henhaobio
Sign in

Nature CommunicationsSynthetic BiologyPeer-reviewed journal article

Controlling CRISPR-Cas9 with ligand-activated and ligand-deactivated sgRNAs

Kale KundertiD, James E. LucasiD, Kyle E. WattersiD, Christof FellmanniD, Andrew H. NgiD, Benjamin Murray HeineikeiD, Christina M. FitzsimmonsiD, Benjamin L. Oakes, Jiuxin QuiD, Neha K. PrasadiD, Oren S. RosenbergiD, David F. SavageiD, Hana El‐SamadiD, Jennifer A. DoudnaiD, Tanja KortemmeiD

PDFDOIPreprint
In-depth full-text digest available · Jump to digest ↓

Abstract

The CRISPR-Cas9 system provides the ability to edit, repress, activate, or mark any gene (or DNA element) by pairing of a programmable single guide RNA (sgRNA) with a complementary sequence on the DNA target. Here we present a new method for small-molecule control of CRISPR-Cas9 function through insertion of RNA aptamers into the sgRNA. We show that CRISPR-Cas9-based gene repression (CRISPRi) can be either activated or deactivated in a dose-dependent fashion over a >10-fold dynamic range in response to two different small-molecule ligands. Since our system acts directly on each target-specific sgRNA, it enables new applications that require differential and opposing temporal control of multiple genes.

In-depth summary

Claude Sonnet 4.6 · xhigh effort · 8 Nov 2026 · from the full text of version 1 (15 May 2018) · CC BY 4.0

Inserting theophylline aptamers into the sgRNA yields ligand-activated and ligand-deactivated guides that tune CRISPRi in E. coli over 11x to 13x dynamic ranges, enabling two-ligand multiplexing.

Most Cas9 regulation switches the protein, so every target responds the same way. Aptamer-containing sgRNAs allow per-guide control, both activating and deactivating, and require only Cas9 plus the designed sgRNA.

Key results

  1. ShownAn in vitro cleavage screen of rationally designed aptamer-sgRNAs found a minority of responsive designs. Most were ligand-activated (aptamer in upper stem or hairpin), and one was ligand-deactivated (aptamer in nexus).

    86 designs tested; 10 responsive to theophylline; 9 activated and 1 (#61) deactivated. Strand-displacement linking gave the most successful designs. · Results and Discussion; Figure legends

  2. ShownFACS selection and counter-selection of randomized-linker libraries in E. coli gave ligRNA(+) and ligRNA(-), which switch CRISPRi repression with theophylline. A weaker sgRNA promoter narrowed the ranges.

    ligRNA(+) and ligRNA(-) showed 11x and 13x dynamic ranges, spanning 55% and 59% of the control range, with negligible overlap. Moving from strong J23119 to weak J23150 promoter narrowed ranges from 10.2±0.7-fold to 5.9±0.5-fold (+) and from 16.2±1.0-fold to 11.6±0.9-fold (-). · Results and Discussion; Identification of functional ligRNAs in E. coli…; Figure legends

  3. ShownBoth ligRNAs respond smoothly and nearly linearly to theophylline dose in the E. coli CRISPRi assay. Apparent EC50s are far above the free aptamer affinity, so the ligRNAs allow expression tuning, not only on/off switching.

    Apparent EC50 134.3±11.3 μM (ligRNA(+)) and 177.6±17.3 μM (ligRNA(-)) vs aptamer KD of 320 nM; fits to a two-state equilibrium model. · Results and Discussion; Figure legends

  4. SuggestedBoth ligRNAs work with most of 24 tested spacers in vitro. For ligRNA(-), higher predicted spacer-aptamer affinity correlated with a smaller ligand response, giving a design criterion. The ligRNA(+) correlation was negligible.

    Theophylline-responsive with 15 of 24 spacers (ligRNA(+)) and 21 of 24 (ligRNA(-)); ViennaRNA duplex free energy used for spacer-aptamer affinity. · Results and Discussion; Test of different spacers; Figure legends

  5. ShownSwapping in a 3-methylxanthine-specific aptamer added a second ligand input, although the theophylline aptamer recognizes both ligands. Two ligRNA(-) variants reversibly and sequentially controlled GFP and RFP in one E. coli system.

    Timecourse of activation, deactivation and reactivation of both reporters by flow cytometry. The ligRNA(+) with the 3-methylxanthine aptamer showed an undesirable ~2-fold response to theophylline. · Results and Discussion; Figure legends

Methods, data and reproducibility

Aptamers were inserted at three sgRNA sites and 86 rational designs tested by in vitro cleavage. Randomized-linker libraries then underwent FACS selection and counter-selection in an E. coli CRISPRi GFP/RFP strain, followed by ViennaRNA structure analysis.

  • Rationally designed aptamer-sgRNAs tested in an in vitro Cas9 DNA cleavage assay · in vitro (S. pyogenes Cas9) · n = 86 designs
  • E. coli MG1655 CRISPRi reporter strain (dCas9, chromosomal sfGFP and mRFP) used for library FACS screens and dose-response assays · E. coli · n = Approximately 10,000 events per measurement
  • ligRNA(+) and ligRNA(-) tested with diverse spacers in the in vitro cleavage assay · in vitro (S. pyogenes Cas9) · n = 24 spacers; 3 or 4 measurements each (n=2 for spacer #24)
  • S. cerevisiae strains with estradiol-inducible dCas9-Mxi1, fluorescent reporters and ligRNA constructs (methods only; results in supplementary figures) · S. cerevisiae
  • HEK293T-derived GFP reporter cells (HEK-RT1) with lentiviral ligRNA constructs, theophylline 0.1 mM to 10 mM (methods only) · human · n = 10,000-30,000 events per sample
Methods & tools
S. pyogenes Cas9 and dCas9 (D10A, H840A), Theophylline aptamer, 3-methylxanthine aptamer, ViennaRNA 2.4.3 (RNAfold, RNAduplex), Fiji 1.51r, BD LSRII flow cytometer, BD FACSAria II cell sorter, Gaussian kernel density estimation with BFGS mode finding, Hill equation fitting, MoClo Yeast Toolkit, Addgene 44251 and 44249 plasmids
Reagents & environment
ViennaRNA 2.4.3; Fiji 1.51r; dCas9 with D10A and H840A mutations; plasmids Addgene 44251 and 44249 for the E. coli CRISPRi strain.
Reproducibility gaps
No code or data repository URLs given; sequences are in supplementary tables. The mammalian reporter line is described in a separate work in revision.
Limitations (6)
  • Transfer to eukaryotic systems (yeast, mammalian cells) will need further optimization; ligRNAs are described as robust only in bacteria. In HEK293T reporter cells, 5 mM and 10 mM theophylline were considerably toxic. · stated by the authors
  • The mechanism of ligRNA(-) is unclear. ViennaRNA predicted the same structure with and without ligand; the U95 accessibility model rests on mutational tests and is offered as a possible explanation. · stated by the authors
  • Not all spacers worked: ligRNA(+) responded with 15 of 24 spacers and ligRNA(-) with 21 of 24. For ligRNA(-), non-functional spacers may reflect spacer-aptamer base pairing, a hypothesis assessed only by predictions; the ligRNA(+) correlation was negligible. · stated by the authors
  • The ligRNA(+) built with the 3-methylxanthine aptamer kept an undesirable but small ~2-fold response to theophylline, so the multiplexing system used the two ligRNA(-) variants instead. · stated by the authors
  • Apparent EC50s (134.3±11.3 and 177.6±17.3 μM) and the millimolar ligand doses used (1 mM in E. coli, 2.5 mM in yeast) are far above the aptamer KD of 320 nM, which may limit use in sensitive or toxicity-prone settings. · noted in this summary
  • Performance was shown on fluorescent reporter genes and in vitro cleavage. Endogenous gene targets, off-target effects and the number of replicates in the multiplexing timecourse are not reported in the text. · noted in this summary

Read this if you want to engineer RNA-aptamer-controlled sgRNAs for dose-dependent, ligand-activated or ligand-deactivated CRISPRi, or multiplexed temporal gene control, in bacteria.

Citations
189
FWCI
9.27
Field percentile
top 1%

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISBaccouche '22Pelea '25Huang '21Wei '22Zhou '20Zhang '21
71% semantic matchBioengineering
A. Baccouche, A. Adel, N. Yachie+2 more·bioRxiv·Oct 26, 2022