A component of the mir-17-92 polycistronic oncomir promotes oncogene-dependent apoptosis

Thank you for sending your work entitled “A component of the mir-17-92 polycistronic oncomir promotes oncogene-dependent apoptosis” for consideration at eLife. Your article has been favorably evaluated by a Senior editor and 3 reviewers, one of whom, Chi Van Dang, is a member of our Board of Reviewing Editors.

The Reviewing editor and the other reviewers discussed their comments before we reached this decision, and the Reviewing editor has assembled the following comments to help you prepare a revised submission.

The manuscript by Olive et al. describes a very intriguing finding: while as a whole the miR-17∼92 cluster accelerates Myc-driven lymphomagenesis, its miR-92 component acts as a built-in damper, which induces apoptosis. Furthermore, deleting this component results in earlier-onset lymphomas. This central discovery was made using a model developed by Dr. He and her collaborators (Nature 2005; G&D 2009), wherein premalignant hematopoietic progenitors from Eμ-myc mice are transduced with various miR-encoding retroviruses and used to reconstitute irradiated recipients. The key data presented in Figures 1–3 are generally crisp, compelling, and easy to interpret. However several issues remain to be resolved.

Specificity of miR-92 activity:

1) The authors documented that miR-92 targets Fbw7 and thereby enhances MYC protein levels. While Figure 5D shows that miR-92 can repress a Fbw7 3'UTR luciferase reporter construct or Fbw7 cDNA expression construct, specificity of miR-92 should be established via mutating the predicted miR-92 binding site(s) within the 3'UTR and determine whether they are required for this repression.

2) Further experiments are needed to show whether the miR-92-Fbw7-Myc axis is fully responsible for miR-92's pro-apoptotic effects. In Figure 5F, it is shown that Fbw7 shRNA partially recapitulates the effect of miR-92 expression on Myc-mediated apoptosis in MEFs. The levels of Myc protein should be shown in this experiment. If Fbw7 knockdown fully recapitulates the miR-92-induced Myc levels yet does not fully recapitulate the degree of miR-92-induced apoptosis, it suggests that miR-92 engages additional mechanisms to induce apoptosis. To further investigate this issue, the authors should express ectopic Fbw7 (which does not have the miR seed sequence) at physiologic levels and see if this rescues the apoptotic phenotype of miR-92. These experiments should establish whether upregulation of Myc by miR-92 is the entire story or whether additional pro-apoptotic mechanisms exist. The authors are not required to identify such additional mechanisms in the current paper but it is important to know whether they exist.

3) To further substantiate their model, the authors should measure Fbw7 and Myc protein levels in Eμ-myc lymphoma cells expressing wild-type miR-17-92 versus those expressing miR-17-92Δ92. If miR-92-mediated Fbw7 repression/Myc induction cannot be demonstrated in this setting, the proposed mechanism, while elegant, could be completely irrelevant.

4) Along the same lines, what is the evidence that the effects of miR-92 on Myc levels are Fbw7-mediated? Perhaps Fbw7-null HCT116 cells could be used to establish causality.

5) The authors make a claim that that miR-92 is processed less efficiently in murine and human lymphomas than miR-19, the main oncogenic component of the cluster. This is an important claim, however, some of the quantifications are difficult to understand. For example, in panel 7D, miR-19 and -92 levels in Burkitt's cell lines are normalized separately for those found in “normal PB-cells”. Assuming that PB stands for “peripheral blood”, this does not appear to be a relevant control, since a circulating lymphocyte is not a cell of origin for Burkitt's – or other human lymphomas, for that matter. A direct comparison between miR-19 and miR-92 levels would be more helpful. According to a recent profiling paper [doi:10.1038/bcj.2012.1], miR-19b and miR-92 are overexpressed at comparable levels in Burkitt's samples.

Conceptual framework:

1) Although the authors focused on this ‘oncoMir’ cluster and studied its oncogenic properties, it would be terrific for the authors to discuss the potential physiological importance of this cluster with regard to its evolution as presented in the manuscript. In particular, it would be safe to assume that this cluster evolved to regulate cell growth and proliferation downstream or independent of MYC. Hence, the different miRs in the cluster might be subject to regulation via microRNA processing in addition to the expression of the cluster mRNA precursor. In this regard, are the relative levels of miR-92 to other miRs in the cluster differentially affected by cellular stresses that lead to apoptosis (serum or growth factor deprivation, nutrient deprivation?)? Some discussion on this aspect of miR-17-92 function could be very useful for the field.

2) In the Discussion, the authors describe miR-92 as conferring negative feedback on the oncogenic activity of miR-17-92. Given that miR-17-92 is transcriptionally activated by Myc and Myc dosage is positively regulated by miR-92, a positive feedback loop is also established. This concept should be discussed.

Influence of miR-92 on the miR cluster:

1) The expression of the miRNAs derived from the various MSCV constructs (miR-17-92; miR-17-92Δ92; miR-17-92Mut92) is tested by transducing 3T3 cells with these retroviruses (Figure 1B). However, the conclusions of the paper rest heavily upon the assumption that mutating miR-92 does not affect the expression of other miRNAs in the cluster in B cells (where the oncogenic activity is examined most extensively). Therefore it is important to examine the miRNA levels in Eμ-Myc lymphoma cells or primary B cells infected with the various viruses to confirm their findings in 3T3 cells.

Specificity of miR-92 activity:

1) The authors documented that miR-92 targets Fbw7 and thereby enhances MYC protein levels. While Figure 5D shows that miR-92 can repress a Fbw7 3'UTR luciferase reporter construct or Fbw7 cDNA expression construct, specificity of miR-92 should be established via mutating the predicted miR-92 binding site(s) within the 3'UTR and determine whether they are required for this repression.

We thank the reviewers for this comment. We have constructed luciferase reporters that carry either a wild type fbw7 3’UTR, or a mutated fbw7 3’UTR with defective miR-92 binding sites. Using these reporters, we clearly demonstrated that miR-92 overexpression could downregulate the expression of the luciferase reporter carrying the wild type fbw7 3’UTR, but not the luciferase reporter with the mutated fbw7 3’UTR. This result, shown in Figure 5D, demonstrates that Fbw7 is specifically repressed by miR-92, and that the miR-92 binding is required for its repression on Fbw7.

2) Further experiments are needed to show whether the miR-92-Fbw7-Myc axis is fully responsible for miR-92's pro-apoptotic effects. In Figure 5F, it is shown that Fbw7 shRNA partially recapitulates the effect of miR-92 expression on Myc-mediated apoptosis in MEFs. The levels of Myc protein should be shown in this experiment. If Fbw7 knockdown fully recapitulates the miR-92-induced Myc levels yet does not fully recapitulate the degree of miR-92-induced apoptosis, it suggests that miR-92 engages additional mechanisms to induce apoptosis. To further investigate this issue, the authors should express ectopic Fbw7 (which does not have the miR seed sequence) at physiologic levels and see if this rescues the apoptotic phenotype of miR-92. These experiments should establish whether upregulation of Myc by miR-92 is the entire story or whether additional pro-apoptotic mechanisms exist. The authors are not required to identify such additional mechanisms in the current paper but it is important to know whether they exist.

To investigate if the miR-92-Fbw7-Myc axis is fully responsible for miR-92's pro-apoptotic effects in vitro, we compared the effect of miR-92 overexpression and fbw7 knockdown on c-Myc protein level in R26^MER/MER mouse embryonic fibroblasts (MEFs) (Figure 5–figure supplement 1F). In this experiment, fbw7 knockdown largely recapitulated the extent of c-Myc upregulation by miR-92. This is consistent with our observation that the repression of fbw7 by miR-92 is essential for miR-92 to upregulate c-Myc (Figure 5–figure supplement 1E, also see our response to #4). Since fbw7 knockdown only partially phenocopies miR-92 in promoting c-Myc induced apoptosis, one possible scenario is that miR-92 engages additional mechanisms to promote c-Myc apoptosis. Nevertheless, the miR-92-Fbw7-Myc axis does constitute a major mechanism to mediate the pro-apoptotic effects of miR-92. To examine the importance of fbw7 in mediating the apoptotic effects by miR-92, we expressed fbw7 in R26^MER/MER mouse embryonic fibroblasts (MEFs) with and without miR-92 overexpression. In this experiment, the fbw7 cDNA introduced did not contain its 3’UTR, thus was not regulated by miR-92. Although miR-92 overexpression in R26^MER/MER MEFs invariably enhanced c-Myc induced apoptotic response upon MycERT(Bartel, 2009) activation, expression of fbw7 abolished this apoptotic effect of miR-92 (Figure 5H). Thus, the ability of miR-92 to increase c-Myc protein level through fbw7 repression constitutes the major mechanism underlying its pro-apoptotic effects.

3) To further substantiate their model, the authors should measure Fbw7 and Myc protein levels in Eμ-myc lymphoma cells expressing wild-type miR-17-92 versus those expressing miR-17-92Δ92. If miR-92-mediated Fbw7 repression/Myc induction cannot be demonstrated in this setting, the proposed mechanism, while elegant, could be completely irrelevant.

We thank the reviewers for this insightful comment. The experiment proposed here, if performed successfully, would strongly support our hypothesis. However, we have encountered technical limitations in detecting the endogenous Fbw7 protein in our tumor lysates. In our experience, we have not found any Fbw7 antibodies that can reliably detect endogenous Fbw7 proteins by simple immunoblotting. We have tested several commercial antibodies for detection of endogenous Fbw7, including those sold by Abcam, Sigma, and Invitrogen, and we are unable to detect endogenous Fbw7 cleanly, using proper controls (Fbw7-null HCT116 cell lysate). In lysates from cultured MEFs, which we can expand greatly, we use an immunoprecipitation-western blot method that does detect endogenous Fbw7 (Figures 5E), as detailed in our Methods section. The limitation of this approach is that one needs a large amount of cell pellet for this experiment. As an alternative, we performed fbw7 QPCR analyses, using Eμ-myc/17-92, Eμ-myc/17-19b, and Eμ-myc/MSCV lymphoma cells. Consistent with our hypothesis, Eμ-myc/17-92 B-lymphoma cells exhibited significantly decreased levels of fbw7 mRNA, when compared to those in Eμ-myc/17-19b or Eμ-myc/MSCV lymphoma cells (Figure 5–figuresupplement 1F).

We also measured c-Myc protein levels in several lines of Eμ-myc/17-92, Eμ-myc/17-19b, and Eμ-myc/MSCV lymphoma cells, to determine if there is a correlation between miR-92 overexpression and increased c-Myc dosage. However, we observed no differences in the c-Myc protein levels among these terminal tumor cells (data not shown). Previous studies have demonstrated that the terminal E-myc tumors, which are defective for c-Myc-induced apoptosis, clearly favor a high dosage of c-Myc to promote and maintain oncogenesis. In addition to the miR-92-Fbw7 axis that regulates c-Myc dosage, a miR-92 andfbw7 independent mechanism can also enhance c-Myc dosage in the transformed Eμ-myc lymphoma cells. Thus, comparing the c-Myc level in the terminal Eμ-myc/17-92, Eμ-myc/17-19b, and Eμ-myc/MSCV lymphoma cells is unlikely to reveal the importance of c-Myc regulation by the miR-92-Fbw7 axis, because this regulation plays an essential role in the early stages of lymphoma development (Figure 3A, 3B, Figure 6B).

4) Along the same lines, what is the evidence that the effects of miR-92 on Myc levels are Fbw7-mediated? Perhaps Fbw7-null HCT116 cells could be used to establish causality.

In the revised manuscript, we have clearly demonstrated that the overexpression of miR-92 increases c-MYC protein levels in a FBW7-dependent manner. The effect of miR-92to upregulate c-MYC protein level was observed in wild type Hct116 cells, but was largely absent in FBW7^-/- Hct116 cells (Figure 5–figure supplement 1E). These results argue that the repression of FBW7 by miR-92 is essential for miR-92 to upregulate the protein level of c-MYC.

5) The authors make a claim that that miR-92 is processed less efficiently in murine and human lymphomas than miR-19, the main oncogenic component of the cluster. This is an important claim, however, some of the quantifications are difficult to understand. For example, in panel 7D, miR-19 and -92 levels in Burkitt's cell lines are normalized separately for those found in “normal PB-cells”. Assuming that PB stands for “peripheral blood”, this does not appear to be a relevant control, since a circulating lymphocyte is not a cell of origin for Burkitt's – or other human lymphomas, for that matter. A direct comparison between miR-19 and miR-92 levels would be more helpful. According to a recent profiling paper [doi:10.1038/bcj.2012.1], miR-19b and miR-92 are overexpressed at comparable levels in Burkitt's samples.

We thank the reviewers for the constructive comments. We have realized that our wording in the previous manuscript has caused confusion. What is clear from our studies is that the ratio of miR-19 to miR-92 is greater in B-lymphomas than in normal B-cells. In other words, when normalized to the respective miRNA levels in normal B-cells, mature miR-19 exhibited a greater increase in premalignant and malignant Eμ-myc B-cells than mature miR-92 (Figure 7A, 7B, 7C). Since miR-19and miR-92 are coregulated transcriptionally, we speculate, but do not claim, that a differential miRNA biogenesis and/or turn over could explain the differential increase of these two miRNAs. Given their functional antagonism, the ratio between miR-19 and miR-92 is the key determinate for the oncogenic activity of mir-17-92 in the context of the Eμ-myc B-lymphoma model. What we showed here strongly supported an altered miR-19:miR-92 ratio in premalignant and malignant Eμ-myc B-cells, which favored a greater miR-19 increase to drive oncogenesis.

Per the reviewers’ request, we directly compared the miR-19 and miR-92 levels using miRNA Taqman asssays. Our data suggest a ∼2-5 fold increase in the absolute level of miR-19b than miR-92 in transformed B-cells, both in mouse and in human (data not shown). However, we must point out the intrinsic caveats associated with absolute quantitation of different mature miRNAs. Currently, two methods are most popular for the absolute quantitation of mature miRNAs miRNA Taqman assays or high-throughput sequencing (HTS). However, both methods have technical caveats that prevent an accurate quantitation. For the Taqman miRNA assays, the different RT efficiency for different mature miRNAs can introduce systematic bias in quantitation and preclude an accurate quantitation of different mature miRNAs. For the HTS approach, different mature miRNAs have different cloning efficiency due to RNA-ligase-dependent bias (Hafner et al., RNA 2011). Given the intrinsic technical limitations to accurately compare copy numbers of different mature miRNAs, we think it is the most appropriate to leave this out for our manuscript. We included a discussion about this issue in the revised manuscript.

We also clarified the legend of our Figure 7 to indicate the use of normal B-cells from periphery blood as a control for our Burkitt’s lymphoma cell lines. We admit that using B-cells from peripheral blood to control for human Burkitt’s lymphoma cell lines is less than ideal. However, such comparison has been used routinely for many published studies due to the difficulty to acquire human GC B-cell RNA as a proper control. We have included a statement in our revised manuscript to discuss this caveat for our comparison.

Conceptual framework:

1) Although the authors focused on this ‘oncoMir’ cluster and studied its oncogenic properties, it would be terrific for the authors to discuss the potential physiological importance of this cluster with regard to its evolution as presented in the manuscript. In particular, it would be safe to assume that this cluster evolved to regulate cell growth and proliferation downstream or independent of MYC. Hence, the different miRs in the cluster might be subject to regulation via microRNA processing in addition to the expression of the cluster mRNA precursor. In this regard, are the relative levels of miR-92 to other miRs in the cluster differentially affected by cellular stresses that lead to apoptosis (serum or growth factor deprivation, nutrient deprivation?)? Some discussion on this aspect of miR-17-92 function could be very useful for the field.

We thank the reviewers for the constructive comment. We have included a brief discussion on the functional significance of the mir-17-92 polycistronic structure in its physiological functions.

2) In the Discussion, the authors describe miR-92 as conferring negative feedback on the oncogenic activity of miR-17-92. Given that miR-17-92 is transcriptionally activated by Myc and Myc dosage is positively regulated by miR-92, a positive feedback loop is also established. This concept should be discussed.

We thank the reviewers for the insightful comment. In the revised manuscript, we have included a discussion on the positive feedback loop between mir-17-92 and c-Myc.

Influence of miR-92 on the miR cluster:

1) The expression of the miRNAs derived from the various MSCV constructs (miR-17-92; miR-17-92Δ92; miR-17-92Mut92) is tested by transducing 3T3 cells with these retroviruses (Figure 1B). However, the conclusions of the paper rest heavily upon the assumption that mutating miR-92 does not affect the expression of other miRNAs in the cluster in B cells (where the oncogenic activity is examined most extensively). Therefore it is important to examine the miRNA levels in Eμ-Myc lymphoma cells or primary B cells infected with the various viruses to confirm their findings in 3T3 cells.

We have examined the expression of all mir-17-92 components in the Eμ-myc B-lymphoma cells that overexpress mir-17-92, mir-17-92Δ92, or mir-17-92Mut92. Consistent with our results in the 3T3 cells (Figure 1–figure supplement 1D), mutation or deletion of miR-92 specifically disrupted the miR-92 expression in B-cell, without affecting the expression of the remaining mir-17-92 components (Figure 1D).