Abstract
Background CRISPR-Cas9 has been developed as a therapeutic agent for various infectious and genetic diseases. In many clinically relevant applications, constitutively active CRISPR-Cas9 is delivered into human cells without a temporal control system. Excessive and prolonged expression CRISPR-Cas9 can lead to elevated off-target cleavage. The need for modulating CRISPR-Cas9 activity over the dimensions of time and dose has created the demand of developing CRISPR-Cas off-switches. Protein and small molecule-based CRISPR-Cas inhibitors have been reported in previous studies.
Results We report the discovery of Cas9-inhibiting peptides from inoviridae bacteriophages. These peptides, derived from the periplasmic domain of phage major coat protein G8P (G8PPD), can inhibit the in vitro activity of Streptococcus pyogenes Cas9 (SpyCas9) proteins in an allosteric manner. Ectopic expression of full-length G8P (G8PFL) or G8PPD in human cells can inactivate the genome-editing activity of SpyCas9 with minimum alterations of the mutation patterns. Furthermore, unlike the anti-CRISPR protein AcrII4A that completely abolishes the cellular activity of CRISPR-Cas9, G8P co-transfection can reduce the off-target activity of co-transfected SpyCas9 while retaining its on-target activity.
Conclusion G8Ps discovered in the current study represent the first anti-CRISPR peptides that can allosterically inactivate CRISPR-Cas9. This finding may provide insights into developing next-generation CRISPR-Cas inhibitors for precision genome engineering.
Background
Clustered Regularly-Interspaced Short Palindromic Repeats (CRISPR) is the bacterial adaptive immune system to defend bacteriophage infections [1–3]. During infection, invader DNA is captured and integrated into bacterial genome as CRISPR array. Sequences from CRISPR array are transcribed and processed into CRISPR RNAs (crRNAs), which direct CRISPR-associated (Cas) proteins to foreign nucleic acids [1, 2]. Type II CRISPR-Cas systems function with streamlined components comprising of a single nuclease protein such as Cas9 [4]. The modular and programmable features make CRISPR-Cas9 one of the most widely used tools for genome engineering applications [5–8]. However, CRISPR-Cas9 is associated with off-target cleavage [9], chromosomal rearrangement [10] and genotoxicity [11]. These side effects mainly arise from the excessive or prolonged expression of CRISPR-Cas9 [12–14]. As a therapeutic agent, CRISPR-Cas9 is often constitutively expressed in host cells [15], making the elevated off-activity a major safety concern. Temporal control of SpyCas9 activity have been investigated as an approach to improving its specificity in human cells. Technologies enabling the temporal control of CRISPR-Cas9 include optogenetic tools, intein splicing system, small molecule inducers or inhibitors [16–20] and anti-CRISPR proteins (Acrs) [21, 22].
The most investigated CRISPR-Cas inhibitors are the naturally occurring, phage-derived Acrs. Bacteriophages can use Acrs to antagonize the CRISPR-Cas immunity in bacteria [23, 24]. A number of Acrs have been identified for type I [23, 25–28], type II [29–32] and type V [28, 33] CRISPR-Cas systems. Acrs can be adapted to regulate CRISPR-Cas activities in bacteria [34], yeast [35] and mammalian cells [29, 31, 34, 36–38]. Biosensor [39] and synthetic circuits [40] can be devised based on Acr-coupled CRISPR-Cas systems. Moreover, Acrs can be harnessed to enable the temperature-responsive [41] and optogenetic [42] control of CRISPR-Cas activity. Importantly, Acrs can enhance the editing [38] and cell-type [43] specificities and reduce the cytotoxicity [11] of CRISPR-Cas-mediated genome editing. It has also been reported that Acrs can facilitate the production of CRISPR-carrying viral vectors by restricting CRISPR self-cleavage [44].
Currently known Acrs inhibit CRISPR-Cas systems by interfering with Cas protein-mediated DNA surveillance or cleavage [21]. For instance, AcrIIA4 mimics double-stranded DNA (dsDNA) and occupies the protospacer adjacent motif (PAM) recognition site of SpyCas9, thereby preventing Cas9 protein from binding to the target DNA [38, 45, 46]. Using a different mechanism, AcrIIC3 perturbs DNA binding by inducing the dimerization of Cas9 protein [36]. An alternative Cas inactivation strategy by Acrs is to interact with DNA-bound Cas proteins and block subsequent DNA cleavage, as seen with AcrF3 [47, 48] and AcrIIC1 [36]. In addition, some Acrs can function as acetyltransferase and inactivate CRISPR-Cas activity by post-translational modifications [49]. Different Acrs may inactivate CRISPR-Cas via identical mechanisms while possessing low sequence similarities [21]. The Acrs characterized to date share no common sequence motifs except for a putative transcriptional element referred to as anti-CRISPR associated genes (Acas) that are commonly found downstream of the Acr genes in the bacteriophage genome [26]. The poorly understood sequence-structure-activity relationship largely hampers the systemic discovery of novel Acrs. In addition to protein-based inhibitors, small-molecule CRISPR-Cas inhibitors have been developed. However, relative high concentrations of 10 μM or above are required for small molecules to achieve efficient inhibition.
Along with small molecules and proteins, peptides represent an alternative class of CRISPR-inhibiting agents with distinct biochemical features. In this study, we report the discovery of Cas9-inactivating peptides from inoviridae bacteriophages. In an attempt to develop anti-CRISPR antibodies using well established phage display technology, we are surprised to find that the commonly used laboratory bacteriophage strain M13 served as a source of Cas9-inactivating agent. Subsequent analyses showed that the periplasmic domain of the major coat protein G8P (G8PPD) from several inoviridae bacteriophages, which contain M13 phage, inhibited the in vitro and in vivo activity of SpyCas9 in an allosteric manner. Our study hence expands the inhibitor toolbox for the temporal control of CRISPR-Cas activity.
Results
Intact M13 phage inhibits the in vitro DNA cleavage activity of SpyCas9
In a conventional phage display experiment, we surprisingly discovered that intact M13 phage [50] itself could block the DNA cleavage activity of purified SpyCas9 proteins in dose-dependent manner with an approximate half maximum inhibitory concentration (IC50) of 5 nM (Fig. 1a), which corresponds to a phage titer of 3×109 PFU/μL. Interestingly, phage-mediated SpyCas9 inactivation occurred only if phage was supplemented to the reaction prior to the addition of sgRNA, but not post the formation of SpyCas9-sgRNA ribonucleoproteins (RNPs) (Fig. 1b). The order-of-addition-dependent inhibition suggested that competition for sgRNA-binding site in SpyCas9 is a possible mechanism of the inhibitory activity of M13 phage.
M13 phage major coat protein-derived peptide inhibits the in vitro DNA cleavage activity of SpyCas9
Next we sought to determine the components in M13 phage that contribute to SpyCas9 inactivation. M13 phage has a simple, compact genome comprising of 11 protein-coding sequences. Considering the accessibility, surface-exposed phage proteins most likely serve as the sources of Cas9-inhibiting agents. Fully packaged M13 phage contains approximately 2700 copies of major coat protein pVIII (G8P) and 5 copies each of minor coat proteins pIII, pVI, pVII and pXI [51–53]. We initiated our investigation on the major coat protein G8P due to its abundance on M13 phage surface. The 73-amino acid major coat protein G8P contains four segments including signal peptide, periplasmic domain, transmembrane helix and cytoplasmic domain (Fig. 1c). After G8P maturation, signal peptide is cleaved and transmembrane helix is inserted into phage membrane, leaving periplasmic domain the only region on phage surface. Therefore, we synthesized a 21-amino acid peptide constituting the periplasmic domain of G8P (G8PPD) and examined its inhibitory activity on SpyCas9.
We found that G8PPD inhibited the activity of SpyCas9 with an IC50 of 5 μM (Fig. 1d), which was 1000-fold lower than that of intact M13 phage. Similar to the intact phage, G8PPD suppressed the in vitro DNA cleavage of SpyCas9 in an order-of-addition-dependent manner (Fig. 1e), indicating that G8PPD may inactivate SpyCas9 by specifically interfering with apo-SpyCas9.
In an effort to Blast search peptide homologs of M13 G8PPD, we identified G8PPD from f1 phage which has only one amino acid difference from M13 G8PPD (Additional file 1: Figure S1a). Because M13 and f1 G8PPD peptides have similar amino acid sequences and Cas9-inhibiting activities (Additional file 1: Figure S1b), they were used interchangeably in the following studies.
G8PPD prevents the assembly of Cas9 and sgRNA by binding to apo-Cas9
SpyCas9 can be inactivated at distinct steps during its action including guide RNA binding, substrate DNA binding and DNA cleavage. Most previously known Acrs exert their inhibitory activity by interfering with DNA surveillance or cleavage. To understand the mechanism of actions of G8PPD, we examined the interactions between G8PPD and apo-Cas9 or Cas9-sgRNA RNP using electrophoresis mobility shift assay (EMSA). Under fixed sgRNA concentration of 15 μM, the fraction of Cas9-bound sgRNA increased with the increasing molar ratio of Cas9 and sgRNA in the absence of G8PPD. By contrast, pre-incubation of SpyCas9 proteins with 300 μM G8PPD reduced the fraction of Cas9-bound sgRNA in a G8PPD dose-dependent manner (Fig. 2a). We also noticed that under high molar ratio of Cas9 and sgRNA, G8PPD did not fully block the formation of Cas9-sgRNA RNP (Fig. 2a), suggesting a weak interaction between Cas9 and G8PPD.
G8PPD-mediated inhibition of Cas9-sgRNA assembly is dependent on the order of sgRNA addition. Under a fixed Cas9-sgRNA molar ratio of 0.3, pre-incubation of Cas9 with 300 or 600 μM f1 G8PPD prior to sgRNA addition abolished the assembly of Cas9 and sgRNA (Fig. 2b). By contrast, supplementation of 300 or 600 μM G8PPD post sgRNA addition had minor or no effect on the formation of Cas9-sgRNA complex (Fig. 2b). These results suggest that G8PPD prevents Cas9-sgRNA binding by interacting with apo-Cas9 but not sgRNA-bound Cas9 and may explain why G8PPD-mediated inactivation of SpyCas9 cleavage is dependent on the order of sgRNA addition.
Identification of G8PPD binding site in SpyCas9
In order to dissect the mechanism of interactions between SpyCas9 and G8PPD, we sought to determine the binding region of G8PPD on SpyCas9 using high-resolution mass spectrometry (MS). SpyCas9 protein and M13 G8PPD were crosslinked using collision-induced dissociation (CID)-cleavable cross-linker disuccinimido sulfoxide (DSSO) [54] and then were subject to digestion with chymotrypsin. The integration analyses of CID-induced cleavage of interlinked peptides in MS/MS and MS3 of single peptide chain fragment ions revealed high crosslinking scores (Fig. 3a) on residues K1158 of [K]SVKEL peptide and K1176 of E[K]NPIDFLEAKGY peptide from SpyCas9 (Fig. 3b-c and Additional file 1: Figure S2a-b). These peptides occupy a continuous region in the PAM-interacting (PI) domain of SpyCas9 (Fig. 3d) that is responsible for recognizing the PAM sequence on the non-complementary DNA strand [55]. Interestingly, this candidate G8PPD binding site dose not locate in the sgRNA or DNA binding pockets of SpyCas9. These results suggest that G8PPD does not directly compete with sgRNA but may instead function as an allosteric inhibitor.
Next we sought to perform mutational analyses on the candidate G8PPD binding sites in SpyCas9. Residues KSKVEL in K1158 mutant and EKNPID in K1176 mutant were mutated into alanines respectively. Mutant SpyCas9 proteins were purified into high homogeneity (Additional file 1: Figure S3a-b). In vitro cleavage reaction illustrated that alanine mutations at positions KSKVEL (K1158 mutant) and EKNPID (K1176 mutant) markedly reduced the DNA cleavage activity of SpyCas9 (Fig. 3e), suggesting the importance of the G8PPD binding sites for SpyCas9 activity.
α-helical structure is critical for the inhibitory activity of G8PPD
Next we performed an alanine scan on f1 phage G8PPD to illustrate its structural determinants for Cas9 inhibition. Four peptide mutants are designed with alanine mutations spanning the entire G8PPD sequence (Fig. 4a). Alanine mutations at positions 6 to 11 abolished the inhibitory activity of G8PPD while mutants 1, 3 and 4 retained the majority of Cas9-inhibiting activity (Fig. 4a). The major coat protein G8P adopts an α-helical structure [53]. Residues 6 to 11 are located at the N-terminus of the α helix and contain Pro6, Lys8, Phe11 and three native alanines (Fig. 4b). The abolished activity in mutant 2 suggested a critical role of positions 6 to 11, particularly the residues Pro6, Lys8 and Phe11. Circular dichroism (CD) spectra study revealed an α-helical structure-enriched feature for f1 G8PPD WT but not for mutant 2 (Fig. 4c), suggesting that Pro6, Lys8 and Phe11 are important for maintaining the α-helical structure in f1 G8PPD. Although the presence of residues 6 to 11 could directly participate in the interaction between Cas9 and G8PPD, the overall α-helical structure of G8PPD could be also important for its Cas9-inhibiting activity.
Modulation of SpyCas9 activity in human cells using inoviradae phage G8PPD
We next explored the potential application of G8PPD as an off-switch for the genome-editing activity of SpyCas9 in human cells. Ectopic expression of the full-length (G8PFL) or periplasmic domain (G8PPD) of M13 and f1 G8P at 24 h prior to Cas9-sgRNA transfection significantly suppressed the genome-editing activity of SpyCas9 in HEK293 cells (Fig. 5a). In comparison, Acr protein AcrII4A [38] blocked SpyCas9 cleavage on AAVS1 site whereas Neisseria meningitidis Cas9 (NmeCas9)-specific Acr protein AcrIIC3 [56] partially inhibited SpyCas9. Importantly, f1 G8PPD was capable to inhibit SpyCas9 activity across different genes and cell types (Fig. 5b). Consistent with the in vitro experiments, significant inhibition of the on-target activity of SpyCas9 in human cells was observed only when G8PPD was overexpressed prior to sgRNA transfection. Co-transfection of G8PPD and SpyCas9-sgRNA did not inhibit SpyCas9 cleavage (P>0.05) (Fig. 5c).
In order to have detailed understanding of the effects of G8PPD on the genome-editing activity of SpyCas9, we performed next-generation sequencing to analyze the profiles of edited genomic loci in the absence and presence of G8PPD. Despite reduced mutation rate, the mutation pattern of SpyCas9 along the 20 bp sgRNA-targeting site was not altered by G8PPD treatment, as characterized by the high-frequency editing events at 3 bp upstream of the PAM sequence [4] (Fig. 5d). Importantly, G8PPD treatment retained the distribution pattern of indel length, with 1-5 bp indel being predominant in the population (Additional file 1: Figure S4a). In addition, we observed modest decrease in the in-frame mutations (3N) (Additional file: Figure S4b), the mechanism of which is yet to be elucidated. Collectively, these data suggested that G8PPD treatment did not cause major alterations in the profiles of SpyCas9-induced mutations, thus highlighting the potential of G8PPD as a safe off-switch for the therapeutic applications of SpyCas9.
To expand peptide-based anti-CRISPR toolbox, we examined the G8Ps from other inoviridae phages (Additional file 1: Figure S5). Peptides constituting the periplasmic domain of these G8P (G8PPD) are synthesized and evaluated for the in vitro and in vivo activities. At a concentration of 100 μM, the G8PPD from M13, f1, Pf1 and I2-2 phage markedly inhibited the in vitro DNA cleavage activity of SpyCas9 while other G8PPD orthologs showed little inhibitory effects (Fig. 5e). Ectopic expression of M13, f1 and pf1 G8PPD in K562 cells significantly reduced the activity of SpyCas9 in HEK293 cells whereas G8PPD mutant 2 did not show inhibitory activity (P>0.05) (Fig. 5f). Our results suggested that inoviridae phage G8Ps could be leveraged to inhibit both the in vitro and in vivo activity of SpyCas9. The variations of G8PPD sequences and the difference in their inhibitory activities indicate that further engineering endeavors could be made to improve Cas9-inhibiting peptides.
G8P co-transfection improves the specificity of SpyCas9 at AAVS1 site in human cells
It has been reported that timed delivery of AcrIIA4 can improve the genome-editing specificity of nucleofected Cas9-sgRNA RNP complex [38]. We intended to investigate the effects of G8P peptides on the specificity of constitutively expressed SpyCas9, a more therapeutically relevant model. The aforementioned in vitro and in vivo data have suggested that the inhibitory effects of G8P on CRISPR-Cas9 is dependent on the accessibility of Cas9 protein to sgRNA. This observation prompted us to explore whether G8P can be leveraged, via timed delivery, to improve the specificity of SpyCas9. Unlike the above experiments using G8Ps as CRISPR-Cas9 off switch (Fig. 5), here we co-transfected Hela cells with Acr or G8P plasmid and sgRNA and SpyCas9-coding plasmids. AcrIIA4 suppressed the on- and off-activities of SpyCas9 to undetectable levels, as determined by T7E1 assay (Fig. 6a). Next-generation sequencing analyses of edited Hela cells showed that M13 G8PFL and f1 G8PFL significantly reduced (P<0.05) both the on-target and off-target activity of SpyCas9 though the inhibitory effects on the off-target site appeared to be more prominent (Fig. 6b). AcrIIC3 significantly reduced the on-target activity (P<0.05) but not the off-target activity (P>0.05) (Fig. 6b). Importantly, M13 G8PPD could reduce the off-target events without affecting the on-target cleavage (Fig. 6a-b). Similarly, M13 G8PPD reduced the off-target events of SpyCas9 in K562 cells but retained the on-target activity (Fig. 6c). Co-transfection of M13 G8PFL or f1 G8PFL in K562 cells led to minimum or no significant decrease of the on-target activity of SpyCas9 but markedly reduced the off-target events (Fig. 6c). Surprisingly, in K562 cells AcrIIC3 had little effect on the on-target activity of SpyCas9 but notably increase the off-target activity (Fig. 6c). Collectively, these data suggest that co-transfection of G8P-based anti-CRISPR agents can reduce the off-target events of constitutively expressed SpyCas9 in human cells with minimum perturbation on the on-target activity (Fig. 6d).
Discussion
In this study, we reported the discovery of anti-CRISPR peptides from inoviridae bacteriophages. The clue leading to this discovery was based on the observation that intact M13 phage inhibited the DNA cleavage activity of SpyCas9 in in vitro reactions. Although we focused the investigation on the major coat protein G8P in the present study, it is likely that other surface-exposed minor coat proteins may attribute to intact phage-mediated SpyCas9 inhibition.
Unlike previously described Acrs that inhibit CRISPR-Cas by disrupting DNA binding [36, 45, 57] or DNA cleavage [36], G8P inactivates CRISPR-Cas9 via a distinct mechanism by preventing Cas9 from sgRNA binding. Moreover, to achieve efficient in vitro and in vivo inhibition, G8PPD must access SpyCas9 prior to its binding with sgRNA, indicating that G8PPD binds to apo-Cas9, but not sgRNA-bound Cas9. These results suggest that G8PPD and sgRNA are mutually exclusive for binding with Cas9 nuclease. One straightforward explanation is that G8PPD and sgRNA compete for the same binding pocket. However, MS and mutational studies suggest that the binding site of f1 G8PPD is located on the PI domain of Cas9, distal from sgRNA or DNA binding pockets, thus suggesting against direct competition between G8PPD and sgRNA. It is known that the binding of guide RNA can induce conformational rearrangements of Cas9 [58], thus it is possible that upon G8PPD or sgRNA binding SpyCas9 can transform from a flexible conformation to a closed conformation, which prevents SpyCas9 from binding with the other counterpart. Collectively, G8P-mediated allosteric inhibition of Cas9 and sgRNA binding represents a unique CRISPR-inactivating mechanism that may have important implications for developing next-generation anti-CRISPR agents.
Existing Acr proteins can display nanomolar binding affinity to Cas9 [36]. By contrast, inoviridae phage G8PPD peptides exhibit weak affinity as evidenced by the micromolar IC50 in the in vitro cleavage reaction (Fig. 1) and by its inability to completely block the assembly of Cas9 and sgRNA (Fig. 2). Interestingly, intact M13 phage inhibits Cas9 with an IC50 of approximately 5 nM, 1000-fold lower than that with G8PPD. The increased potency with intact phage may result from the enhanced cooperativity and avidity afforded by the multimeric assembly of the phage capsid that carries 2700 copies of G8P.
Although we have demonstrated that inoviridae phage G8PPD can function as anti-CRISPR peptides, the biological relevance of our discovery is yet to be explored. Under native context, the major coat protein G8P does not enter bacterial cytoplasm during phage infection. It is thus unlikely for G8P to exert anti-CRISPR function at the early stage of phage infection. Instead, G8P could interact with CRISPR-Cas in bacterial cytoplasm after the phage genome is translated. This post-translational inhibitory activity will require G8P-coding DNA to evade CRISPR attack during phage infection. In addition, the SpyCas9-inactivating G8PPD discovered in the present study are encoded by phages that infect Escherichia coli and Pseudomonas aeruginosa, which do not harbor type II CRISPR-Cas system. Horizontal gene transfer could explain for the cross-species CRISPR inactivation [26, 32], however a systemic phylogenetic analysis is necessary to reveal the evolutionary implications of the anti-CRISPR activity of G8Ps.
Despite the elusive biological relevance of the anti-CRISPR activity of G8P, we nevertheless demonstrated that the genome-editing activity of SpyCas9 in human cells can be modulated by G8PPD. To the best of our knowledge, our discovery represents the first peptides known to exhibit Cas9-inhibiting activity and expands the anti-CRISPR agent toolbox which is currently composed of anti-CRISPR proteins [21], small-molecules [20] and synthetic oligonucleotides [59]. Compared with Acr proteins that are typically 10 to 20 kD in size, G8PPD peptides are small in size and can be chemically synthesized in large scale. The facile manufacturing process is critical for the rapid evaluation of the structure-activity relationship to identify enhanced anti-CRISPR peptides. The varied Cas9-inhibiting activities observed among different inoviridae phage G8PPD supports the notion of improving SpyCas9-inhibiting peptides by sequence optimization.
G8PPD does not alter the pattern of SpyCas9-induced mutations, suggesting that G8PPD-mediated Cas9 inactivation does not interfere with the downstream DNA repair pathway. This feature facilitates the therapeutic applications of G8PPD as CRISPR-Cas off-switches by restricting the alterations of genome-editing outcome. It has been proposed that the specificity of SpyCas9 in human cells may be increased by partial inhibition with weak CRISPR-Cas inhibitors [20]. In the current study, we found that transfection of strong CRISPR-Cas inhibitor AcrII4A inhibited SpyCas9 activity to near complete in human cells when transfected prior to (Fig. 5a) or simultaneously with (Fig. 6a) sgRNA- and SpyCas9-coding plasmids. By contrast, co-transfection of G8P with CRISPR-Cas9 showed little inhibition of the on-target activity but reduced the off-target events at AAVS1 site. Interestingly, it was observed that the AcrIIC3, a weak inhibitor to SpyCas9 (Fig. 5a), did not increase the specificity of SpyCas9 in human cells (Fig. 6b-c). This result along with the results of G8PPD suggest that the mechanism of actions of the inhibitors may be also important for their effects on the specificity of CRISPR-Cas9. It has to be noted, however, that the applicability of G8Ps as agents to improve the specificity of CRISPR-Cas9 requires further investigation on additional cell types, genomic loci or genome-wide mutation profile.
Conclusion
In the present study, we report the surprising discovery of phage-derived peptides that can inhibit the in vitro and in vivo activities of SpyCas9. These peptides inhibited Cas9 activity by disrupting Cas9 and sgRNA binding in an allosteric manner. We show that the genome-editing activity of CRISPR-Cas9 can be harnessed by G8P. This unique mechanism of action of G8P may provide insights into developing anti-CRISPR agents towards genome- or base-editing systems.
Methods
Production of M13 bacteriophage
To produce M13 phage, XL1-Blue E. coli was inoculated in SB medium supplemented with 2% glucose and grown until OD600 reached 0.5. M13 phages were transfected and incubated with XL1-Blue E. coli at 37 ℃ for 30 min and the culture was shaken at 37 ℃ for 1 h. Thereafter, bacteria were centrifuged twice at 3,000 rpm for 15 min at 4 ℃ and washed with SB medium to remove glucose and free phage particles. The cells were re-suspended with SB medium and diluted to an OD600 of 0.8 in 200 mL culture and grown at 30 ℃ overnight with shaking. The next day, the culture was centrifuged at 5,000 rpm for 15 min at 4 ℃ and the supernatant containing phage particles was collected. The supernatant was gently mixed with NaCl-PEG buffer containing 20% PEG and 2.5 M NaCl and kept on ice for 1 h. The PEG-phage supernatant was centrifuged at 9,000 rpm for 30 min at 4 ℃. The supernatant was discarded and phage pellet was centrifuged again at 9,000 rpm for 5 min at 4 ℃ to remove residual medium. The precipitated phage particle was re-suspended in phosphate buffered saline (PBS) for further studies.
Cell culture
HEK293T cells were maintained in Dulbecco’s Modified Eagle’s medium (DMEM, Gibco/Thermo Fisher Scientific, Shanghai, China) supplemented with 10% fetal bovine serum (FBS, Gibco/Thermo Fisher Scientific) at 5% CO2 and 37 °C in a fully humidified incubator and were passaged when 70-90% confluency was reached. K562 cells were cultured in RPMI-1640 medium supplemented with 10% FBS, 100 IU/mL of penicillin and 100 μg/mL of streptomycin at 37 °C under 5% CO2. Cell lines were validated by VivaCell Biosciences (Shanghai, China).
Expression and purification of SpyCas9 proteins
pET28b plasmids coding SpyCas9 WT, K1158 mutant and K1176 mutant proteins were transformed into E. coli BL21 (DE3) cells. Single colonies were picked and grown in 2 liter LB media supplemented with 50 μg/mL kanamycin. Culture was grown to an OD600 of 0.8. Protein expression was induced with 0.2 mM isopropyl-β-D-thiogalactopyranoside (IPTG) at 16 °C overnight. Cells from 2 liter culture were pelleted by centrifugation at 6,000 g at 4 °C for 15 min and then re-suspended in 40 mL binding buffer containing 20 mM TrisHCl, pH 8.0 and 0.5 M NaCl. Cell suspension was then supplemented with 1 mM Tris (2-carboxyethyl) phosphine (TCEP) and 1× complete inhibitor cocktail (Roche). Cells were lysed by sonication on ice and then centrifuged at 80,000 g at 4 °C for 30 min. The supernatant of cell lysate was incubated with 1 mL Ni-NTA agarose beads (QIAGEN) at 4 °C for 1 h. Th e resin was washed with 20 mL wash buffer that was made by supplementing binding buffer with 30 mM imidazole. Proteins were eluted with 5 mL elute buffer that was made by supplementing binding buffer with 300 mM imidazole. Eluted protein samples were further purified by gel filtration using Superose 6 10/300 column (GE Healthcare). These proteins were buffer-exchanged to storage buffer containing 20 mM HEPES, pH 8.0 and 200 mM NaCl, aliquoted and stored at −80 °C.
Construction of G8PPD overexpression plasmids
Human codon-optimized DNA sequences encoding M13, f1, f1 mutant 2 and pf1 G8PPD were cloned into the BamHI/XbarI sites of pcDNA3.1(+) by plasmid recombination kit Clone Express (Vazyme). These G8PPD peptides carry an N-terminal SV40 nuclear localization signal (NLS) for co-localization with Cas9 proteins. G8PPD peptides were cloned into plv-EF1α-mCherry plasmid harboring mCherry fluorescent protein marker. sgRNA was cloned into pGL3-U6-gRNA plasmid carrying green fluorescent protein (GFP).
In vitro transcription of sgRNA
sgRNA was transcribed from a sgRNA-coding PCR product with a 5′ T7 promoter sequence using HiScibe T7 Quick High yield RNA Synthesis kit (NEB). The transcription was performed at 37 °C overnight and then purified by phenol: chloroform extraction, followed by ethanol precipitation. Purified sgRNA was quantified by spectrometry and stored at −80 °C.
In vitro cleavage assay
Cas9 protein and transcribed sgRNA were incubated for 10 min at room temperature in reaction buffer containing 1× NEB buffer 3.1 (NEB Biolabs) supplemented with 1 mM DTT) to form Cas9-sgRNA RNP complex. Cleavage was performed in 10 μL reactions containing 100 ng of substrate DNA and 1 μL RNP complex of indicated concentrations at room temperature for 1 h. Reactions were terminated by addition of 1× DNA loading buffer and resolved on 2% agarose gels. For inhibition experiments, G8PPD peptides were dissolved in deionized distilled water and incubated with Cas9 protein or pre-assembled Cas9-sgRNA RNP for 10 min at room temperature and the mixed solution was then added to the in vitro cleavage reaction. For experiments comparing the inhibitory activities of G8PPD peptides, 0.1% dimethyl sulfoxide (DMSO) was included in the reaction solution to solubilize lyophilized peptide samples.
Electrophoresis mobility shift assay (EMSA)
sgRNA concentration is fixed to 15 μM and Cas9 protein was titrated with a molar ratio of Cas9 over sgRNA ranging from 0.05 to 1. f1 G8PPD of 150, 300 and 600 μM was incubated with Cas9 protein or pre-assembled Cas9-sgRNA RNP complex for 20 min at 25 °C and quenched by addition of 1× native DNA loading buffer containing 40 mM Tris, pH 8.2, 40 mM acetate, 1 mM ethylenediaminetetraacetic acid (EDTA), 12.5% (v/v) glycerol, 0.025% (m/v) bromophenol blue. The samples were run on 2% agarose gels. For the order-of-addition experiment, Cas9 and sgRNA concentrations are fixed to 7.5 and 15 μM respectively. SpyCas9 was pre-incubated with sgRNA or G8PPD for 20 min, followed by incubation with the counterpart for 20 min at 25 °C.
Chemical crosslinking and mass spectrometry
Cas9 protein and M13 G8PPD were crosslinked using collision-induced dissociation (CID)-cleavable cross-linker-disuccinimido sulfoxide (DSSO) following the described procedure [54] with minor modification. Cas9 protein and peptides were mixed in PBS and incubated for 30 min at room temperature. Crosslinking was performed for 30 min by adding DSSO (Thermo Scientific) to protein/peptide solution with 1,000 molar excess. The crosslinking reaction was quenched by excess Tris and the crosslinked products were digested with Chymotrypsin. The LC MSn data of digested peptides were collected on Orbitrap Fusion Tribrid (Thermo Scientific) with an on-line NanoLC system and analyzed using CID-MS2-MS3 strategy as previously described [60]. Monoisotopic mass of parent ions and corresponding fragment ions, parent ion charge states and ion intensities from LC MS2 and LC MS3 spectra were extracted using Xcalibur v 3.0 (Thermo Scientific). Database searching was performed using Proteome Discoverer v 2.2 software (Thermo Scientific). Chymotrypsin was set as the enzyme with two missed cleavages being allowed as the maximum values. Protein N-terminal acetylation, methionine oxidation (15.995 Da), carbamidomethyl cysteine (57.021 Da), hydrolyzed lysine DSSO (176.014 Da) and lysine DSSO TRIS (279.078 Da) were selected as variable modifications. In addition, to account for the residual crosslinker three defined modifications on uncleaved lysines were chosen including alkene (C3H2O, 54 Da), sulfenic acid (C3H4O2S,104 Da) and thiol (C3H2SO, 86 Da) modifications. A false discovery rate (FDR) of 1% was employed to filter out false positive results. The MS, MS2 and MS3 mass tolerances were set as 10 ppm, 20 ppm and 0.6 Da respectively.
The XlinkX detect program (Thermo Scientific) was used to search MS2 data and identify the list of putative DSSO-interlinked products based on their unique DSSO fragmentation patterns. Monoisotopic masses and charges of parent ions measured in MS3 scans for those putative cross-linked peptides were further validated and scored by XlinkX. The final results were confirmed by manual inspection of the MS2 and MS3 spectra, respectively.
Circular dichroism
Circular dichroism spectroscopy (Chirascan-plus CD spectrometer, Applied Photophysics) was used to probe the peptide conformational changes. 100 μM WT and mutant 2 G8PPD peptides were dissolved in deionized water. The CD data were recorded at 25 °C and the average value of three biological replicates was presented.
Inhibition of CRISPR-Cas9 activity in human cells by G8PPD overexpression
K562 cells (2 × 105) were harvested, washed with PBS and re-suspended in 20 μL of SF nucleofection buffer (Lonza). G8PPD-coding plasmid (1 μg) was mixed with re-suspended K562 cells and nucleofected by Lonza 4D nucleofector with program FF-120. Immediately following the nucleofection, 100 μL pre-warmed RPMI-1640 medium was added into nucleofection cuvettes and the cells were transferred to culture dishes. At 24 post nucleofection, plasmids encoding SpyCas9 (500 ng) and AAVS1 or EMX1-targeting sgRNA (250 ng) (Additional file 2: Table S1) were transfected into G8PPD-expressing K562 cells by nucleofection as described above.
Low-passage HEK293T cells were seeded into 24-well plates at a density of 150,000 cells per well. The next day, G8PPD plasmid (1 μg) were transfected into cells using Lipofectamine 3000 (Invitrogen). At 24 h post G8PPD transfection, plasmids encoding SpyCas9 (0.5 μg) and AAVS1-targeting sgRNA (250 ng) were co-transfected into G8PPD-expressing HEK293T cells. For co-transfection experiments, K562 and Hela cells were nucleofected with 1 μg of G8PPD plasmid, 0.5 μg of SpyCas9 plasmid and 0.5 μg sgRNA plasmid.
At 48 h after transfection of SpyCas9 and sgRNA plasmids, mCherry and GFP dual positive cells were collected using a BD FACSAria III flow cytometer (BD Biosciences). At least 2,000 cells were collected for subsequent analyses. The genomic DNA of sorted cells was extracted using QuickExtract DNA Extraction Solution (Epicentre). Genomic PCR reaction was performed using 100 ng genomic DNA, corresponding primers (Additional file 2: Table S2), Phanta Max Super-fidelity DNA Polymerase (Vazyme) or KOD plus (Takara) using a touchdown cycling protocol (30 cycles of 98 °C for 10 s, 68-58 °C for 15 s and 68 °C for 60 s). The PCR products were digested by T7E1 enzyme (NEB), resolved on 2% agarose gel and then analyzed by densitometry measurements as described [61]. Three biological replicates were performed for each condition.
Next-generation sequencing of edited genomic sites
Genomic DNA (100 ng) from AAVS1-edited K562 cells was subject to PCR reactions using stubbed primers (Additional file 2: Table S2). PCR products were purified using Gel Extraction Kit (OMEGA). A high-throughput library preparation kit (Hiseq3000 SBS&Cluster kit) was used to generate dual-indexed sequence. Two or three biological replicates were processed by Genergy Biotech (Shanghai, China) or Genewiz (Suzhou, Jiangsu, China) using Illumina HiSeq 3000 platform.
Next generation sequencing library preparations were constructed following the manufacturer’s protocol (VAHTS Universal DNA Library Prep Kit, Illumina). For each sample, more than 50 ng purified PCR fragment was used for direct library preparation. The fragments were treated with End Prep Enzyme Mix for end repairing, 5′ phosphorylation and dA-tailing in one reaction, followed by a T-A ligation to add adaptors to both ends. Size selection of adaptor-ligated DNA was then performed using VAHTSTM DNA Clean Beads. Each sample was then amplified by PCR for 8 cycles using P5 and P7 primers. Both P5 and P7 primers carry sequences that can anneal with flowcell to perform bridge PCR. In addition, P7 primer carries a six-base index allowing for multiplexing. The PCR products were cleaned up using VAHTSTM DNA Clean Beads, validated using an Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA, USA) and quantified by Qubit2.0 Fluorometer (Invitrogen, Carlsbad, CA, USA). Then libraries with different indexes were multiplexed and loaded on an Illumina HiSeq instrument according to manufacturer’s instructions (Illumina, San Diego, CA, USA). Sequencing was carried out using a 2 x 150 paired-end (PE) configuration. Image analyses and base calling were conducted by the HiSeq Control Software (HCS) + OLB + GAPipeline-1.6 (Illumina) on the HiSeq instrument. Sequencing reads were obtained in the Fastq format.
Amplicons with less than 6 M read counts were excluded from the analyses. Short reads were aligned to the reference sequence by Bowtie2 [62] with the following parameters: -D 5 -R 3 -N 1 --gbar 1 --rdg 5,1 --rfg 5,1 --dovetail. Aligned reads were sorted by SAMtools [63] and INDEL and SNP calling was performed by mpileup [64] with maximum read depth per sample equal to the total reads mapped. VarScan v2.4 [65] was used for the quality control of SNPs and INDELs in mpileup output with a minimum variant frequency of ≥ 0.001, and a P value threshold of ≤ 0.05. With the above settings, the following items were quantified including the proportions of reads with INDELs/SNPs at each position in the 20 bp target region, the proportions of INDEL with different insertion or deletion length, the proportions of INDEL reading frames (3N, 3N+1 and 3N+2) and the proportions of reads harboring variants over the total number of aligned reads.
Statistical analyses
Two or three biological replicates were performed for each experimental condition. Significant difference was analyzed using one-way ANOVA with Dunnett’s multiple comparisons test unless otherwise noted.
Author contributions
P.M. and J.L. conceptualized study. P.M., J.L., Y.-R.C. and S.-J.W. designed the experiments and analyzed data. Y.-R.C. and S.-J.W. performed the in vitro and in vivo Cas9-inhibiting experiments. J.C. analyzed next-generation sequencing data. J.Li manufactured M13 phage and initial phage selections. P.M. and S.W performed the crosslink experimenting. W.C. and W.Z. performed MS experiment and analyzed the data. B.M. and B.Y. provided purified WT SpyCas9 proteins and advice on the purification of other SpyCas9 variants. S.W. and Z.Z. helped to express and purify SpyCas9 proteins. B.J. and G.Y. provided critical resources. P.M. and J.L. wrote the manuscript. All authors discussed the results and commented on and approved the manuscript.
Funding
This work is supported by the National Natural Science Foundation of China (31600686 to J.L. and 31500632 to P.M.) and ShanghaiTech University Startup Fund (2019F0301-000-01 to J.L.).
Availability of data and materials
The mass spectrometry data have been deposited into ProteomeXchange with the accession number PXD012466. Deep sequencing data have been deposited into NCBI SRA database with the accession number SRP180801. Additional data that support the findings of this study are available upon request from the lead contact author J. Liu.
Ethics approval and consent to participate
Not applicable.
Consent for publication
Not applicable.
Competing interests
ShanghaiTech University has filed a patent application including the work described herein.
Acknowledgements
We thank the Analytical Platform and High-Throughput Screening Platform at Shanghai Institute for Advanced Immunochemical Studies (SIAIS) at ShanghaiTech University for the support of mass spectrometry and flow cytometry experiments.
Footnotes
The manuscript has been revised to address reviewers' comments following peer-review and to correct typos.
Abbreviations
- CRISPR
- Clustered regularly-interspaced short palindromic repeats
- Cas
- CRISPR-associated protein
- sgRNA
- Single guide RNA
- Acrs
- Anti-CRISPR proteins
- dsDNA
- Double-stranded DNA
- G8PPD
- Periplasmic domain of the major coat protein G8P
- G8PFL
- Full-length major coat protein G8P
- RNP
- Ribonucleoprotein
- EMSA
- Electrophoresis mobility shift assay
- MS
- Mass spectrometry
- CID
- Collision-induced dissociation
- DSSO
- Disuccinimido sulfoxide
- PI domain
- PAM-interacting domain
- CD
- Circular dichroism