mia, an increase in leukemic blasts in the blood and bone marrow, and myeloid sarcomas in a small subset of animals.8 Inducible deletion of Dicer1 using Mx1-Cre resulted in defective competitive repopulation

nucleus by the Drosha RNase into premiRNAs (comprising 60 to 70 nucleotides). After transport into the cytoplasm, the premiRNAs are further processed by an RNase enzyme termed Dicer. miRNAs (and experimentally designed siRNAs) are generated consisting of 22 nucleotide RNA duplexes with two-nucleotide 3 overhangs. Dicer is part of the RNA-induced silencing complex that functions downstream of miRNA processing. Early studies demonstrated that constitutive loss of Dicer1 results in embryonic lethality at E8.5 because of loss of brachyuryexpressing stem cells.4 Conditional alleles of Dicer1 have shown the requirement of the enzyme in B cell5 and T cell development.6,7 Interestingly, conditional deletion of Dicer1 in mouse osteoprogenitors resulted in a myelodysplasia phenotype, which progressed into a secondary leukemia with splenomegaly, anemia, an increase in leukemic blasts in the blood and bone marrow, and myeloid sarcomas in a small subset of animals.8 Inducible deletion of Dicer1 using Mx1-Cre resulted in defective competitive repopulation and reduced reconstitution in secondary transplant recipients. miRNA profiling revealed a locus on chromosome 19 expressing 3 miRNAs. Remarkably, expression of miR-125a increased long-term multilineage reconstitution and one of miR125a targets was shown to be the proapoptotic protein Bak1.9 These previous studies help to frame the current analysis. A conditional Dicer1 allele was crossed with Cebpa-Cre to delete the enzyme in granulocyte-macrophage progenitors (GMPs). Cebpa-Cre;Dicer1fl/fl mice died shortly after birth, likely due to defective Cebpa-dependent induction of the lung epithelium in late gestation,10 similar to mice with conditional deletion of Dicer1 driven by a Sonic Hedgehog transgene.11 To circumvent the perinatal lethality observed in this model, Alemdehy et al transplanted fetal liver cells from mutant and control embryos into lethally irradiated recipient mice. No quantitative differences in Lin-Sca-1negKitpos, common myeloid progenitors (CMPs), GMPs, or megakaryocyte erythroid progenitors (MEPs) were observed in Dicer1-deficient mice. However, there was a 50% reduction in colonyforming unit– granulocyte macrophages (CFUGMs) from Dicer1 mutants. Culture of Lin ;Dicer1 / progenitors in GM-CSF resulted in myeloid cell dysplasia with myeloid cells with a hyposegmented nucleus, typical of Pelger-Huet Anomaly. Hyposegmented and bilobed neutrophils were also observed in vivo. This disease has been correlated with mutations in the Lamin B receptor.12 No changes in Lamin B receptor expression was observed in Dicer1-null GMPs. Whether Dicer1 affects other genes in the Lamin B receptor pathway or whether additional genes play a role in Pelger-Huet Anomaly remains to be resolved. Dicer1-deficient neutrophils also appear to have defect migration as Ly-6GposDicer1 / granulocytes were decreased in the peripheral blood and spleen. However, unlike the Osterix-Dicer1fl/fl mice,8 no myeloproliferative disease or acute myeloid leukemia was observed. The effect of Dicer1 deletion on gene expression in GMPs was examined by expression profiling. There were 300 significantly up-regulated transcripts including Bim, KRas, Hmga2, Hoxa9, and p21. Gene dosage appeared to also play a role as loss of 1 allele of Dicer1 affected transcript levels. While it is not surprising that Dicer1 plays a critical role in many cellular lineages, there is remarkable specificity in phenotypes observed with the conditional alleles of Dicer1 reported to date. Whether individual lineages have a unique miRNA target as has been shown with inducible Mx1-Cre remains to be explored.9 Several miRNAs play critical roles in the regulation of hematologic malignancies when gainof-function or loss-of-function mouse modeling experiments are analyzed.13 To identify the precise targets involved in lineage-specific determination, miRNA add-back screens should be devised. It will be interesting to determine whether single miRNAs will regulate specific cell progenitors as has been demonstrated for miR-125a.9 Gaining an understanding of the underlying biology of miRNAs will help validate this interesting group of macromolecules for therapeutic modulation. Conflict-of-interest disclosure: The author declares no competing financial interests. ■