5' DREDGE: Direct Repeat-Enabled Downregulation of Gene Expression via the 5' UTR of Target Genes

Abstract

Despite the availability of numerous methods for controlling gene expression, there remains a strong need for technologies that maximize two key properties: selectivity and reversibility. To this end, we developed a novel approach that exploits the highly sequence-specific nature of CRISPR-associated endoribonucleases (Cas RNases), which recognize and cleave short RNA sequences known as direct repeats (DRs). In this approach, referred to as DREDGE (direct repeat-enabled downregulation of gene expression), selective control of gene expression is enabled by introducing one or more DRs into the untranslated regions (UTRs) of target mRNAs, which can then be cleaved upon expression of the cognate Cas RNase. We previously demonstrated that the expression of target genes with DRs in their 3' UTRs are efficiently controlled by the DNase-dead version of Cas12a (dCas12a) with a high degree of selectivity and complete reversibility. Here, we assess the feasibility of using DREDGE to regulate the expression of genes with DRs inserted in their 5' UTRs. Among the five different Cas RNases tested, Csy4 was found to be the most efficient in this context, yielding robust downregulation with rapid onset in doxycycline-regulatable systems targeting either a stably expressed fluorescent protein or an endogenous gene, both in a fully reversible manner. Unexpectedly, dCas12a was also found to be modestly effective despite binding essentially irreversibly to the cut mRNA on its 5' end and thereby boosting mRNA levels. Our results expand the utility of DREDGE as an attractive method for regulating gene expression in a targeted, highly selective, and fully reversible manner.

Keywords: CRISPR; DREDGE; direct repeat; endoribonuclease; gene regulation.

Figure 1

Cas RNases screened for suitability for 5′ DREDGE. (A) Cartoon illustrating the basic principles of 5′ DREDGE. DRs (green) placed into the 5′ UTR of an mRNA are cleaved by Cas RNases (red), resulting in the removal of the 7-methylguanosine (m7G) 5′ cap, which in turn impairs translation by multiple mechanisms and also triggers mRNA degradation via 5′-3′ exonuclease activity (see text). (B) DRs for the five Cas RNases tested in this study. Cleavage sites are indicated with red arrows and the sizes in nucleotides (nts) are shown. Note the “synSeparator” (AAAU) upstream of the 5′ end of the Cas12a DR.

Figure 2

Comparative efficacy of 5′ DREDGE implemented with five different Cas RNases and their cognate DRs. (A) Overview of the construction of the vectors encoding GFPd2 with individual cognate DRs (green) in the 5′ UTR (top) and vectors co-expressing different Cas RNases (red) together with mCherry (bottom). Constructs lacking a DR or RNase served as controls. (B) Cell cytometry results from MEFs transiently transfected with the constructs in (A). Depicted are the percentages of cells in Q2 relative to controls (top), which were quantified from log-log plots of GFP vs. mCherry RFU values (n = 3 replicates), with representative plots for each condition provided (bottom). (C) GFP intensity in mCherry-positive cells, normalized to controls, derived from the RFU plots shown in (B), together with representative images of GFP fluorescence in cells from the various conditions (bottom), which were acquired immediately prior to cytometry.

Figure 3

Characterization of 5′ DREDGE using an inducible system. (A) Cartoon depicting the generation of clonal cell lines constitutively expressing GFPd2 with a single Cas12a or Csy4 DR in the 5′ UTR (or no DR as a control). (B) Generation of double-stable lines for inducible Cas RNase expression. The lines in (A) were used to generate lines that also stably express either dCas12a or Csy4 (or no RNase as a control) in a Dox-regulatable manner. (C,D) Performance of the double-stable cell lines from (B), assessed using the percentage of cells in Q2 (C) and mean GFP RFU (D) in the absence or presence of Dox (1 µg/mL). Data are shown as the mean ± SEM normalized to Dox-treated No-RNase controls; n = 2–3 per condition. (E,F) Dox dose–response experiments with double-stable cell lines inducibly expressing dCas12a or Csy4. Graphs depict responsiveness to a range of concentrations of Dox, quantified in terms of the percentage mCherry-positive cells that were also GFP-positive (E) or the mean GFP RFU in all cells (F). Mean IC₅₀ values are shown. Data are shown as the mean ± SEM for 2–3 replicates per condition, normalized to values in the absence of Dox for each line. (G) Temporal dynamics of GFPd2 expression in cell lines inducibly expressing either dCas12a or Csy4 following addition (solid lines) or withdrawal (dashed lines) of Dox (1 µg/mL). Data are shown as the mean ± SEM for 2–3 independent experiments, normalized to No-Dox controls. Mean t_1/2 values are indicated. (H,I) Relative expression of GFPd2 mRNA in the absence versus the presence of Dox (1 µg/mL) in double-stable cell lines instantiating 5′ DREDGE with Cas12a (H) and Csy4 (I). Note the dramatic increase in GFPd2 mRNA levels triggered by dCas12a expression (H). Data are mean ± SEM for 3 independent experiments, expressed as a percentage of No-Dox controls.

Figure 4

Control of an endogenous gene with 5′ DREDGE. (A) Genomic structure of the 5′ end of murine CTSD. (B) Design of the targeting construct used to introduce a single Csy4 DR into the 5′ UTR of CTSD together with a Puro^r resistance cassette within the first intron. Note the presence of two FRT sites flanking the Puro^r resistance cassette (purple). (C) Structure of the modified CTSD allele after removal of Puro^r resistance cassette with Flp-recombinase. The complete sequence of the Csy4 DR is shown (green), highlighting its location relative to the initiation codon (blue). One cell line with sequence-verified insertion of the Csy4 DR was transfected with the indicated constructs to generate individual stable lines expressing Csy4 or no RNase in a Dox-regulatable manner. (D) CatD activity in the latter stable lines in the presence of Dox (1 µg/mL) relative to no Dox. Data are shown as the mean ± SEM; n = 5–6. (E) CatD activity in stable line 1E8 subjected to repeated cycles of Dox addition and withdrawal on a weekly basis. Graphs depict CatD activity in cells initially grown in the absence (left) or presence (right) of Dox (1 µg/mL), normalized to unchanged controls. Data are shown as the mean ± SEM; n = 8 replicates per condition.