The histone chaperone FACT induces Cas9 multi-turnover behavior and modifies genome manipulation in human cells

Introduction

There are few in the scientific, and indeed the non-scientific, community who have not heard of the recent genome engineering revolution that is CRISPR-Cas9. This technology has transformed our ability to precisely manipulate, alter and recode genes in the laboratory and has hit mainstream headlines with its promises of miracle cures for genetic disease as much as for its controversies.

The technique centres around a bacterial nuclease called Cas9, which is directed to cleave specific DNA target sequences by a guide RNA (gRNA) molecule. Cleaved sites are then naturally repaired by the cell by either non-homologous end-joining (NHEJ) or homology directed repair (HDR). As repair is error-prone this usually leads to base pair insertion or deletion mutations (indels) which ‘knockout’ genes, however HDR can also be exploited to introduce exogenous sequences into the genome by providing a single stranded oligodeoxynucleotide (ssODN) to template the repair. Catalytically dead Cas9 (dCas9) fused to nuclear effectors can also be used to target other proteins to the DNA, for example transcriptional repressors (CRISPRi).

The usefulness of Cas9 as a research tool is undeniable – a quick PubMed search for CRISPR Cas9 brings up almost 10 000 publications since 2013 when its technological potential was first reported^1,2,3. However, as anyone who has ever tried this technique will tell you, it is not always as simple to use as you might expect. Engineering DNA by this method, although precise, is not always efficient and success in vitro does not always translate to success in vivo. The reasons for this are only partially understood.

Although the situation is improving, we have very little understanding of how the bacterial Cas9 behaves within the busy environment of the Eukaryotic nucleus. Chromatin structure, enzymes, and DNA post-translational modifications are all likely to influence the accessibility of DNA to the enzyme and the subsequent dynamics of this interaction. It is therefore perhaps not surprising that different DNA regions are more amenable to manipulation than others, or that there are differences between model systems. In order to improve the precision and efficiency with which we can edit genes we must better understand the interplay between Cas9 and other nuclear residents.

The prelight presented here addresses precisely this problem. Using an unbiased proteomics approach, Wang et al discover that the Facilitates Chromatin Transcription (FACT) histone chaperone complex impacts on the outcome of Cas9 activity by determining its residence time on DNA.

Key Results

Using competition assays between dCas9 (preloaded onto the templates but unable to cleave) and active Cas9 (can bind and cleave if dCas9 dissociates), Wang et al demonstrate that a cellular factor present in Xenopus eggs is required for the exchange of dCas9 for Cas9 in vitro. To identify this interactor, they fuse Cas9 to the biotin ligase BirA^* to biotinylate neighbouring proteins for extraction with streptavidin. Three sets of DNA bound Cas9 interactors were discovered by this approach: PIP4K2C; H/ACA-associated proteins DKC1, NHP2, 201 NOP10, and GAR1; and both components of the FACT heterodimer, SPT16 and SSRP1. As the FACT complex is a histone chaperone with roles in nucleosome remodelling it seemed the most likely factor promoting Cas9-DNA dissociation and so became the focus of the study.

To determine whether FACT could influence Cas9 behaviour in vitro, the dCas9/active Cas9 competition assays were repeated following SPT16 or SSRP1 depletion. In the absence of FACT, no DNA cleavage was observed indicating dCas9 turnover was inhibited. The authors then moved on to look in human cells. K562 cells were depleted of SPT16 using siRNA in the presence or absence of an ssODN donor for HDR. In the absence of ssODN there was little change in indel rates after 48 hours of transfection, however in the presence of ssODN, levels of HDR dropped by 50% in SPT16 depleted cells.

The effect of FACT levels on epigenetic marking and transcriptional activation by dCas9 directed modifiers was also tested. In cells expressing dCas9-p300 which directs histone acetyltransferase p300 to target sites, depletion of FACT caused a significant increase in the levels of acetylation specifically at that site. Cas9-effector induced methylation rates were also increased following depletion of FACT in cells, which translated into an increased amount of transcriptional repression.

Conclusions and perspective

The overall conclusion of this manuscript is that depletion of FACT can potentiate CRISPRi and inhibit HDR in cells. As the authors discuss, the mechanism of HDR inhibition is not immediately obvious – multiple rounds of cleavage in the presence of FACT may create more opportunities for HDR to take place, or FACT could be acting indirectly to recruit repair factors and remodel nucleosomes. The same is true for FACT enhancement of CRISPRi, however the direct effect of increasing DNA modifications by increasing Cas9 residence time at the target site is the more compelling model in this instance. Further work is needed to fully understand this process. Nonetheless, I chose this preprint because of the take home message – the outcome of CRISPR genome editing is dependent on the nuclear environment. Although this has been shown to some extent in other ways, continuing to attain such knowledge can help develop combined strategies and treatments that allow better control of outcome and increased efficiencies.

Questions for authors

Can overexpression of FACT push CRISPR editing/CRISPRi efficiencies in the opposite direction, for example could this help promote homology directed repair and be used to improve the efficiency of CRISPR knock-ins?

Is it known how FACT levels vary between, for example, different cell types or developmental stages in vivo and can this information be used to target CRISPR strategies more effectively?

Is it possible to probe the contribution of different domains of SPT16 and SSRP1 to find the region relevant to Cas9 function and determine the mechanism by which FACT influences HDR rates. For example, is DNA binding alone sufficient or do you indeed need binding sites for repair factors, nucleosome remodelling function, etc?

What effect do the other identified interactors have on Cas9 behaviour?

References

1. Cong et al. 2013. Multiplex genome engineering using CRISPR/Cas systems. Science 339(6121):819
2. Mali et al. 2013. RNA guided human genome engineering via Cas9. Science 339(6121):823
3. Jinek et al. 2013. RNA-programmed genome editing in human cells. eLife 29:2

Tags: crispr, dna, fact, genome engineering

doi: https://doi.org/10.1242/prelights.13520

Read preprint (No Ratings Yet)