Response: Context-dependent actions of miR-106b in CLL

Response

We appreciate di Val Cervo et al's comments on our article about the role of miR-106b in repressing the ubiquitin ligase, Itch, to result in increases in the levels of p73 in CLL.¹ 

The ability of microRNA to interact with and suppress endogenous targets likely depends on the cellular context in which it is expressed. Associations of microRNAs and their targets determined in one cell type may not predict their association in another cell type. This concept is illustrated by several examples: miRs-17∼19b targeted Bim in one lymphoma cell line but not another² ; miR-29b targeted DNMT3a/b in lung³  but not in other cell types⁴ ; miR26a targeted EZH2 efficiently in leukemia but minimally in embryonic kidney cells.⁵  Our work demonstrated that ectopic expression of miR-106b in quiescent CLL cells resulted in a repression of Itch (n = 4) that was mechanistically associated with an increase in p73 levels. In contrast, over-expression of miR-106b in H1299 lung cancer cells did not decrease the levels of Itch protein (D.S., unpublished observations, June 2006). However, the levels of Itch protein were readily reduced in response to LBH589 suggesting that mechanisms unrelated to mir-106b were likely to regulate Itch in this cell line. Similarly, the findings presented in the letter show that miR-106 did not repress Itch in Saos2 osteosarcoma or HEK293T embryonic kidney cells. This result likely reflects the role of cellular context in determining miRNA and target gene associations, and may not recapitulate the actions of miR-106b in primary, quiescent CLL cells. In addition, miRNAs are likely to share a reciprocal relationship with their targets in individual cell types.⁶  miR-106b is over-expressed in several solid tumors and cell lines, shares a reciprocal relation with p21,^7,8  and represses p21 in such cells.⁸  However, in CLL, miR-106b and its host gene Mcm7 were epigenetically silenced, shared a reciprocal relation with Itch, not p21, and targeted Itch, not p21, in this disease.¹ 

We demonstrated a 1.4- to 4-fold increase in the levels of mir-106b in 12 of 19 CLL samples exposed to LBH589 (Figure 2A¹ ; P < .01, paired t test), whereas the cumulative data in Table 1 indicate an increase in miR-106b levels in 16 of 23 samples evaluated.¹  di Val Cervo et al demonstrated similar inductions in the levels of miR-106b upon 10 nM LBH exposure but statistically compared the induction of miR-106b with that of p73, DNp73 and miR34a, indicating that the objective of the 2 experiments differed considerably. In addition, we reported that transcriptional induction accounted for increases in p73 in a fraction of the samples evaluated (7/20). This suggests the existence of additional mechanisms that must account for the observed increases in p73 protein in samples that do not demonstrate a transcriptional induction. Finally, the catalytic activity of Itch was unaffected by caspase cleavage.⁹  Both full-length and the cleaved fragment of Itch were efficient at ubiquitilating their physiologic substrates.⁹  Consequently, caspase cleavage of Itch is unlikely to be unrelated mechanistically to the LBH-induced signaling pathway that results in increases in p73 protein, as described in our work.

The comments pertaining to miR-34a refer to work conducted by others¹⁰  and may need to be addressed separately.