Visualizing Nuclear Rnai Activity in Single Living Human Cells
Total Page:16
File Type:pdf, Size:1020Kb
Visualizing nuclear RNAi activity in single living PNAS PLUS human cells Shira Avivia,b,AmirMora,b,IrisDotanc, Sivan Tzadokc, Itamar Kantera,b, Noa Kinora,b, Dan Canaanic, and Yaron Shav-Tala,b,1 aThe Mina & Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat Gan 5290002, Israel; bInstitute of Nanotechnology, Bar-Ilan University, Ramat Gan 5290002, Israel; and cDepartment of Biochemistry & Molecular Biology, Faculty of Life Sciences, Tel-Aviv University, Tel Aviv 6997801, Israel Edited by Robert H. Singer, Albert Einstein College of Medicine, Bronx, NY, and approved September 11, 2017 (received for review May 4, 2017) Nuclear RNA interference (RNAi) is mediated by the canonical RNAi but, currently, there is little evidence for these mRNA:siRNA machinery and can lead to transcriptional silencing, transcriptional interactions in mammalian cells. For instance, short hairpin activation, or modulation of alternative splicing patterns. These RNAs (shRNAs) designed to target a promoter region were effects transpire through changes in histone and DNA modifications shown to cause transcriptional activation of the gene by binding via RNAi-mediated recruitment of chromatin-modifying enzymes. To to long noncoding RNAs (lncRNAs) transcribed from the pro- prove that nuclear RNAi occurs and modulates transcription in moter, although cleavage of the lncRNAs was not observed (19). human cells, we used live-cell imaging to detect and track nuclear Recently, a study in fission yeast has demonstrated that small RNAi transcriptional repression in single living human cells. While RNAs bind directly to the nascent pre-mRNA and yield epige- employing reporter genes constructed with inducible promoters netic repression (20). Moreover, those authors show that tran- and cognate-inducible short hairpin RNA (shRNA) targeted against scription must occur above a minimal threshold in order for the reporter coding region, we have characterized the dynamics of small RNA-directed heterochromatin to form. the nuclear RNAi process in living human cells. We show that the In this study, we set out to detect and follow the kinetics of silencing effect is mediated through the nascent mRNA, followed nuclear RNAi as it occurs in real time in single living human by activity of histone methylating enzymes, but not through cells. We expressed shRNAs targeting a GFP coding sequence DNA methylation. transcribed by genes with two different inducible promoters that can be tracked in living cells (21, 22). Transcription from the transcription | nuclear RNAi | histone methylation | live-cell imaging | targeted genes was significantly reduced in a time window of argonautes several hours after induction of shRNA expression. The silencing effect took place when the location of the GFP sequence in the NA interference (RNAi) is a cellular mechanism for con- gene was at the 5′ or at the 3′ of the gene, and far from the Rtrolling gene expression, in which the cell uses small RNAs to promoter, showing that the pre-mRNA was being targeted. achieve posttranscriptional specific gene silencing (1). The small Specific inhibitors of histone methyltransferases (HMTs) but not RNAs that mediate sequence-specific mRNA degradation, or DNA methylation reduced the silencing effect, as did knock- ∼ protein translation inhibition, are 21-nucleotide-long short- down of Argonaute (AGO) proteins and HP1γ, indicating that interfering RNA molecules (siRNA) generated from Drosha the RNAi-mediated transcriptional silencing was occurring on and Dicer cleavage of longer double-stranded RNA (dsRNA), or the gene via the recruitment of RNAi factors, histone modifying micro-RNAs (miRNA) generated from nuclear transcribed enzymes, and transient epigenetic control of transcription. dsRNA (2). RNAi results in sequence‐specific gene silencing through a variety of mechanisms (3–5). RNAi is considered a Results cytoplasmic process controlling gene expression long after Generation of an Inducible Gene Expression System and an Inducible mRNA transcription has occurred in the nucleus. However, shRNA System That Can Be Visualized in Living Cells. To examine many of the proteins of the RNAi machinery as well as the small RNAs have been detected in the nucleus (6, 7), and therefore Significance RNAi might have nuclear roles and the ability to act on tran- scripts immediately after their release from the parental DNA, or even during transcription. RNA interference (RNAi) is a natural process occurring in cells, and The nuclear process of RNAi in nonmammalian cells was is used to silence genes. Typically, RNAi occurs via small RNA initially identified in plants (8), and later recognized in other molecules generated in the cell nucleus, which are exported to organisms (9–12). In mammalian cells, nuclear RNAi was first the cytoplasm where they silence messenger RNA (mRNA) mol- CELL BIOLOGY described by the identification of small RNAs that shared se- ecules. However, RNAi is thought to occur in the nucleus as well. To demonstrate that this process can occur in the nucleus and to quence homology to promoter sequences, and silencing was at- determine its dynamics, we generated human cell systems that tributed to changes in histone modifications and DNA methylation enabled us to image living cells and to track gene silencing as it (13). While this was highly controversial at the time (6), cur- transpired in real time. We found that the RNAi machinery can rently, it is known that the RNAi machinery is present and active target the mRNA as it is being transcribed, and that silencing is in the nucleus (14), and it is generally accepted that several mediated through modifications occurring on histone proteins mechanisms of mammalian nuclear RNAi do exist. These can bound to the DNA. lead to RNAi-mediated gene silencing, gene activation, or modu- lation of alternative splicing, via recruitment of various enzymes Author contributions: S.A., A.M., I.D., D.C., and Y.S.-T. designed research; S.A., A.M., I.D., that yield changes in histone modifications (6, 7, 15, 16). S.T., and N.K. performed research; S.A., A.M., I.D., and I.K. analyzed data; and Y.S.-T. The small RNAs could act in the nucleus through binding to wrote the paper. DNA, RNA, or proteins of the transcription machinery (17). The authors declare no conflict of interest. Naturally, as in the cytoplasm, the nascent RNA that is com- This article is a PNAS Direct Submission. plementary to the small RNA would be a reasonable candidate 1To whom correspondence should be addressed. Email: [email protected]. for nuclear RNAi targeting. Indeed, cotranscriptional siRNA This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. binding and RNA processing were suggested in yeast cells (18), 1073/pnas.1707440114/-/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.1707440114 PNAS | Published online October 3, 2017 | E8837–E8846 Downloaded by guest on September 25, 2021 whether nuclear RNAi can be observed in human cells as it is also a gene system with a Tet-inducible promoter termed E6, occurs in real time, we decided to target mRNAs transcribed from which contains a gene with six exons and five introns, and is based genes with inducible promoters, for which the mRNAs can be de- on the β-globin minigene (21) (Fig. 1B). The last exon encodes an tected and tracked over time in single living human cells. The ap- in-frame cyan fluorescent protein (CFP) that comprises a peroxi- proach we describe here sets out to assess the levels of gene activity some targeting signal (Ser-Lys-Leu, SKL tripeptide), such that the in the nucleus under RNAi conditions, by using RNA fluorescence protein translated from this gene accumulates in peroxisomes. The in situ hybridization (FISH) with probes that will detect the active 3′UTR contains 18 MS2 sequence repeats that enable [after the genes. The levels of gene activity will be determined by measuring induction of transcription by doxycycline (dox)] the detection of the intensities of the active transcription sites. the transcribed E6 mRNAs in the cells, by a coexpressed YFP- The RNAi activity was designed to target GFP coding se- MS2 coat protein (YFP-MCP) that specifically binds to the quences of the transcribed mRNAs. To this end, we designed a C gene construct with a Tet-inducible promoter that encodes a MS2 loops formed in the transcribed E6 mRNAs (Fig. 1 ). combined transcript of turboRed fluorescent protein (tRFP), The second inducible gene system is induced by Ponasterone A which contains in its 3′UTR an shRNA sequence that targets the (PonA) and transcribes a GFP-fusion protein of dystrophin. The ′ mRNA sequence of GFP (Fig. 1A). The red fluorescent protein mRNA also contains MS2 sequence repeats in its 3 UTR (22) (Fig. D E serves as an indicator for shRNA expression, and the intensity of 1 and ). Both E6 and GFP-Dys genes are stably integrated in the red color should be proportional to the levels of the expressed human U2OS cells. Since the gene constructs were integrated as shRNA in individual cells. tandem gene arrays (23), they formed a strong and detectable site The Tet-inducible shRNA construct was coexpressed in two of transcription, and therefore have been used for tracking the types of inducible gene expression systems. The first cell system transcriptional activity of genes in individual cells (21, 22). Fig. 1. Cell systems used for inducible expression of shRNA and tagged mRNA for the detection of gene activity in fixed and living cells. (A) The shRNA-GFP expressing gene construct contains an inducible promoter (Tet-On + dox), a tRFP coding sequence fused in the 3′UTR to an shRNA module that targets GFP mRNA (sh-GFP). (B) The E6 gene construct containing an inducible promoter (Tet-On + dox) and a gene encoding an mRNA with the following modules: β-globin coding region fused to Cerulean fluorescent protein (CFP, cyan) followed by a 3′UTR containing 18× MS2 sequence repeats (gray), which allows the detection of the transcribed mRNA by YFP-MCP dimers (yellow) that bind the MS2 stem loops in the mRNA.