The median survival of patients with mantle cell lymphoma (MCL) ranges from 3 to 5 years with current chemotherapeutic regimens. A common secondary genomic alteration detected in MCL is chromosome 13q31-q32 gain/amplification, which targets a microRNA (miRNA) cluster, miR-17∼92. On the basis of gene expression profiling, we found that high level expression of C13orf25, the primary transcript from which these miRNAs are processed, was associated with poorer survival in patients with MCL (P=0.021). We demonstrated that the protein phosphatase PHLPP2, an important negative regulator of the PI3K/AKT pathway, was a direct target of miR-17∼92 miRNAs, in addition to PTEN and BIM. These proteins were down-modulated in MCL cells with overexpression of the miR-17∼92 cluster. Overexpression of miR-17∼92 activated the PI3K/AKT pathway and inhibited chemotherapy-induced apoptosis in MCL cell lines. Conversely, inhibition of miR-17∼92 expression suppressed the PI3K/AKT pathway and inhibited tumor growth in a xenograft MCL mouse model. Targeting the miR-17∼92 cluster may therefore provide a novel therapeutic approach for patients with MCL.
Mantle cell lymphoma (MCL) is an aggressive hematological malignancy, comprising 5–10% of human B-cell lymphomas.1 The median survival of patients with MCL ranges between 3 and 5 years.2 Cytogenetically, MCL is characterized by the chromosomal translocation t(11;14)(q13;q32), which results in aberrant expression of cyclin D1.3 However, studies in transgenic mice indicate that the t(11;14)(q13;q32) alone is insufficient to result in lymphoma and additional genetic alterations are necessary.4, 5 Secondary genomic alterations are frequently detected in MCL,6 of which chromosome 13q31-q32 gain/amplification is one of the most common.7 Studies have shown that gain/amplification at 13q31-q32 targets a microRNA (miRNA) cluster, miR-17∼92, which resides within intron 3 of C13orf25 (or MIRH1), a non-protein-coding gene at 13q31.3.8, 9 The miR-17∼92 cluster consists of six miRNAs that are transcribed as a polycistronic unit.9, 10 Overexpression of miR-17∼92 has been observed in lymphomas and other solid tumors.9, 10 The oncoprotein MYC upregulates miR-17∼92 expression, and overexpression of the cluster accelerates MYC-induced lymphomagenesis in mice9, 11 and contributes to tumorigenesis.10, 12, 13, 14
The miR-17∼92 cluster is essential for B-cell development; in its absence, the proapoptotic protein BIM is upregulated, and B-cell development is blocked at the pro-B to pre-B transition.15 Similarly, expression of miR-17∼92 target genes, including BIM, is upregulated in Dicer-deficient pro-B-cells.16 Transgenic mice with higher expression of miR-17∼92 develop a lymphoproliferative disorder and autoimmunity and die prematurely. Lymphocytes from these mice show increased proliferation and attenuated cell death. In these mice, miR-17∼92 suppresses the expression of PTEN and BIM.17
Several genes in the phosphatidylinositol 3-kinase (PI3K)/AKT signaling pathway, including PIK3CA, AKT1, PDK1 and PPP1R2, are overexpressed in primary MCL tumors and MCL cell lines;18, 19 and the PI3K/AKT pathway is constitutively activated in a subset of MCL.20 Activation of PI3K/AKT signaling may be secondary to overexpression of miRNAs via downregulation of several negative regulators such as PTEN,21, 22 SHIP21 and PPP2R2A.13 However, the functional role of miRNAs in MCL remains elusive.
Herein, we show that high expression of miR-17∼92 is correlated with poorer survival in patients with MCL. We found that the protein phosphatase PHLPP2 (PH domain and Leucine-rich repeat Protein Phosphatase 2), a negative regulator of the PI3K/AKT pathway, is a direct target of the miR-17∼92 cluster, in addition to PTEN and BIM. We demonstrated that overexpression of miR-17∼92 in MCL cells activates the PI3K/AKT pathway and impedes chemotherapy-induced apoptosis by collaboratively down-modulating multiple negative regulators in the PI3K/AKT pathway and by decreasing BIM expression. We further showed that knockdown of miR-17∼92 miRNAs inhibits tumor growth in a xenograft MCL mouse model.
Materials and methods
MCL cell lines Z138c, Granta-519 and Jeko-1 were maintained in RPMI 1640, and HEK293T and NIH3T3 cells were maintained in DMEM. Both media were supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin.
Gene expression profiling
Gene expression profiling was performed on 92 patients with cyclin D1-positive MCL as previously described.23 A total of 82 cases were re-profiled using Affymetrix HG-U133 plus 2.0 arrays (http://llmpp.nih.gov.). The expression level of the C13orf25 gene was determined using probe set 232291 (UG cluster ID: Hs. 24115). The following MCL cell lines were also studied: HBL2, Jeko-1, Mino, Rec-1, SP53, UPN1 and Z138c. The study was approved by the Institutional Review Board at the University of Nebraska Medical Center.
Array-based comparative genomic hybridization (aCGH)
aCGH was performed as previously described.24 Genomic DNA from MCL cell lines was extracted using the DNeasy kit (Qiagen, Germantown, MD, USA). Normal male genomic DNA was used as reference. The aCGH protocol from NimbleGen Systems (Madison, WI, USA) was followed. Briefly, tumor and reference DNAs were fragmented by sonication and random-prime labeled with Cy3 and Cy5 dyes, respectively. Labeled material was cohybridized to microarrays consisting of 386 165 oligonucleotide probes spaced at ∼5 kb intervals throughout the human genome, washed and scanned (Axon Instruments, Union City, CA, USA).
The vectors were constructed as described in the Supplementary Materials and methods. The primers used are listed in Supplementary Table S1 and S2.
Establishment of MCL cell lines with ectopic expression or conditional knockdown of miR-17∼92
Z138c and Granta-519 cell lines were stably transduced with pRevTet-On (Clontech, Palo Alto, CA, USA) to generate stable cell lines (Z138c-Tet-On and Granta-Tet-On) according to the manufacturer's instructions. The Z138c-Tet-On and Granta-Tet-On cell lines was then transduced with the TMP2-miR-17∼92 vector (Supplementary Figure S1A). These cells were selected with puromycin, and GFP-expressing cells were isolated by fluorescence-activated cell sorting. The cell lines were grown in the presence of doxycycline (1 μg/ml) to maintain the overexpression of miR-17∼92. For knockdown of miR-17∼92, Jeko-1 cells were transduced with lentivirus packaged with the pTRIPZ-Sponge construct (Supplementary Figure S1B), pMD2G envelope vector and psPAX2 packaging plasmid in HEK293T cells. Transduced cells were selected with puromycin, and doxycycline-induced RFP-expressing cells were isolated by fluorescence-activated cell sorting.
HEK 293T or NIH3T3 cells were plated at 1 × 105 cells per well in a 24-well plate 24 h before transfection. pGL3-promoter plasmids containing the wild-type or mutated 3′-UTRs of various potential targets were co-transfected with pRL-SV40 using Exgen 500 (Fermentas, Hanover, MD, USA). For miRNA inhibition studies, sponge plasmids targeting particular miRNAs were co-transfected into HEK293T cells with 3′-UTR reporter plasmids and pRL-SV40. Luciferase assays were performed 24 h after transfection using the Dual Luciferase Reporter Assay System (Promega, Madison, WI, USA). Firefly luciferase activity was normalized to Renilla luciferase activity for each reaction. Transfected wells were analysed in triplicate for each group.
Six- to eight-week-old female CB-17/SCID mice (The Jackson Lab, Bar Harbor, ME, USA) were subcutaneously inoculated in the flank with 5 × 106 Jeko-pTRIPZ-control or Jeko-pTRIPZ-Sponge cells suspended in 100 μl PBS. Tumor growth was assessed using the two largest perpendicular axes measured with standard calipers. Tumor volume was calculated based on the formula V=π × L × S × S/6 (L, the long axis; S, the short axis). Tumor size and body weight were measured every other day after inoculation of MCL cells. Animals were killed when the largest tumor diameter reached 20 mm or after the loss of >10% of body weight. All animal studies were conducted in accordance with the NIH guidelines for animal care. All experimental procedures and protocols were approved by the Institutional Animal Care and Use Committee at the University of Nebraska Medical Center.
Data were analyzed with the Student's t-test using the SPSS 11.0. program (IBM Corporation, Armonk, NY, USA). P<0.05 was considered statistically significant. Data are presented as mean ± s.e.m.
Expression of C13orf25 correlates with overall survival in patients with MCL
We previously profiled 92 primary samples of MCL23 and found that the proliferation signature is the main predictor of clinical outcome in these patients. Of these, 82 cases were reexamined using Affymetrix HG-U133 plus 2.0, which showed that c13orf25 were highly expressed in these cases. The expression of individual miRNAs within the miR-17∼92 cluster was also evaluated in a subset of these cases by quantitative RT-PCR assay. Compared with normal naive B-cells (n=3), the tumor cells in MCL cases (n=30) expressed two to three-fold higher levels of miR-17∼92 miRNAs (Supplementary Figure S2). To further study the prognostic significance of c13orf25 overexpression, we identified five cases which had significantly higher levels of c13orf25 (> mean + 1.5 s.d.) compared with the other 77 MCL cases. All five cases showed a high proliferation signature (Figure 1a), and four had 13q31-32 gain/amplification by aCGH analysis (data not shown). Gene set enrichment analysis25 revealed the enrichment of gene signatures that were highly associated with cell proliferation in these cases (Supplementary Table S3). Kaplan–Meier analysis showed that patients with high levels of C13orf25 had poorer overall survival (P=0.021) with median survival of only 1.06 years, compared with median survival of 2.75 years for the rest of the cohort (Figure 1b).
MiR-17∼92 overexpression inhibits chemotherapy-induced apoptosis in MCL cells
To study the effects of miR-17∼92 overexpression, we established virally transduced MCL cell lines, Z138c-miR-17∼92 (Z138c-miR) and Granta-519-miR-17∼92 (Granta-miR). Compared with their respective control cells, Z138c-miR and Granta-miR cells expressed ∼3.6-fold and 2.9-fold higher levels of miR-20a, determined by real time RT-PCR. Similarly higher levels of miR-17 were also demonstrated in these cells (Figure 2a).
Topotecan is a potent topoisomerase I inhibitor, which induces DNA damage and leads to apoptosis in cells through the intrinsic pathway in a caspase 3-dependent manner.26 Overexpression of miR-17∼92 significantly decreased cell death in Z138c cells and Granta-519 cells treated with topotecan, as determined by a proliferation assay (Figure 2b) or annexin V-staining (Figure 2c). Similar results were also seen when these cell lines were treated with doxorubicin or etoposide, a topoisomerase II inhibitor (Supplementary Figure S3A and B). In addition, significantly smaller fractions of cleaved poly-(ADP-ribose) polymerase and cleaved caspase 3 were present in Z138c-miR compared with control cells (Figure 2d). Conversely, knockdown of miR-17∼92 activity using a mixture of antagomirs against individual miRNAs within the miR-17∼92 cluster (Applied Biosystems/Ambion, Carlsbad, CA, USA) decreased cell survival when treated with etoposide in Mino and Z-138c cells, but only minimally in Jeko-1 cells (Supplementary Figure S3C and D). Furthermore, treatment with antagomirs increased the number of apoptotic cells and increased caspase 3/7 activity in Z138c (Supplementary Figure S3E and F) and Mino cells (data not shown). Our observations suggest that overexpression of miR-17∼92 protects the MCL cells from apoptosis when challenged with chemotherapeutic agents and, conversely, that knock down of the activity of miR17-92 promotes chemotherapy-induced apoptosis in MCL cells.
Overexpression of miR-17∼92 miRNAs downregulates PTEN and BIM
To gain insight into the molecular mechanisms by which miR-17∼92 overexpression enhances tumor cell resistance to chemotherapy-induced apoptosis, we searched for putative miR-17∼92 target genes predicted by TargetScan, Pictar and miRbase programs. Several genes with known roles in proliferation and apoptosis, such as PTEN and BIM, are among the top predicted targets of miR-17∼92 based on a series of established criteria (Supplementary Table S4).27
As PTEN has been previously reported as direct target of the miR-17∼92, we examined its protein level in Z138c-miR cells by immunoblot analysis. Compared with the control cells, the protein level of PTEN were significantly reduced in cells with miR-17∼92 overexpression (Figure 3a). To demonstrate that downregulation of PTEN was mediated through miR-17∼92 binding to the 3′-UTR of PTEN mRNA, we constructed a luciferase reporter plasmid containing a segment of the PTEN 3′-UTR that includes predicted binding sites for miR-20a/-17-5p and miR-19. We also constructed a luciferase reporter plasmid containing point mutations in the predicted miRNA-binding sites within the PTEN 3′-UTR (PTENmut) (Supplementary Figure S4A). The luciferase activity of the reporter gene containing the wild-type PTEN 3′-UTR was decreased 40% compared with the control construct in HEK293T cells, in which the miR-17∼92 cluster is highly expressed,17 but the decrease was not seen with the reporter gene PTENmut (Supplementary Figure S4B). To further demonstrate that the effects seen were directly mediated by miR-17∼92, we co-transfected the 3′-UTR reporter plasmid PTEN or PTENmut along with TMP2 or TMP2-miR-17∼92 plasmid into NIH3T3 cells, which have much lower levels of endogenous miR-17∼92.17 We found that enforced miR-17∼92 expression in NIH3T3 cells decreased the luciferase activity of the reporter gene with a wild-type PTEN 3′-UTR, but not the reporter gene PTENmut (Supplementary Figure S4C). Furthermore, we studied the effect on targets upon knocking down these miRNAs in MCL cells. The miR-20a ‘sponge’ construct expresses a mRNA with seven tandem miR-20a binding sites in the 3′-UTR and acts to sequester miR-20a miRNA.28 Coexpression of the miR-20a ‘sponge’ construct in HEK293T cells upregulated PTEN protein expression by 72% (Supplementary Figure S4D). Moreover, the luciferase activity in the PTEN 3′-UTR reporter gene was restored in HEK293T cells by coexpression of the miR-20a ‘sponge’ plasmid (Supplementary Figure S4E). However, there was no significant effect on the luciferase activity in reporter genes in which these binding sites were mutated (Supplementary Figure S4E). The findings confirmed that PTEN was a direct target of the miR-17∼92 cluster.
BIM, a BH3-only proapoptotic protein, is also markedly downregulated in Granta-miR cells (Figure 3b). Among the three cloned fragments of the BIM 3′-UTR (Supplementary Figure S5A), B2 contains binding sites for miR-17-5p/-20a and -92, whereas B3 contains binding sites for miR-19 and -92. The luciferase activity of a reporter plasmid containing the wild-type B2 or B3 was decreased 40% compared with the control construct (Supplementary Figure S5B). Luciferase activity was partially restored by co-transfection with sponge plasmids against miR-20a and -92, and to a lesser extent miR-19 (Supplementary Figure S5C). These findings suggest that miR-20/-17-5p and miR-92 have an important role in BIM downregulation.
MiR-17∼92 cluster modulates the PI3K/AKT pathway in MCL cells
We then examined the activity of the PI3K/AKT pathway in MCL cells. Immunoblot analysis showed higher levels of phosphorylated AKT (p-AKT) and its targets, GSK-3β and p70S6K, in Z138c-miR than in control cells (Figure 3c). Similarly, an intracellular immunofluorescence assay (Figure 3d), showing that the level of p-AKTSer473 in Z138c-miR, was significantly higher than that in control cells. Our findings demonstrate that miR-17∼92 overexpression activates the PI3K/AKT pathway, presumably at least in part through downregulation of PTEN.
PHLPP2, a negative regulator of the AKT pathway, is a direct target of the miR-17∼92 cluster
We have also analyzed the 3′-UTR of other known negative regulators of the PI3K/AKT pathway and found that the 3′-UTR of PHLPP2 contains sequences matching the seed sequences for miR-18, miR-20a/-17-5p and miR-92 (Figure 4a). Immunoblotting analysis demonstrated that PHLPP2 protein was also downregulated in Z138c-miR cells compared with the control cells (Figure 3a). The luciferase activities of report plasmids P1, which contains putative binding sites for miR-18 and miR-20a/-17-5p, and P2, which does not contain any known binding sites for the miR-17∼92 miRNAs, were comparable to the control plasmid (Figure 4b). However, the luciferase activity of reporter plasmid P3, which contains two putative binding sites for miR-92, was decreased ∼40% compared with the control, but the decrease was not seen when both binding sites for miR-92 were mutated (P3-92-2M) (Figure 4b). The finding was further confirmed using a miR-92 sponge plasmid that restored ∼80% of the luciferase activity of P3 (Figure 4c). However, sponge plasmids for miR-18 or miR-20 failed to restore the luciferase activity of their respective reporter plasmids. Furthermore, we showed that the sponge plasmids for miR-92 or the entire miR-17∼92 cluster, but not the sponge plasmids for miR-18, -19 or -20, upregulated PHLPP2 protein levels in HEK293T cells (Figure 4d). The findings indicated that PHLPP2 is a direct target of the miR-17∼92 cluster and more specifically, that miR-92 has an important role in PHLPP2 regulation.
We performed gene expression profiling analysis in MCL cell lines including HBL2, Jeko-1, Mino, Rec-1, SP53, UPN1 and Z138c along with 82 primary MCL tumor samples. Expression levels of PHLPP2 in Mino, Rec-1 and SP53 were significantly decreased and reached only 50% of the average expression level in primary MCL tumors and cell lines studied. We also performed aCGH in these cell lines and found that Rec-1 contained a single-copy ∼7 Mb deletion in chromosome 16 that includes PHLPP2 (Figure 4e).
Knockdown of the miR-17∼92 cluster suppresses tumor cell proliferation in vitro and tumor growth in a xenograft MCL mouse model
We constructed a doxycycline-inducible sponge lentiviral construct targeting all the miRNAs within the cluster (Supplementary Figure S1B). By transducing it into Jeko-1 cell line, which expresses a high level of miR-17∼92,29 we generated a stable cell line (Jeko-Sponge (Jeko-S)) in which miR-17∼92 could be conditionally knocked down. As the miRNAs were sequestered instead of being degraded, we used a functional assay such as measuring the levels of target proteins or the activity of AKT pathway to confirm that the miRNAs were knocked down in our system. Specifically, upon doxycycline administration, the protein levels of PHLPP2 and PTEN in Jeko-S cells increased ∼72% and 32%, respectively, compared with the control cells (Jeko-Control) with or without doxycycline; and Jeko-S cells without doxycycline); and p-AKT decreased to ∼75% of the control levels (0.62/0.83) (Figure 5a). As a consequence, Jeko-S cells showed decreased cell proliferation upon doxycycline administration (Figure 5b).
Next, we developed a xenograft MCL mouse model by inoculating Jeko-1 cells into the flanks of NOD-SCID mice. After the inoculation of 3 weeks, all mice developed measurable tumors. There was no difference in tumor size among mice inoculated with Jeko-Control or Jeko-S cells without doxycycline treatment. To study the effect of miR-17∼92 inhibition on tumor growth, half of the mice were administered doxycycline. Whereas there was no difference in tumor growth in mice inoculated with Jeko-Control cells with or without doxycycline, growth of Jeko-S tumors treated with doxycycline was markedly suppressed compared with untreated Jeko-S tumors (1.96±0.89 g versus 3.36±0.74 g, P<0.001) (Figures 5c and d). PHLPP2 and PTEN protein levels were upregulated in doxycycline-treated Jeko-S tumors but not in controls. Conversely, the levels of p-AKT and RPS6 were decreased 20% and 50%, respectively (Figure 5e). Cell cycle analysis showed that knockdown of miR-17∼92 expression blocked the G1–S phase transition without significantly affecting apoptosis (Figure 5f). Similarly, using the TUNEL assay, we did not detect any significant changes in apoptosis between Jeko-S-tumors treated with or without doxycycline (data not shown).
In the current study, we describe a novel oncogenic pathway underlying the pathogenesis of MCL: that is, overexpression of miR-17∼92 targets the protein phosphatase PHLPP2, in addition to PTEN and BIM, and thereby impedes chemotherapy-induced apoptosis in MCL cells. We showed that high expression of the miR-17∼92 cluster was associated with poorer survival in MCL patients. Inhibition of miR-17∼92 augmented the protein level of PTEN and PHLPP2, and inhibited tumor growth in vivo.
Constitutive activation of the PI3K/AKT pathway has been shown to contribute to survival in a subset of MCL. Rudelius et al.,20 found loss of PTEN expression in 5 of the 17 cases with activated PI3K/AKT. Although loss of PTEN expression may be secondary to gene mutation, gene deletion or epigenetic mechanisms, our data demonstrate that miRNAs, specifically miR-17∼92, have an important role and provide a novel mechanism for PTEN downregulation in MCL.
There are clearly many potential causes for PI3K/AKT activation in MCL other than PTEN alterations, such as amplification of one of the AKT genes, activation or mutation of one of the three RAS proto-oncogenes and activation or mutation of a wide range of cellular receptors. Nevertheless, our recent study of diffuse large B-cell lymphoma showed that the gain/amplification of the chromosomal region containing miR-17-92 occurs only in the germinal center-B-like subtype and is mutually exclusive of PTEN deletion,24 suggesting that PI3K/AKT activation may be mediated through the miR-17∼92/PTEN pathway.
PI3K/AKT signaling can be terminated through at least two different mechanisms: removal of the activating lipid second messenger, catalyzed by PTEN,30 or dephosphorylation of activated AKT. Dephosphorylation of AKT is mediated by PP2A-type phosphatases31 and members of a novel phosphatase family, PHLPP, which includes PHLPP1 and PHLPP2 (also known as PHLPPL).32 In the current study, we demonstrated that PHLPP2 is a direct target of miR-17∼92, suggesting that miR-17∼92 may target both PHLPP2 and PTEN to activate the PI3K/AKT pathway in MCL. Although PHLPP1 and PHLPP2 both dephosphorylate the same residue within the hydrophobic phosphorylation motif on AKT, they differentially terminate AKT signaling by regulating distinct AKT isoforms. It has been suggested that PHLPP1 has a role in glucose homeostasis, whereas PHLPP2 has a role in cell survival. Notably, whereas PHLPP1 modulates the phosphorylation of MDM2 and GSK-3α through AKT2, PHLPP2 specially modulates p27 phosphorylation through AKT3 and the phosphorylation of GSK-3β and the FOXO family of forkhead transcription factors through AKT1.32, 33 Therefore, downregulation of PHLPP2 may lead to a constitutive activation of AKT, especially AKT1 and AKT3, resulting in inactivation of FOXO family members, GSK-3β and p27kip1 through phosphorylation. Further investigation of the role of PHLPP2 in MCL is warranted, particularly the potential synergistic effect with the down-modulation of PTEN expression on PI3K/AKT signaling. The notion of PHLPP2 as a tumor suppressor is further supported by our finding that it is deleted in one of the MCL cell lines, Rec-1 and the levels of PHLPP2 mRNA are decreased in several MCL cell lines.
BIM is one of the most potent pro-apoptotic BH3-only proteins, which binds to all pro-survival BCL2 family members with high affinity,34, 35 thereby releasing BAX and BAK proteins, the critical downstream effectors of the BCL2-regulated pathway of apoptosis.36 Homozygous deletions of BIM occur in several MCL cell lines including Jeko-1, SP53 and Z138c, and BIM expression is reduced in Rec-1, but not in Granta-519 or JVM2 cells.37 Our study showed that overexpression of the miR-17∼92 cluster markedly down-modulated BIM and provided a novel mechanism for BIM downregulation in MCL. Repression of BIM by the miR-17∼92 cluster could also explain the observation that miR-17∼92 overexpression synergizes with MYC in a mouse model of B-cell lymphomagenesis, as a similar synergy in lymphomagenesis has been observed by combining MYC overexpression with BIM deficiency.38 Activation of the PI3K/AKT pathway may inactivate the FOXO family of transcription factors, leading to decreased transcription of BIM.39, 40, 41 The marked BIM down-modulation therefore could be attributed to a combined effect of both direct targeting of BIM by miR-17∼92 and indirect transcriptional suppression through FOXO inactivation.39, 40, 41 The miR-17∼92 cluster may therefore control signaling networks regulating cell survival at multiple levels, in a cooperative fashion, in MCL; thus, modest effects on multiple factors in the regulatory pathway may result in marked changes in BIM protein levels and in AKT activation.
High expression of miR-17∼92 facilitates cell proliferation and inhibits apoptosis in lymphocytes.15, 16, 17 MCL patients with high miR-17∼92 expression also showed the enrichment of gene signatures that were highly associated with cell proliferation. Inhibition of miR-17∼92 in Jeko-1 cells modestly decreased proliferation in vitro but induced a dramatic G1/S arrest in vivo. These results suggest that the miR-17∼92 cluster has an important role in tumor proliferation. In addition to PTEN and PHLPP2, other targets of miR-17∼92 may also participate in regulating proliferation, such as p2142 and p57.43
The much greater effect of inhibition of miR-17∼92 in vivo than in vitro is intriguing and suggests that miR-17∼92 may modulate tumor/microenvironment interactions to facilitate tumor proliferation in addition to its effects on tumor cells per se. In particular, miR-17∼92 has been shown to promote angiogenesis through several pathways. In RAS-expressing cells, miR-17∼92 promotes tumor angiogenesis in vivo by targeting the anti-angiogenic proteins thrombospondin-1 and connective tissue growth factor.12 Mir-17∼92 may also inhibit the TGFβ pathway and attenuate its antiangiogenic effects.44 Furthermore, it has been shown that miR-92 may target the von Hippel–Lindau gene product, and increases VEGF expression.45
The miR-17∼92 cluster is known to be upregulated by the E2F family.46 In addition, not only is it upregulated by MYC, but we also recently reported evidence that inhibitory members of the MYC family repress transcription of the locus.47 Indirect activation of AKT by miR-17∼92 is expected to provide positive feedback, as AKT, by inhibiting GSK3β, decreases phosphorylation of MYC at threonine-58 and thereby decreases its degradation.48 Similarly, activation of AKT is expected to result in inhibition of p27 and p21, with activation of CDKs, inhibition of the RB family, and activation of E2F.49 Thus, the miR-17∼92 miRNAs may promote their own transcription through AKT activation.
The finding that inhibition of miR-17∼92 activity significantly reduced tumor growth in xenograft mice indicates that miR-17∼92 may serve as a target for tumor therapy. Inhibition of individual miRNAs in the miR-17∼92 cluster has previously been shown to inhibit tumorigenesis and tumor growth in different cancers, including multiple myeloma50 and neuroblastoma.51 Targeting miR-17∼92 cluster in MCL may therefore provide a new therapeutic approach for this incurable disease.
This work was supported in part by the National Institutes of Health grant U01 CA114778 to WCC; the Lymphoma Research Foundation/Millennium Pharmaceuticals, Inc. Clinical Investigator Career Development Award to KF; and UNMC Eppley Cancer Center Pilot Grant to KF; MJ and XH are supported by a scholarship from the China Scholarship Council.
About this article
TWM, WCC and KF designed the study and provided administrative support; WCC and KF provided the study materials and patient information; ER, CJ, MJ, JI, XH, GL, GW and KF collected and assembled the data; ER, CJ, JI, GL, GW, LS, YZ, TWM, WCC and KF performed data analysis and interpretation; ER, CJ, TWM and KF wrote the manuscript; all authors have checked and approved the final version of the paper.
Supplementary Information accompanies the paper on the Leukemia website (http://www.nature.com/leu)
Molecular Biology Reports (2018)