Repurposing type I–F CRISPR–Cas system as a transcriptional activation tool in human cells

Class 2 CRISPR–Cas proteins have been widely developed as genome editing and transcriptional regulating tools. Class 1 type I CRISPR–Cas constitutes ~60% of all the CRISPR–Cas systems.

However, only type I–B and I–E systems have been used to control mammalian gene expression and for genome editing. Here we demonstrate the feasibility of using type I–F system to regulate

human gene expression. By fusing transcription activation domain to Pseudomonas aeruginosa type I–F Cas proteins, we activate gene transcription in human cells. In most cases, type I–F

system is more efficient than other CRISPR-based systems. Transcription activation is enhanced by elongating the crRNA. In addition, we achieve multiplexed gene activation with a crRNA

array. Furthermore, type I–F system activates target genes specifically without off-target transcription activation. These data demonstrate the robustness and programmability of type I–F

CRISPR–Cas in human cells.

Clustered regularly interspaced short palindromic repeats (CRISPR) and CRISPR-associated (cas) genes-based defence systems protect bacteria and archaea against phage and other foreign

genetic elements1,2,3. Since the identification of increasing number of cas genes, the CRISPR–Cas systems have been classified into two Classes (Class 1 and Class 2) and six types (Type

I–VI)4 based on the different arrangements of cas genes and the subunits of effector complexes5,6,7. Class 2 CRISPR–Cas systems, the best-studied system with single effector protein (e.g.,

Cas9, Cas12, or Cas13) for foreign DNA or RNA interference, are subdivided into Type II (Cas9), Type V (Cas12), and Type VI (Cas13). In the past few years, Class 2 CRISPR–Cas systems have

revolutionized both basic and clinical researches, enabling more rapid, precise, and robust genome editing and modifications in cultured cells and animals8,9,10,11,12,13,14,15,16,17.

However, there were only a few applications of Class 1 CRISPR–Cas (Type I, Type III and Type IV) system.

Class 1 type I CRISPR–Cas systems are the most prevalent (~60%) in both bacteria and archaea, whereas class 2 only makes up ~10% of all CRISPR–Cas systems18,19. Differing from the Class 2

CRISPR–Cas systems, the Class 1 type I system relies on Cascade (CRISPR-associated complex for antiviral defense complex) for DNA binding, which further recruits Cas3 to degrade the foreign

DNA20. Cascade, which recognizes and binds specific DNA, is a complex consist of multiple Cas proteins and CRISPR RNA (crRNA). CRISPR–Cas expression involves cas genes expression and CRISPR

transcription, yielding a precursor crRNA (pre-crRNA). The pre-crRNA is processed at the repeat regions by Cse33, Cas621 or Csy422 to generate mature crRNA with different characteristics.

Other Cas proteins then bind onto the crRNA and assemble into a functional Cascade23,24,25,26. Cascade discriminates the self and non-self DNAs by recognizing the PAM (proto-spacer adjacent

motif) sequence27, which triggers a conformational change upon binding28,29. The conformational change finally recruits Cas3 for invasive DNA degradation20,30,31,32.

Compared to the widely used class 2 CRISPR–Cas systems, the multiple-subunit class 1 type I CRISPR–Cas system has distinct properties, for example, generating large fragment deletion in

genome editing with Cas333,34, and multiple subunits for different Cas protein–effector fusion strategies35. These differences between the class 1 and class 2 CRISPR–Cas system may

contribute to the advantages of Class 1 CRISPR–Cas system in some applications. Accroding to recent classification studies, there are seven subtypes (I–A to I–G) in type I CRISPR–Cas

system7,36. In recent years, the type I–A37, I–B38,39, I–E40, and I–F41,42 CRISPR–Cas have been used for prokaryotic gene engineering in Sulfolobus islandicus (I–A), Clostridium pasteurianum

(I–B), Lactobacillus crispatus (I–E), Zymomonas mobilis (I–F), and Pseudomonas aeruginosa (I–F). Besides, type I–B43 and type I–E44,45,46 Cascades can work as transcription repressor in

Sulfolobus islandicus (I–B) and Escherichia coli (I–E). Furthermore, type I–E and I–B CRISPR–Cas systems have been used in human cells33,34,35,47 and plants48 for gene editing and

transcription regulation. Therefore, developing tools based on type I CRISPR–Cas system might provide alternative tools for genome editing and gene regulation.

Type I–F CRISPR–Cas system is among the well-studied CRISPR–Cas systems. It has fewer Cascade components than type I–E CRISPR–Cas system (4 vs 5), which will be easier to be controlled and

delivered. The type I–F CRISPR–Cas system was first discovered as CRISPR subtype Ypest from Yersinia pestis49,50. The Cascade components of type I–F CRISPR–Cas system were also named as Csy

(CRISPR subtype Ypest) subunits, which includes Csy1 (Cas8f1), Csy2 (Cas5f1), Csy3 (Cas7f1), and Csy4 (Cas6f)7,26 (Fig. 1a). In addition, the Cascade of type I–F variant (type I–Fv, or type

I–F2) CRISPR–Cas system, derived from type I–F system, consists of only three subunits: Cas5fv (Cas5f2), Cas6f, and Cas7fv (Cas7f2)4,7 (Fig. 1a). The type I–F and type I–Fv Cascade

recognizes 5′-CC PAM on the non-target strand for target binding51,52. Their crRNAs consist of 8-nt 5′ handle for Csy1 and Csy2 binding, 32-nt spacers bound by six copies of Csy3 for target

recognition, and 20-nt 3′ hairpin for Csy4 binding and pre-crRNA processing22. Recently, type I–F CRISPR–Cas system has been used for genome engineering in Zymomonas mobilis41 and

Pseudomonas aeruginosa42. However, there has not been any report on the exploitation of the type I–F or type I–Fv CRISPR–Cas system for genome manipulation application in human cells yet.

a Schematic diagram of the Pseudomonas aeruginosa type I–F and Shewanella putrefaciens type I–Fv CRISPR–Cas locus. Cas proteins are presented with arrows in different colors. CRISPR repeats

are indicated with gray diamonds. b Electrophoresis mobility shift assays to detect target DNA binding by PaeCascade. Up, schematic representation of the processed crRNA with 5′-CC-3′ PAM

recognition and base pairing at the DNA target site. Down, the result of the EMSA assay. The arrow indicates PaeCascade–crRNA–DNA complex. “*” Indicates free ssDNA. c Electrophoresis

mobility shift assays to detect target DNA binding by SpuCascade. Up, schematic representation of the processed crRNA with 5′-CC-3′ PAM recognition and base pairing at the DNA target site.

Down, the result of the EMSA assay. The arrow indicates the SpuCascade–crRNA–DNA complex. “*” Indicates free ssDNA. d A schematic of the integrated sequence in the TRE-eGFP reporter cell.

The target sequences of type I–F and type I–Fv CRISPR–Cas system containing a 5′-CC-3′ PAM is shown. PAM is in red, and the target sequence is in blue. rtTA: reverse

tetracyclin-transactivator. TRE: tetracyclin response element. e Flow cytometric analysis of GFP activation in TRE-eGFP reporter cells transfected with type I–F PaeCascade all-in-one vectors

and crRNA expression vectors. Left: the all-in-one constructs used in the experiment. PaeCascade subunits linked by self-cleaving P2A peptides was driven by PGK promoter. Right:

Quantification of GFP positive cell induced by type I–F PaeCascade all-in-one vectors. f Flow cytometric analysis of GFP activation in TRE-eGFP reporter cells transfected with type I–F

SpuCascade all-in-one vectors and crRNA expression vectors. g Flow cytometric analysis of GFP activation in TRE-eGFP reporter cells transfected with type I–F PaeCascade 2-vector systems and

crRNA expression vectors. Left: 2-vector systems used in the experiments. Right: quantification of GFP positive cells induced by type I–F PaeCascade 2-vector systems. Ctrl: untransfected

control. Data represented three biological repeats and displayed as mean ± S.E.M. Statistical significance was calculated using one-way ANOVA (n.s., not significant; *P