Live-cell imaging of microRNA expression with post-transcriptional feedback control

Abstract

MicroRNAs (miRNAs) are small noncoding RNAs that regulate complex gene expression networks in eukaryotic cells. Because of their unique expression patterns, miRNAs are potential molecular markers for specific cell states. Although a system capable of imaging miRNA in living cells is needed to visually detect miRNA expression, very few fluorescence signal-on sensors that respond to expression of target miRNA (miR-ON sensors) are available. Here we report an miR-ON sensor containing a bidirectional promoter-driven Csy4 endoribonuclease and green fluorescent protein, ZsGreen1, for live-cell imaging of miRNAs with post-transcriptional feedback control. Csy4-assisted miR-ON (Csy4-miR-ON) sensors generate negligible background but respond sensitively to target miRNAs, allowing high-contrast fluorescence detection of miRNAs in various human cells. We show that Csy4-miR-ON sensors enabled imaging of various miRNAs, including miR-21, miR-302a, and miR-133, in vitro as well as in vivo. This robust tool can be used to evaluate miRNA expression in diverse biological and medical applications.

Keywords: Csy4; fluorescent protein; live-cell imaging; microRNA; post-transcriptional feedback control.

Graphical abstract

Figure 1

Design of a Csy4-miR-ON sensor (A) Construction of a Csy4-miR-ON system. The Csy4-miR-ON sensor plasmid contains the bidirectional PGK and EFS promoters to drive expression of Csy4 and ZsGreen1 (ZsG), respectively. Four copies of miRNA target sequence are incorporated into the 3′ UTR of Csy4, whereas the Csy4 recognition site is incorporated immediately after the start codon of ZsG. (B) Schematic representation of the feedback mechanism of the Csy4-miR-ON system. Csy4 binds to the Csy4-specific stem loop within the ZsG mRNA and then cleaves it in the absence of the target miRNA, resulting in maintenance of the fluorescence-OFF state. In contrast, Csy4 mRNA is degraded when the miRNA binds to its target sequences. Next, ZsG mRNA is translated, which, in turn, generates a fluorescence signal.

Figure 2

Imaging of miR-21 expression by the Csy4-miR-ON sensor (A) Imaging of miR-21 expression in HeLa cells. HeLa cells were transfected with the Csy4-miR-ON sensor containing target sequences for miR-21 (21T) or scrambled sequences (CtrlT). One day after transfection, ZsG expression was examined via fluorescence microscopy and flow cytometry. Mean fluorescence intensity (MFI) of non-transfected cells (mock) was set to 1.0, and the relative MFI of Ctrl- and 21T-transfected cells is indicated. Data are presented as the mean ± SD (n = 3). ∗∗p < 0.001 versus mock. Pseudo-colored and phase contrast images are shown. (B) Imaging of miR-21 expression in HCT116 cells. Transfection and analyses were performed as described for (A). Data are presented as the mean ± SD (n = 3). ∗∗p < 0.001 versus mock. Scale bars, 200 μm.

Figure 3

Sensitive response of the Csy4-miR-ON sensors to miRNA activity (A) Evaluation of the Csy4-miR-ON sensor using an miRNA inhibitor. HeLa cells were co-transfected with the 21T sensor and miR-21 inhibitor, and ZsG expression was examined by fluorescence microscopy and flow cytometry 1 day after transfection. As a control, the control inhibitor (Ctrl inhibitor) was used rather than the miR-21 inhibitor. The MFI of non-transfected cells (mock) was set to 1.0, and the relative MFI of Ctrl inhibitor- and miR-21 inhibitor-transfected cells is indicated. Data are presented as the mean ± SD (n = 3). ∗p < 0.005, ∗∗p < 0.001 versus mock. Pseudo-colored and phase contrast images are shown; scale bar, 200 μm. (B) Evaluation of the Csy4-miR-ON sensor using an miRNA mimic. HeLa cells were co-transfected with the 302aT sensor and the miR-302a mimic, and analyses were performed as described for (A). As a control, the control mimic (Ctrl mimic) was used rather than the miR-302a mimic. The MFI of mock cells was set to 1.0, and the relative MFI of Ctrl mimic- and miR-302a mimic-transfected cells is indicated. Values and scale bar are the same as those described for (A). ∗∗p < 0.001 versus mock.

Figure 4

Improved sensitivity of the Csy4-miR-ON sensors (A) Construction of the Csy4-miR-ON sensor with modifications. The modified Csy4-miR-ON sensor (miRNA-TT′) contains four copies of miRNA target sequences at the 3′ UTR as well as a single copy of the miRNA target sequence at the 5′ UTR of Csy4, whereas the original Csy4-miR-ON sensor (miRNA-T) contains four copies of miRNA target sequences at the 3′ UTR of the Csy4 gene. (B) Imaging of miR-302a expression by the modified Csy4-miR-ON sensor. NT-2 cells were transfected with CtrlT, CtrlTT′, 302aT, or 302aTT′, and ZsG expression was examined via fluorescence microscopy 1 day after transfection. CtrlT and 302aT were constructed using the miRNA-T backbone, whereas CtrlTT′ and 302aTT′ were constructed using miRNA-TT′. Pseudo-colored and phase contrast images are shown; scale bar, 200 μm. (C) Quantitative evaluation of the modified Csy4-miR-ON sensor. Sensor plasmid-transfected cells were analyzed via flow cytometry. The MFI of non-transfected cells (mock) was set to 1.0, and the relative MFI of each sensor-transfected cell is indicated. Data are presented as the mean ± SD (n = 3). ∗p < 0.005, ∗∗p < 0.001 versus mock.

Figure 5

Capability of Csy4-miR-ON sensors for imaging analyses (A) Imaging of miRNA expression in hiPSCs and NHDFs. hiPSCs and NHDFs were transfected with CtrlTT′, Let7aTT′, or 302aTT′, and ZsG expression was examined via fluorescence microscopy 1 day after transfection. Pseudo-colored and phase contrast images are shown; scale bars, 200 μm. (B) Co-imaging of miR-302a and NANOG expression. hiPSCs were transfected with CtrlTT′ or 302aTT′, and NANOG expression was analyzed by immunofluorescence staining. Nuclei were counterstained with DAPI. Pseudo-colored images are shown; scale bars, 200 μm.

Figure 6

Imaging of a muscle-specific miRNA with the Csy4-miR-ON sensor (A) Live-cell imaging of miR-133 expression. C2C12 cells were transfected with CtrlTT′ or 133TT′, and miR-133 expression was monitored during myogenesis. Nuclei were counterstained with Hoechst 33342. Pseudo-colored images are shown; scale bar, 200 μm. MB, myoblasts; MT, myotubes. (B) Representative images of muscle section with Csy4-miR-ON sensors. Electric pulse-mediated gene transfer was performed to express CtrlTT′ or 133TT′ in the tibialis anterior (TA) muscle of C57Bl6 wild-type (WT) mice. Five days after gene transfer, the mice were sacrificed to collect TA muscles. Cross-section of TA were stained with Hoechst 33342 (blue, nucleus) and anti-laminin antibody (red, sarcolemma). Scale bar, 100 μm.