A facile method to increase titers of miRNA‐encoding retroviruses by inhibition of the RNaseIII enzyme Drosha

Ectopic expression of small non-coding microRNAs (miRNAs) through retroviral gene transfer is a powerful tool to decipher miRNA function and identify their cellular targets. miRNAs are non-coding ∼22-nt-long molecules that modulate gene expression at the post-transcriptional level by hybridizing to complementary sequences, mostly in the 3′-untranslated region of their corresponding mRNAs 1. Depending on the degree of base pairing, an miRNA either accelerates the degradation of the corresponding transcript or restricts its translation. miRNAs play an important role in T- and B-cell differentiation (e.g. miR-150, miR-155, miR-181 and the miRNA cluster miR-17∼92) 2. To address the function of miRNAs in B-cell activation, we adapted a retroviral system 3 to ectopically express selected miRNAs in freshly isolated splenic murine B cells. We first constructed the retroviral vector pCLEP, which is based on the murine stem cell virus-derived vector pCru5 4. Expression of miRNAs was accomplished by transcribing inserted genomic fragments of approximately 500 bp of the respective miRNA gene from promoter/enhancer elements in the long terminal repeat (LTR, Fig. 1A). pCLEP also encodes for enhanced green fluorescent protein (EGFP), which is linked to a puromycin resistance gene via an IRES element and in which expression is driven by an internal phosphoglycerate kinase promoter (PGK). The inhibition of Drosha increases frequency of retrovirally infected mouse B cells. (A) Schematic of retroviral vector pCLEP. The miRNA gene of interest is amplified by PCR from genomic DNA. (A, B), restriction enzymes used for cloning; PCMV/5′ LTR, hybrid of cytomegalovirus and long terminal repeat promoters; Ψ, packaging signal; PPGK, phosphoglycerate kinase promoter; EGFP, enhanced green fluorescent protein; IRES, internal ribosome entry site; Puro, puromycin resistance; ORI, origin of replication; Amp, ampicillin resistance. (B–C) 4×106 Phoenix Eco (ΦNX-E) cells in 100-mm dishes were CaPO4-transfected without (mock) or with pCLEP, pCLEP-106b only (-siRNA), and with pCLEP-106b and 200 pmol of either Drosha siRNA (siDrosha) or Luciferase siRNA (siControl). (B) After 48 h following transfection, cellular lysates from 5×105 cells were analyzed on western blots with anti-Drosha and anti-Dicer antibodies, an appropriate HRP-conjugated secondary antibody and the ECL method. Loading of equal cell equivalents was assessed with mouse antibodies against HSP90. (C) Transfected ΦNX-E and NIH3T3 cells infected with ΦNX-E supernatants were analyzed by flow cytometry 48 and 24 h after transfection and infection, respectively, for GFP-expressing cells. A representative example is shown. The graphs on the right represent the mean of three independent experiments and bars indicate the standard deviation. (D) MACS-sorted murine splenic B cells were pre-stimulated with LPS for 24 h before infection with viral supernatants from ΦNX-E transfected with pCLEP (control virus) or miRNA-expressing pCLEPs without (miR-106b and miR-30c virus) or with Drosha siRNA (+siDrosha). After 48 h, GFP was determined by flow cytometry in living cells (R1). Data are representative of three independent experiments. A representative example is shown. The graph on the right represents the mean of three independent experiments and bars indicate the standard deviation. The pCLEP control vector and pCLEP vectors encoding miR-150, miR-106b and miR-30c were transfected by the calcium phosphate method into the ecotropic retrovirus packaging cell line Phoenix Eco 5. As revealed by flow cytometry, transfection of Phoenix cultures with both miRNA-encoding and "miRNA-empty" pCLEP vectors resulted in similar frequencies (approximately 70–80%) of GFP-positive cells (Supporting Information Fig. 1A and Table 2). When NIH3T3 cells were infected with viral Phoenix supernatant, however, frequencies of GFP-positive cells were 1.5- (for miR-150 virus) to 18-fold lower (for miR-30c) in miRNA virus-infected NIH3T3 cultures compared to control virus-infected NIH3T3 cultures (Supporting Information Fig. 1B). We hypothesized that the full-length viral RNA carrying an miRNA gene could be recognized in Phoenix cells by the miRNA processing machinery, especially the RNaseIII enzyme Drosha. Drosha cleaves the primary miRNA transcript in the nucleus to generate the precursor hairpin miRNA 6. If full-length viral transcripts were suitable miRNA cleavage substrates, Drosha processing should reduce the amount of full-length viral transcripts available for packaging into virus particles. In this case, downregulation of Drosha using the siRNA technique should increase titers of miRNA-encoding retroviral particles. To test this hypothesis, we first determined the amount of Drosha-specific siRNA required to efficiently downregulate the enzyme by transfecting Phoenix cells via the calcium phosphate method with synthetic siRNA against Drosha. In western blots, the signal for Drosha was already reduced with 50 pmol and barely detectable with 800 pmol siRNA (Supporting Information Fig. 2). Next, we co-transfected Phoenix cells with a retroviral expression vector encoding miR-106b and with 200 pmol of Drosha siRNA or a control siRNA against luciferase. As expected, western blot analysis verified the successful downregulation of Drosha only in cultures that were co-transfected with siRNA against Drosha (Fig. 1B). As revealed by flow cytometry, frequencies of GFP-positive cells were quite similar in all transfected Phoenix cultures, ranging from 87 to 98% (Fig. 1C). Downregulation of Drosha did not lead to altered abundance of Dicer, the second RNaseIII enzyme needed to release mature miRNAs from the hairpin precursors. siRNA-mediated downregulation of Drosha in Phoenix cells should increase the amount of viral particles in the culture medium of cells transfected with retroviral constructs. As summarized in Supporting Information Fig. 3, this was indeed the case. More importantly, flow cytometry detected approximately 80% GFP-positive cells in NIH3T3 cultures that were infected with retroviral supernatants of Phoenix cells co-transfected with pCLEP-106b and 200 pmol Drosha siRNA (Fig. 1C), similar to the frequency of GFP-positive cells in NIH3T3 cultures infected with the empty control virus. In contrast, only 32 and 47% of NIH3T3 cells could be infected with miR-106b virus from Phoenix cells transfected without Drosha siRNAs or a control siRNA against luciferase, respectively. Transfection of Phoenix cells with 800 pmol of Drosha siRNA yielded a very similar picture (data not shown). We next confirmed the effect of Drosha siRNA with pCLEP-30c. Addition of siRNA against Drosha in the Phoenix transfection cocktail led to a three- to four-fold increase in GFP-positive cells in infected NIH3T3 cultures (data not shown). Therefore, titers of miRNA-encoding retroviral particles were increased by co-transfecting the packaging line with the retroviral expression vector and an siRNA against Drosha. To test whether the addition of Drosha siRNA also improves the transduction efficiency in primary B-cell cultures, we infected pre-activated primary splenic (CD43−) B cells with supernatants from Phoenix cells transfected either with the control vector pCLEP, with pCLEP-106b or with pCLEP-106b together with 200 pmol of Drosha siRNA (Fig. 1D). Similar to NIH3T3 cultures, the addition of Drosha-specific siRNA to the Phoenix transfection cocktail resulted in a twofold increase of GFP-positive cells in miRNA-virus infected B-cell cultures compared to cultures infected with supernatant from Phoenix cells transfected with pCLEP-106b in the absence of Drosha siRNA. Likewise, five-fold more GFP-positive cells were detected by flow cytometry in B-cell cultures infected with supernatants from Phoenix cells co-transfected with the miR-30c vector and Drosha siRNA (Fig. 1D). To verify that reduced transduction efficiencies of miRNA-encoding retroviral particles were due to Drosha-dependent processing of the primary RNA transcripts in the packaging cell line, we determined the abundance of mature miR-106b in Phoenix cells transfected with pCLEP-106b together with Drosha- or control siRNAs using quantitative TaqMan RT-PCR analysis (Supporting Information Fig. 4). If Drosha processes miRNA-carrying viral transcripts, reduction of Drosha abundance by Drosha siRNA should lead to a decrease in the abundance of mature miR-106b. This was indeed the case, as co-transfection of Phoenix cells with the expression vector pCLEP-106b and Drosha siRNA reduced the relative abundance of mature miR-106b by 50% when compared to that observed in Phoenix cells transfected with the miRNA vector either without Drosha siRNA or with a control siRNA against luciferase. Drosha siRNA transfection does not affect gag-pol- and env expression in the Phoenix packaging line, which shows that the observed effects are rather due to an increase in the abundance of proviral vector RNA than viral packaging proteins (Supporting Information Fig. 5 and Table 3). Hence, the inhibition of Drosha in the packaging cells results in impaired processing of mature miRNA from full-length retroviral transcripts, which leads to more full-length viral transcripts that can be packaged into infectious virus particles. Similar findings were recently reported by Poluri and Sutton, who showed that the titers of shRNA-containing lentiviral particles could be increased by co-transfection of Dicer siRNAs 7. In their study, however, processing of shRNAs did not rely on Drosha processing. In summary, if retroviral vectors carrying genomic miRNA genes are being used to ectopically express miRNAs, Drosha siRNAs should be used to increase infectivity. The authors thank Matthias Wabl (San Francisco) for providing pCru5, Javier Martinez (Vienna) for Dicer antibodies and Edith Roth for excellent technical assistance. This work was supported, in part, by the Deutsche Forschungsgemeinschaft (FOR832 & GRK592) to H.-M. J., the Hertha Löw Foundation to H.-M. J., the IZKF Erlangen and the Hiege Foundation to H.-M. J. and J. W., as well as the intramural ELAN Fonds to J. W. A. B. was supported by the DFG Training Grant GRK592. Conflict of interest: The authors declare no financial or commercial conflict of interest. Detailed facts of importance to specialist readers are published as "Supporting Information". Such documents are peer-reviewed, but not copy-edited or typeset. They are made available as submitted by the authors. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.