Epigenetics classically refers to the inheritable changes of hereditary information without perturbing DNA sequences. Understanding mechanisms of how epigenetic factors contribute to inheritable phenotype changes and cell identity will pave the way for us to understand diverse biological processes. In recent years, the emergence of CRISPR/Cas9 technology has provided us with new routes to the epigenetic field. In this review, novel epigenetic techniques utilizing the CRISPR/Cas9 system are the main contents to be discussed, including epigenome editing, temporal and spatial control of epigenetic effectors, noncoding RNA manipulation, chromatin in vivo imaging, and epigenetic element screening.
Epigenetics classically refers to the inheritable changes of hereditary information without perturbing DNA sequences. DNA methylation, demethylation, hydroxyl-methylation, histone modification, chromatin remodeling, gene imprinting, and noncoding RNA are the central mechanisms involved. They play important roles in diverse biological processes including gene regulation, iPSC reprogramming and maintenance, genomic imprinting, X-chromosome inactivation, aging, neurodegeneration, autoimmune modulation, and tumorigenesis [
Adapted from a natural immune defense system in bacteria, the clustered regularly interspaced short palindromic repeat- (CRISPR-) associated protein 9 (Cas9) system, abbreviated as the CRISPR/Cas9 system, is a site-specific genome editing tool that could be implemented to target and mutate specific genomic regions in eukaryotic cells, especially in mammalian cells [
Notably, since the emergence of CRISPR/Cas9 technology, diverse applications have been explored beyond genome editing. Here, we will focus on the new toolkit that CRISPR/Cas9 has provided to us for epigenetic research.
Epigenome editing refers to the targeted rewriting of epigenetic markers [
To achieve CRISPR/Cas9-mediated epigenome editing, the main strategy is fusing the Cas9 protein with a transcription repressor or activator domain, which was known as an epigenetic effector (epieffector) [
For example, when the fused epigenetic effector domain was Krüppel-associated box (KRAB), using the dCas9-KRAB complex to target and induce locus-specific deposition of H3K9me3 at the
When the fused domain was LSD1, using the dCas9-LSD1 complex to target
When the fused domain was p300 core, which was an acetyltransferase, using this dCas9-p300 core complex, researchers activated the
The fusion domain could also be the transactivation domain VP64 or VPR, the DNA methyltransferase 3A (DNMT3A), or the DNA demethylase TET. The dCas9-VPR complex is an escalated version of dCas9-VP64, where VP64 was further fused by two other transcription factors p65 and Rta to increase the transactivation efficiency. Both of them have well been demonstrated to be reliable gene activation tools. The dCas9-VP64 complex could directly activate the silenced
Efforts have been made to improve epigenome editing efficiency. The first strategy is incorporating SunTag into the dCas9-epieffector complex. SunTag is a repeating peptide array that can simultaneously bind with multiple copies of a certain protein. Several studies have highlighted the utility of the dCas9-SunTag system in improving transactivation robustness. For example, via recruiting multiple transactivation domains, the dCas9-SunTag complex has been proved to be able to augment a transactivation effect significantly compared to the conventional dCas9-single activator domain complex [
Briefly, the dCas9-epieffector complex could achieve methylation and demethylation at DNA level, rewriting histone marks by inducing methylation or acetylation at nucleosome level, and be optimized to improve the editing efficiency (Figures
DNA methylation and demethylation mediated by Cas9. DNA sequences are represented by the blue lines. PAM sequences are highlighted by red color. Guide RNA sequences are shown in purple color. DNA polymerase II (PoI II) is represented by the green irregular shape. The circled “me” represents methylation at a specific CpG site. Basically, dCas9 functions as a DNA-binding protein. DNMT3A or TET is the epieffector. The dCas9-epieffector complex is guided to the DNA target by guide RNA via Watson-Crick base pairing to execute (a) DNA methylation or (b) demethylation, thus inducing decreased or increased gene expression.
Histone acetylation and methylation mediated by Cas9. The empty circle attached to the nucleosome represents a specific amino acid of the histone side chain. The circled “acetyl” and “me” represents acetylation and methylation of amino acids, respectively. The dCas9-epieffector complex could be guided to a selected DNA target to achieve (a) acetylation or (b) methylation at histone level to regulate gene expression.
Validated representative dCas9-epieffector complex.
Type | Mechanism | Function | Reference |
---|---|---|---|
dCas9-KRAB | Histone methylation | Gene downregulation | [ |
dCas9-DNMT3A | DNA methylation | ||
dCas9-LSD1 | Histone acetylation | Gene upregulation | [ |
dCas9-p300 | Histone acetylation | ||
dCas9-TET | DNA demethylation | ||
dCas9-VP64 or VPR | Recruitment of active transcription factors | ||
dCas9-SunTag-epieffector | Recruitment of multiple epieffector domains by the repeating peptide array SunTag | Effect augmentation | [ |
dCas9-SAM-epieffector | Recruitment of multiple epieffector domains by the protruding guide RNA loops |
In a broad sense, epigenetics refers to the temporal and spatial control of gene expression, of which the total effect would determine cell phenotypes during development, aging, and disease pathogenesis [
Studies prior to the emergence of CRISPR have proved the utility of optogenetics in temporal control of gene expression. Basically, there need to be two components, a DNA-binding protein fusing to a light-sensitive cryptochrome protein, and a chromatin modifier fusing to a cryptochrome protein interaction partner, vice versa. [
By the same principle, the dCas9 DNA-binding domain paired with optical-inducible proteins could also be utilized to recruit the epieffector domain to the targeted DNA site in an inducible and reversible manner described above.
In a light-activated CRISPR/Cas9 effector (LACE) system, CRY2 and CIB1 were fused to transactivation domain VP64 and catalytically deactivated Cas9, respectively. Cotransfection of these fusion protein pairs with guide RNA resulted in detectable levels of gene activation in the presence of blue light. Interestingly, when the N-terminal fragment of CIB1 was fused to both N- and C-terminus of dCas9, gene activation level in response to blue light was significantly increased, which is consistent with previous observation that simultaneous recruitment of VP64 domains to the target site had a synergistic effect on gene activation [
Another similar strategy from a different angle is to split Cas9 into fragments and further tether them with optical inducible protein pairs. Upon light stimulation, Cas9 fragments would be brought together via light-induced dimerization to reconstruct nuclease activity. Interestingly, in a photoactivatable CRISPR/Cas9 system, when split Cas9 fragments were fused with the CRY2-CIB1 pair, no light-induced Cas9 activity was induced. Given the steric block effect may be at play, a different pair of proteins with much smaller molecular weight called magnets was used. The nuclease activity was successfully reconstructed upon light stimulation when Cas9 fragments were tethered with positive and negative magnets, respectively [
As to the chemical-controlled strategy, the optimized inducible gene knockdown or knockout (OptiKO) system was a good example. It was a strategy harnessing the CRISPR/Cas9 technology and tetracycline-inducible expression cassette in human iPSC. The OptiKO system has two main components. Cas9 was constitutively expressed while guide RNA (gRNA) was tetracycline-inducible. Clonal lines incorporating both the CAG-Cas9 cassette and the TET-ON/CAG-gRNA cassette was first isolated. In the following experiments, the addition of tetracycline to the isolated clonal cells at different time points could achieve targeted region knockdown or knockout flexibly. In particular, inducible knockdown of the
Although more and more evidence is being reported regarding the feasibility of optical or chemical controllable epigenome editing, the detailed mechanisms remain to be elusive. Whether the local chromosomal or nucleosome context affects the binding affinity of the DNA-binding protein remains to be determined. What are the exact effects of locus-specific epigenome editing on DNA epigenetic profiling, histone modifications, and chromosomal interactions? What are the transcription factors participating in transcription activation or repression mediated by epigenome editing? Do we have the confidence to ensure every shot with unfailing accuracy? Such questions need to be answered to achieve reliable epigenome editing consequence and bona fide conclusions.
Elusive as the mechanism is, studies discussed above indeed enlightened us with new possible routes to perturb epigenome precisely. It is conceivable that through temporal and spatial control of genes playing indispensable roles in epigenetic dynamic changes, we may acquire new insights into the underlying mechanisms (Figure
Schematic of temporal and spatial control of epigenome editing. The semicircles labeled with (A) and (B) represent a protein pair. The magenta comb-like lines represent guide RNA sequences. (A) is bound to dCas9, which would be directed by the guide RNA to the DNA target. (B) is bound to the epieffector. (a) Upon stimulation of light or chemicals, (A and B) would pair with each other thus bringing the dCas9 and epieffector together to achieve site-specific epigenome editing at a given time point. (b) Another strategy is splitting the dCas9 into two parts, each of which is bound by protein (A) or (B). Upon light or chemical stimulation, (A) and (B) would gather together to reconstruct an intact dCas9-epieffector complex to achieve site-specific epigenome editing at a given time point.
Noncoding RNA plays important roles in gene imprinting and chromatin remodeling; thus, it is an indispensable topic in the epigenetic research field [
Most recently, Feng Zhang Lab developed a RNA editing tool, named as the RNA Editing for Programmable A to I Replacement (REPAIR) system. It was adapted from the type IV CRISPR/Cas13 system, where Cas13b was deactivated for its RNA cleavage capacity while the RNA-binding ability was reserved. Deactivated Cas13b was further fused with an adenosine deaminase enzyme called ADAR. After being directed to the targeted RNA transcript, the Cas13b-ADAR complex would deaminize adenosine to inosine, which is an analog to guanine in diverse biological processes, resulting in A to I conversion. Although this CRISPR/Cas13-based RNA editing tool seemed to have no overwhelming advantage over the CRISPR/Cas9-based DNA editing method in terms of specificity due to a substantial number of off-target events across the transcriptome, it still presented some promising merits. For example, no PAM sequence constraints within the RNA target meant more flexible target binding capacity of Cas13b compared to the classical Cas9. Direct conversion of RNA base without reliance on endogenous repair mechanism also rendered it applicable in mitotic cells such as neurons. Additionally, the transient nature of RNA editing may provide a basis for the temporal control of targeted RNA transcription and downstream translation [
Altogether, it seems that Cas9 instead of dCas9 is playing important roles in noncoding RNA manipulation, as more and more Cas proteins are being characterized, such as cpf1 [
DNA wraps around histones to form nucleosomes. Nucleosomes continue to assemble in a spiral pattern. Through several rounds of such repeated twisting, DNA and protein are finally packaged into the superhelix structure, chromatin. Chromatin has two statuses, the tightly packed heterochromatin and the loosely packed euchromatin [
Live cell imaging is a valuable method for studying structure changes and interactions of chromatin in epigenetic regulation. Efforts have been made to adapt the CRISPR/Cas9 system for the in vivo imaging experiments. Generally, the design rationale is tagging the guide RNA loops or the dCas9 protein with fluorescent dyes to form a dCas9-gRNA-fluorescent dye complex. In a recent study, a dCas9 expression cassette was first incorporated into cells of interest. Guide RNA scaffold was redesigned, with two new stem loops protruded. One of the loops was tethered by protein MS2; the other was bound by protein PP7, which were able to recruit EGFP-conjugated MCP and mCherry-conjugated PCP, respectively. Clonal cells expressing dCas9, EGFP, and mCherry were selected as the tested platform. The delivery of guide RNAs into these cells would bring dCas9 and fluorescent proteins together at targeted loci. Colocalization and separation of the two-color fluorescent signals indicated possible interactions between the targeted loci on chromosomal level. Using this dual-color system, researchers achieved real-time labeling of major satellites, minor satellites, and two single loci on chromosomes [
In vivo imaging of chromatin interaction mediated by Cas9. Guide RNAs direct dCas9 and fluorescent proteins to bind with selected DNA targets, forming the dCas9-gRNA-fluorescent dye complex. At chromosomal level, spatial position changes between chromosomal regions could be reflected by the (a) seperation or (b) interaction of the dual-color fluorescent signals.
In addition to live cell imaging, CRISPR/Cas9 could also be used to characterize locus-specific regulatory composition. In a recent study, the deactivated cas9 nuclease was biotinylated to capture telomeric factors and components of enhancers regulating human
Chromatin structures and interactions are gaining increasing academic attention in recent years. It has been widely accepted that the dynamic chromatin changes play pivotal roles in the temporal-spatial regulation of genes. The in vivo imaging technique and several other attempts described above indicated that CRISPR/Cas9 is a promising research tool for this topic.
The accumulation of work has demonstrated the capacity of CRISPR/Cas9 for screening both coding and noncoding DNA sequences on a genome-wide scale [
By transcriptionally activating or deactivating a specific gene that is normally silent or active, epigenome editing exhibited therapeutic potentials. It is easy to imagine that diseases with aberrant epigenetic marks as the underlying pathophysiologic mechanisms would be excellent targets to test epigenome editing-based therapeutics. The disease spectrum may include cancers, neurodegenerative diseases, neuropsychiatric diseases, neurodevelopmental diseases, and imprinting disorders. For example, using a dCas9-SAM-VPR complex, researchers reactivated a heavily methylated tumor suppressor gene,
In addition, epigenome editing has the potential to facilitate stem cell therapy. In recent years, iPSCs have been a major source for cell replacement therapy due to its accessibility and pluripotency [
Taken together, these studies demonstrated the therapeutic potential of CRISPR/Cas9-mediated epigenome editing in various diseases and stem cell therapy.
In conclusion, all of the studies described above indicated that CRISPR/Cas9-mediated epigenome editing holds a great promise for epigenetic studies and therapeutics. However, there are still some limitations to be scrutinized. First of all, in terms of basic science studies, although most studies claimed high specificity in their experiments, however, the high specificity usually is the result of repeated optimization. A precise model that could predict deleterious off-target effects during the experiment design stage is still lacking. In addition, although transactivation or repression effects on multiple genes were well documented in publications, mechanisms underlying the phenomenon were not clear. Epigenetic mark profiling on epigenome scale was not sufficient. Local CHIP-seq data usually only focused on the characterization of one or few histone marks. Theoretically, we hope that epigenome editing could achieve targeted gene regulation by changing epigenetic marks specifically and freely according to our wills. To achieve this goal, high specificity and clarified mechanisms are the prerequisite. Therefore, more thorough off-target event assessments and more studies focusing on mechanisms underlying epigenome editing are needed.
Moreover, in terms of clinical applications, several issues need to be addressed prior to successful clinical translation. Firstly, the endurance of gene activation or a repression effect mediated by CRISPR/Cas9 remains to be undetermined. It has been thought that epigenome-editing-induced gene activation or repression is short-term [
The authors have declared that there is no conflict of interest.
This work was supported by the National Key Plan for Scientific Research and Development of China (2016YFC1306000) and the National Natural Science Foundation of China (81430023, 91132000, and 81171198).