‐155 were upregulated and miR‐150 was downregulated in primary NK‐cell tumors and cell lines.47, 48 We also found that expression of miR‐21 and miR‐155 was mutually exclusive, suggesting that these two miRNA target different downstream genes in the same signal cascade. Indeed, miR‐21 negatively regulates the tumor suppressors Pten and PDCD4 in NK‐cell leukemia, while miR‐155 regulates an inositol phospholipid phosphatase, Ship1, in NK/T‐cell lymphoma.47, 48, 49, 50, 51 Pten and Ship1, respectively, regulate dephosphorylation of phosphatidylinositol (3–5)‐trisphosphate (PIP3) to phosphatidylinositol 4,5‐bisphosphate (PIP2) and PIP5 to PIP3, and their downregulation likely leads to activation of AKT signaling.50 We also found that miR‐150 is downregulated in NK and T‐cell lymphomas, which could directly affect AKT2 and dyekerin, which induce cellular senescence. Consequently, downregulation of miR‐150 has an anti‐aging effect, leading to immortalization within NK‐cell tumors (Fig. 4).48 These findings suggest that EBV infection may cause the upregulation of several miRNA, including miR‐21 and miR‐155, as infection with EBV is associated with immortalization of lymphoid cells.52

Figure 4.

microRNA (miRNA) dysregulation in NK/T‐cell lymphoma. (a) Schematic illustration of the role of miR‐21, miR‐155 and miR‐150 in NK‐cell tumor. (b). Expressions of miR‐21, miR‐155 and miR‐150 in normal NK‐cell, NK‐cell lymphoma/leukemia cell lines and primary samples. (c). Immunostaining of CD56, pAKT ^ser473/4 and EBER against an example of NK/T‐cell lymphoma case (extra nodal type). EBV, Epstein–Barr virus.

Bellon et al. (53) found that several miRNA were dysregulated in adult T‐cell leukemia, which is an aggressive tumor entity.53 Furthermore, Yamagishi et al. (54) discovered that miR‐31 regulates an NF‐κB‐inducing kinase that plays a central role in non‐canonical signaling and constitutive activation of the NF‐κB pathway.54 However, there is also evidence that the gene expression and genomic profiles distinctly differ between leukemia and lymphoma,55 suggesting that miR‐31 contributes to tumorigenesis in the former but not the latter. Further study will be required to determine whether miR‐31 plays an important role as a tumor‐suppressive miRNA in lymphomas.

MiRNA May Contribute to Disease Progression and Transformation of Low Grade Lymphomas

MicroRNA do not appear to be as important for the pathogenesis of low grade B‐cell lymphomas (e.g. follicular lymphoma [FL] and marginal zone lymphoma [MZL]) as for high‐grade and transformed lymphomas (e.g. FL/MZL to DLBCL).

Follicular lymphoma is characterized as a indolent B‐cell lymphoma, with approximately 80% of cases possessing t(14;18)(q32;q21). In FL with the translocation, the miRNA profile showed upregulation of miR‐20a/b and miR‐194, which target CDKN1A and SOCS2, respectively, potentially contributing to tumor‐cell proliferation and survival.56 Karube et al. (57) demonstrate that CD10‐negative FL cases are usually t(14;18)‐negative and/or morphologically high‐grade (Grade 3a or b), and, therefore, high‐dose intensive chemotherapy (R‐CHOP) is required. Leich et al. (58) report that these t(14;18)‐negative cases, whose subtype was initially described by Karube et al.,57 have a distinct miRNA profile frequently characterized by downregulation of miR‐16‐1, miR‐26a, miR‐101, miR‐29 and miR‐138.58 Because these miRNA are known to function as tumor‐suppressive miRNA, their downregulation may be associated with the pathogenesis of some FL subtypes and high‐grade FL.

Expression of miRNA in marginal zone lymphomas (MZL) has also been analyzed. The results suggest they are likely important in advanced stage or transformed cases of MZL and DLBCL. MiRNA expression has been analyzed in MALT type, nodal type and splenic type MZL. A study of gastric MALT type MZL revealed that high levels of miR‐223 expression are a marker of MALT stratification and correlate with increased E2A⁺ expression, higher clinical stage and diminished response to Helicobacter pylori eradication therapy.59 Large B‐cell lymphomas that originate in the stomach, and which are presumably derived from the MALT, exhibit a MYC‐miRNA signature, and transformation of MALT to DLBCL is associated with MYC, which negatively regulates miR‐34a, leading to downregulation of FOXP1.60 Studies of nodal MZL confirm that these tumors have distinctive features that distinguish them from FL cases. As compared to FL, nodal MZL shows greater expression of miR‐221, miR‐223 and let‐7f, which is a signature very similar to that exhibited by memory B‐cells and cells isolated from the normal marginal zone. Expression of these miRNA is enhanced in nodal MZL, whereas FL strongly expresses miR‐494. Upregulation of miR‐223 and miR‐221, which target the germinal center‐related genes LMO2 and CD10, could be partially responsible for expression of a marginal zone signature.61 In splenic MZL, the miR‐29 cluster is commonly lost and its expression silenced.62, 63, 64 In addition, increased expression of miR‐21 is associated with an adverse outcome in splenic MZL.63

Conclusion

MicroRNA have now been shown to play both oncogenic (e.g. miR‐17‐92) and tumor‐suppressive (e.g. miR‐15/16) roles in aggressive lymphoma subtypes (e.g. MCL and NK/T‐cell lymphoma) and relapsed cases (e.g. MCL). Some miRNA (e.g. miR‐34a) have also been shown to contribute to phenotypic transformation of malignant lymphoma (e.g. FL to DLBCL). However, only two miRNA (miR‐16‐1 and miR‐21) are known to be cancer‐inducible based on their activity: miR‐16‐1 knockout mice and transgenic mice overexpressing miR‐21, respectively, develop CLL and B‐cell leukemia with no original genetic event.65, 66 More often, however, miRNA dysregulation likely adds to original genetic events, and the resultant aberrations enhance tumorigenicity through activation of additional signaling pathways. Consequently, analysis of miRNA expression may be more useful for evaluation of disease progression and transformation than for classification of various lymphoma entities. For novel treatments against lymphoma, miRNA itself or the appropriate antisense could be useful therapeutic agents, but future functional studies with distinct lymphoma subtypes will be required to determine whether that is the case.

Disclosure Statement

The authors have no conflict of interest.

(Cancer Sci 2013; 104: 871–879)