RNA-activated protein cleavage with a CRISPR-associated endopeptidase

Abstract

In prokaryotes, CRISPR-Cas systems provide adaptive immune responses against foreign genetic elements through RNA-guided nuclease activity. Recently, additional genes with non-nuclease functions have been found in genetic association with CRISPR systems, suggesting that there may be other RNA-guided non-nucleolytic enzymes. One such gene from Desulfonema ishimotonii encodes the TPR-CHAT protease Csx29, which is associated with the CRISPR effector Cas7-11. Here, we demonstrate that this CRISPR-associated protease (CASP) exhibits programmable RNA-activated endopeptidase activity against a sigma factor inhibitor to regulate a transcriptional response. Cryo-electron microscopy of an active and substrate-bound CASP complex reveals an allosteric activation mechanism that reorganizes Csx29 catalytic residues upon target RNA binding. This work reveals an RNA-guided function in nature that can be leveraged for RNA-sensing applications in vitro and in human cells.

Fig. 1.. The type III-E CRISPR-associated protease Csx29 cleaves Csx30.

(A) Schematic of selected CRISPR-associated protease (CASP) loci and three additional conserved genes in type III-E loci. (B) Immunoblot analysis of in vitro reactions with Cas7–11-Csx29 and HA-tagged Csx30, Csx31, and CASP-σ produced by cell-free transcription-translation. (C) A Cas7–11-Csx29-crRNA complex cleaves Csx30 protein in response to target RNA. (D) Csx30 cleavage requires target RNA and the Csx29 protease catalytic residues, but not the catalytic residues of Cas7–11. (E) Schematic of Csx30 highlighting the cleavage site (aa 427–429), linker (aa 377–406), and a potential effector domain annotated from HHpred (aa 452–545). (F) AlphaFold2 prediction of Csx30 (G) Analysis of dCas7–11-Csx29 proteolytic activity on truncated Csx30 proteins. (H) Immunoblot analysis of HA-tagged Csx30 mutants produced by cell free transcription-translation. Panels C, D, and G are SDS-PAGE gels stained with Coomassie.

Fig. 2.. Allosteric activation of Csx29 upon RNA binding.

(A) Schematic of Cas7–11, Csx29, and Csx30 proteins domains, and the crRNA and target RNA used in structural studies. (B) Structures of the inactive (Cas7–11-Csx29-crRNA) and active (Cas7–11-Csx29-crRNA-target RNA-Csx30) CASP complexes. (C) Structural organization of the Csx29 AR in inactive and active CASP complexes. (D) Electrostatic and hydrogen bonded network within the Csx29 catalytic site in the inactive state. (E and F) Catalytic H615 and C658 residues in inactive and active Csx29 shown with EM density. (G) Contacts between Cas7–11 and the DR-mismatched portion of the target RNA in the active state. (H) Electrostatic and hydrogen bonded network extending from the AR to the Csx29 catalytic site in the active state. (I) Mutations disrupting allosteric activation residues impair Csx30 cleavage by Cas7–11-Csx29. SDS-PAGE gel stained with Coomassie.

Fig. 3.. Csx30 substrate recognition by Csx29.

(A) Csx29-Csx30 interface in the active CASP structure. Electrostatic interactions and hydrogen bonds are drawn as dashed lines, and the hydrophobic pocket as a dashed oval. (B) Close-up of the Csx29-Csx30 interface near the catalytic H615 and C658 residues.

Fig. 4.. Csx30 binds and inhibits the transcription factor CASP-σ.

(A) Schematic of Csx30 and CASP-σ proteins. (B) AlphaFold2 prediction of a Csx30-CASP-σ interaction. (C) Purification of a Csx30-CASP-σ complex that is cleaved by dCas7–11-Csx29. SDS-PAGE gel stained with Coomassie. (D) Representative CASP-σ ChIP-seq peaks in E. coli with a 1 kb window, input coverage shown in gray. (E) Identification of a CASP-σ binding motif from ChIP-seq peaks. (F) Enrichment of CASP-σ at four E. coli peaks by ChIP-qPCR. n = 3 replicates. (G) Predicted CASP-σ targets in the D. ishimotonii CASP locus. (H) Schematic of a CASP-σ transcriptional reporter assay. (I) CASP-σ-mediated transcriptional activity in E. coli. GFP expression was normalized to cells with a scrambled promoter sequence. n = 3 replicates. ** denotes p < 0.01, Student’s t-test. Error bars represent standard deviation from the mean in all panels.

Fig. 5.. RNA sensing applications and a proposed model for CASP systems.

(A) Schematic of in vitro RNA detection using CASP systems and fluorescent Csx30 reporters. (B) In vitro detection of RNA as measured by released fluorescence. n = 3 replicates. (C) Immunoblot analysis of HA-tagged Csx30 in HEK293T human cells transfected with DiCASP components. (D) Schematic of engineered proteins containing a cell membrane tether, a Csx30 linker, and an effector domain. (E) Flow cytometry of DiCASP activity in mouse Neuro2A loxP:GFP cells using a Chrm3-Csx30_250–565⁻Cre reporter. n = 3–6 replicates. (F) Model for a three-pronged strategy of CASP systems in the defense against foreign genetic elements including Cas7–11 mediated RNA endonuclease activity, a Csx30-regulated CASP-σ transcriptional response, and a possible third arm involving Csx31. Error bars represent standard deviation from the mean in all panels.