INOM
EXAMENSARBETE BIOTEKNIK, AVANCERAD NIVÅ, 30 HP
STOCKHOLM SVERIGE 2017,
NOVEL AND EFFICIENT DELIVERY OF CRISPR/CAS9 FOR GENOME ENGINEERING IN EUKARYOTIC CELLS.
OSKAR GUSTAFSSON
KTH
SKOLAN FÖR BIOTEKNOLOGI
1
EXAMENSARBETE INOM BIOTEKNIK, AVANCERAD NIVÅ, 30 HP
STOCKHOLM, SVERIGE 2017
NOVEL AND EFFICIENT DELIVERY OF CRISPR/CAS9 FOR GENOME ENGINEERING
IN EUKARYOTIC CELLS.
Oskar Gustafsson
ogu@kth.se 19920503-5539
2
Contents
1. Abstract ... 3
2. Introduction ... 3
3. Materials and methods ... 8
3.1. Materials ... 8
3.2. Cell culture ... 8
3.3. Reporter system testing. ... 8
3.4. Flow cytometry ... 9
3.5. Synthesis of sgRNA ...10
3.6. sgRNA genome editing evaluation ...12
3.7. CPP to RNP ratio optimization ...12
3.8. Transfection buffer optimization ...12
3.9. sgRNA modification...13
3.10. In vitro cleavage assay ...13
3.11 Transmission electron microscopy ...14
3.12 Gel retardation analysis ...14
4. Results ... 15
4.1. Reporter cell evaluation and RNP-lipid transfection. ...15
4.2. Ratio optimization. ...16
4.3. Complexation buffer impact on genome editing efficiency. ...17
4.4. sgRNA length modification ...18
4.5. Complex investigation...20
5. Discussion ... 21
6. Future perspectives ... 24
7. Appendix ... 25
3
1. Abstract
The CRISPR (clustered, regularly interspaced, short palindromic repeat)/Cas9 system holds tremendous applications and therapeutic value. However, a key limiting factor is the delivery of the system into cells, both in vitro and in vivo. Existing delivery methods leaves much to be desired. An alternative delivery method is cell-penetrating peptides (CPP) which are short peptides with proven cell- penetrating abilities which has the capability to deliver cargo into eukaryotic cells. In this master thesis, CPPs was studied as a delivery method for CRISPR/Cas9 ribonucleoprotein (RNP). The CRISPR/Cas9 RNP was complexed with CPPs in a simple, non-covalent manner and successfully delivered to eukaryotic cells. The genome editing efficiency of CPP delivery was found to be
comparable to leading lipid transfection agents. However, further optimization of this delivery method is needed for both in vitro and in vivo applications.
Abstrakt
CRISPR/Cas9 systemet har oerhört mycket applikationer och terapeutiskt värde. Dock så hålls
CRISPR/Cas9 systemet tillbaka utav svårigheter att transportera systemet in i celler, både in vivo och in vitro. Existerande transportmetoder lämnar mycket att önskas, en alternativ transportmetod är
cellpenetrerande peptider (CPP), vilka har bevisad kapacitet att transportera last över cellmembranet. I denna masteruppsats studerades CPPs som en transportmetod för CRISPR/Cas9 ribonucleoprotein (RNP). CRISPR/Cas9 komplexerades tillsammans med peptiden på ett enkelt, icke-kovalent sätt och blev framgångsrikt transporterat in i eukaryotiska celler. Effektivitet av den genomiska redigeringen var jämförelsebar med ledande lipid transformations agenter. Dock krävs ytterligare optimering för denna transportmetod, både för applikationer in vitro och in vivo.
2. Introduction
Programmable endonucleases have revolutionized the targeted genome engineering field, and made possible applications that only a decade ago would have been daunting1. The foremost of these
endonucleases is the type II CRISPR/Cas9 system. Which acts in nature as an adaptive immune response in bacteria and archaea against phages and plasmids by cleaving foreign DNA2,3. The system has been adapted for use outside of its intended purpose in nature, and is today used in genome editing in a large variety of organisms, such as bacteria4, plants5,6, diverse eukaryotic cells7–10 and human cells11–14. The system has also been used in vivo mice models to treat dystrophic mouse cells15–17, with human in vivo trials likely not far away.
A large difference between the CRISPR/Cas9 system and other targetable endonucleases, such as zinc finger nucleases (ZFNs) 18,19 and transcription activator-like effector nucleases (TALENs)20,21, is that the
4 DNA-specificity of the CRISPR/Cas9 is based on RNA-DNA interactions, instead of protein-DNA
interactions.
The CRISPR/Cas9 system can be retargeted by exchanging the target sequence of the RNA-guide, which is comparably easier than the retargeting ZFNs or TALENs. These systems require the protein sequence to be altered, requiring specialized protein engineering expertise22. This combined with difficulties in protein design, synthesis and validation have hampered previous targetable endonuclease systems from routine use. Which has made it possible for the CRISPR/Cas9 system to swiftly catch up in popularity1,23.
In nature Cas9 interacts with CRISPR RNA (crRNA) and trans-activating crRNA (tracrRNA) to form a RNP.
The crRNA confers specificity to the target sequence by base pairing and the tracrRNA performs a structural role in the RNP1, the structure of the RNP bound to DNA can be seen in figure 1. In genome editing applications it’s possible to use tracrRNA/crRNA in their original form or combine them into a single guide RNA (sgRNA). Henceforth, this thesis will use the sgRNA terminology. The binding of Cas9/sgRNA to the DNA triggers a double stranded break (DSB) were the cleaving of the DNA is carried out by two putative nuclease domains, HNH and RuvC. HNH cleaves the DNA strand complementary to the target sgRNA sequence while RuvC cleaves the opposite DNA strand. However, for the binding to DNA to occur, Cas9 needs a protospacer-adjacent motif (PAM), which is a short sequence downstream of the target sequence (e.g. NGG for S. pyrogenes) that is recognized by protein residues in Cas924. This must be taken into consideration when sgRNAs are designed, since all targeted sequences must end with the PAM sequence. Orthologs of the S.pyrogenes Cas9 belonging to other bacterial species have been found with different PAM sequences. It is believed that deeper analysis of bacterial CRISPR systems might unveil Cas9 enzymes with specificity against different PAM sequences1,25. It could also be possible in the future to change the PAM sequence at will through protein engineering.
Figure 1. sgRNA interacts with Cas9 forming the RNP which in turn interacts with DNA. The sgRNA forms several loops, three of which extrude from the Cas9.
5 The introduction of a DSB activates DNA repair pathways such as endogenous homologous
recombination (HR) or non-homologous end joining (NHEJ). Both of these pathways have valuable therapeutic applications, which arise from the site directed mutagenesis caused by NHEJ and the repair and replacement capabilities of HR14. These pathways are present in all three domains of life and serves a vital purpose in DNA repair. An unrepaired or misrepaired DSB can lead to chromosomal aberrations, senescence and apoptosis, it is therefore of importance that the generation of DSB by introduced
endonucleases is controlled and specific26. The HR results, in nature, in precise repair of the DSB resulting in an identical sequence as was present before the DSB, this is done using a DNA template with
homologous overlap on each side of the DSB. HR can be exploited in genome engineering to introduce a new gene or exchange nucleotides of a gene, which is done by introducing a DSB in the target sequence and providing a homologous DNA template containing the insert27. However, the overwhelming majority of DSB are repaired by NHEJ which frequently results in base insertions or deletions (indels), often causing frameshift mutations and premature stop codons26. Therefore NHEJ is often induced for gene knockout28. NHEJ mediates direct religation of the broken DNA molecules which is done independently of the type of DNA end and without the need for a DNA template26. The potential of HR to repair or insert new genes has generated considerable interest since the discovery of the pathway. However, efficient DSB repair through the HR pathway has proven difficult due to the majority of DSB being quickly repaired by the NHEJ29.
The specificity of targetable endonucleases is highly important, since off-target cleavage could have highly detrimental consequences for the cell and/or organism. It has been shown that the specificity and activity of Cas9/sgRNA are at least on par with previous targetable endonucleases23. However,
Cas9/sgRNA can in certain cases tolerate some mismatches in the RNA-DNA pairing, possibly introducing off-target cleaving30. Several approaches have been developed to decrease the off-target cleaving. One of which is the double-nicking approach. Cas9 nickase is a Cas9 that has had one of its nuclease sites
deactivated. The pairing of two of these nickases using two different sgRNAs makes it possible to induce a DSB, dependent on the recognition of two Cas9 enzymes instead of one. This was shown to decrease relative off-target effects by 50-1000 fold29. It has also been shown, quite counterintuitively, that using a truncated (<20 bp) sgRNA target sequence decreases relative off-target effects by a 5000-fold while not sacrificing on-target cleavage29.
Cas9 is unique in that it is a DNA targeted protein with specificity derived from a short RNA sequence.
The applications for this system extend far beyond the creation of DSBs. Catalytically deactivated Cas9 (dCas9) can be used for gene regulation on a genome-wide scale by targeting genes and interfering with their transcription. This method is known as CRISPR interference (CRISPRi) and has been shown to block RNA polymerase binding, transcription elongation or transcription factor binding, depending on the target. The interference was done with no detectable off-target effects due to the highly specific nature of Cas9/sgRNA31. The ability to multiplex Cas9/dCas9 to several targets, by having a pool of different sgRNA, makes is possible to repress several genes simultaneously in a reversible manner31–34. The CRISPRi system has also been fused to regulatory proteins, thus further modulating gene regulation. This results in stable and efficient transcriptional activation and repression in human and yeast cells33. One such application of CRISPRi could be the repression of the Huntingtin gene in humans, thus decreasing expression of m(HTT) protein which is the primary cause for pathology in Huntington’s syndrome. This
6 approach is currently being done using oligonucleotides29. There are numerous applications of a specific DNA targetable enzyme, the future no doubt holds further applications.
Despite attractive applications in a broad range of therapeutic settings, the efficient delivery of the CRISPR/Cas9 remains a key limiting factor. The successful delivery of the system is paramount to any application. The CRISPR/Cas9 can be delivered in three different forms: as DNA, as RNA and as an RNP22. The transfection of Cas9/sgRNA encoding plasmids through lipofection, electroporation or viral delivery has long been established. However, electroporation is not possible to perform on a systemic level in vivo and viral delivery suffers from size restrictions to the cargo35. The small cargo size of viral delivery vectors necessitates the need to separate delivery of Cas9 mRNA and sgRNA15. Furthermore, studies into
potential immune responses to viral vectors are needed36. Lipofection is the delivery of cargo, such as DNA, RNA or protein using lipid molecules. However, lipofection is to an extent toxic, inflammatory and immunogenic37,38. The plasmid CRISPR/Cas9 delivery may not be a suitable delivery method as there are many drawbacks connected to it. For example, plasmid DNA can be inserted into on-/off-target Cas9 cleavage sites, this can result in permanent expression of Cas9 / sgRNA39,40. Another side effect of introducing plasmid DNA into cells is the induced cell-stress and cyclic GMP-AMP synthase activation as well as cell immune responses due to the presence of bacterial sequences41,42. Lastly, one of the large drawbacks of plasmid transfection of the CRISPR/Cas9 system is the overexpression of CRISPR/Cas9 RNP.
It has been shown that overexpression of Cas9 RNPs over long periods of time leads to an significantly increased off-target cleavage risk43. The second approach is the delivery of sgRNA and mRNA coding for Cas9, this solves the issues of cell immune response to DNA in the cytoplasm and genome integration.
The delivery of mRNA has inherent issues with stability and immunogenicity which can addressed through the modification of the RNA44,45. However, the delivery of CRISPR/Cas9 system in RNA form still results in Cas9 RNP overexpression, albeit not in as high levels as plasmid transfection, leading to increased risk of off-target effects. The delivery of Cas9 RNP would solve many of the issues associated with the previous mentioned delivery methods. However, the RNP by itself holds no cell-penetrating activity and must be transported over the cell membrane by carrier agents or through physical methods such as electroporation46,47. One already established method for delivering RNPs is through lipid nanoparticles, these have achieved impressive levels of indels creation, 70%, after 72h, in vitro48. However, lipids can be detrimental to cells and organisms as mentioned earlier.
A promising carrier molecule is cell-penetrating peptides, which generally are cationic and/or
amphipathic peptides with cell-penetrating properties. The structure of CPPs varies greatly, from simple linear forms to branched structures with multiple attached functional groups. Cargo transfection using CPPs can be done by either conjugating CPP and cargo, thus creating a defined unit, or by non-covalent nanoparticle formation between CPP and cargo. These nanoparticles are formed mainly due to
electrostatic and hydrophobic interactions49. One of the early problems of these nanoparticles was the low stability in serum conditions; therefore, nanoparticle stability must be taken into consideration.
CPPs hold considerable interest for therapeutic therapies, given that CPPs can enable protein, RNA and DNA delivery into cells in vivo50–53. CPPs have previously been used to directly deliver programmable endonucleases, such as covalently attached CPP-TALENs or CPP-Cas9 / CPP-sgRNA47,49,54. In these studies,
7 one or more CPPs were covalently attached to the endonuclease granting it cell-penetrating properties.
While the uptake mechanism for CPPs is not fully determined and is heavily debated, it is generally accepted to be an active uptake involving endocytosis pathways49. CPPs entering the cell are to a large extent sequestered into the endo-/lysosomal compartments, as the uptake of the CPPs is believed to be largely dependent on endocytosis55. Therefore the escape from these compartments is vital for the effective delivery of cargos though the use of CPPs56,57. There are different approaches to increasing the endo-/lysosomal escape of the CPP and their cargo. One such approach is the incorporation of
protonatable domains to counteract pH-change in the endo-/lysosomal compartments55. Another approach would be to increasing the membrane interaction properties of the CPP to increase interaction with endo-/lysosomal membranes49. However, the drawback of these modifications are often increased cellular toxicity, thus a balance must be stuck between endo/lysosomal escape and toxicity49.
Ramakrishna et al successfully delivered the CRISPR/Cas9 system into human cells using conjugated CPP- Cas9 and sgRNA-CPP complexes. The CPP was covalently attached to the Cas9 protein and to the sgRNA.
However, a disadvantage of using conjugated CPP-Cas9 is that the sgRNA and CPP-Cas9 must be transfected separately since mixing of the two does not result in indel creation. Ramakrishna et al speculated that the negatively charged sgRNA neutralize the positive charge of the CPP thus blocking the CPP-mediated uptake. This results in the need of two separate treatments, one with Cas9-CPP and one with CPP-sgRNA47. This could result in decreased overlap between cells transfected with sgRNA and cells transfected with Cas9, reducing genome editing efficiency.
This project used RNP to form complexes with CPP in a non-covalent manner. Thus, making it possible to first form Cas9/sgRNA RNP, followed by complexation with the CPP, called PepSe, to form nanoparticles. The Cas9 used had 3xNLS (nuclear localisation signals) covalently attached to it, facilitating nuclear transport after delivery.
The PepSe in a amphipathic peptide with lipids attached to it. The proposed interaction between RNP and PepSe would be for the positively charged groups of the PepSe to interact with exposed negatively charged sgRNA, this
interaction is illustrated in figure 2. The PepSe has in previous work in the lab been non-toxic in vivo.
Figure 2. The envisioned interaction between PepSe and RNP is through the positively charged amino acids present on the PepSe and the negatively charged sgRNA loops.
8
3. Materials and methods
3.1. Materials
Human embryonic kidney (HEK) cells 293T (ATCC) with a stable incorporation of a Stop-light construct, Dulbecco’s Modified Eagle Medium (DMEM)(1X) + GlutaMAX™-I (Gibco®), Fetal Bovine Serume (FBS) (Gibco®), Opti-MEM® I (1X) (Gibco®), 0,05% Trypsin-EDTA (1X) (Gibco®), Penicillin-Streptomycin (P4333) (Sigma-Aldrich). LB-media pH 7,5 +/- 0.2, LA+Ampicilin agar plate, EndoFree® Plasmid Maxi Kit (Qiagen).
HiScribe™ T7 High Yield RNA Synthesis Kit (NEB), NucleoSpin® Gel, PCR Clean-Up kit (Macherey-Nagel), SeaKem® LE Agarose. Phusion® High-Fidelity DNA Polymerase master mix (NEB). Polyvinylpyrrolidone average mol wt 10,000 (PVP) (Sigma-Aldrich) (C6H9NO)x, Hepes (Roche), CaCl2 (Sigma-Aldrich), NaCl (Sigma-Aldrich). Glucose (Merck). Alt-R® S.p. Cas9 Nuclease 3NLS (Integrated DNA technologies), Peptide 2815 (PepSe)(Pepscan), Lipofectamine® 2000 (Thermo-Fisher), Lipofectamine® RNAiMAX (Thermo- Fisher). DAPI dye (Thermo-Fisher), Uranyl Acetate (1%), Formvar/carbon 200 mesh nickel grids (Agar Scientific).
3.2. Cell culture
HEK293T Stop-Light cells were maintained in DMEM(1X) + GlutaMax™ - I supplemented with 10% FBS and penicillin/streptomycin (100 U/mL and 100 µg/mL, respectively). The cells were passaged every 48 to 72 h. The cells were seeded at a concentration of 14.000 cells / 0.32 cm2 in an attachment-factor coated 96- well plate 24 h before transfection in a total volume of 100 µl, 30-70% confluence was achieved at the time of transfection.
3.3. Reporter system testing.
The stop-light reporter system is a stable construct permanently expressing mCherry with eGFP inducible by indel formation of +/- 1 or 2. The simplified construct displaying the relevant genetic elements is displayed in figure 4. The system only responds to indels of +/-1 or 2, meaning any frameshift multiple of 3 will be missed. Therefore, it is likely that 1/3 of indels are not discovered using this reporter system.
Figure 4. Frameshift of +/-1 or 2 will result in expression of both mCherry and eGFP, no indel induction or indels a multiple of 3 will result in only mCherry expression. The F2A element is a self-cleaving peptide and functions to cleave mCherry from any expressed eGFP.
9 The reporter system was tested by transfection. Different plasmid combinations were used: negative control plasmid and positive control. The positive control was two plasmids containing either Cas9 insert or target sgRNA insert, while the negative control consisted of either of the Cas9 or the target sgRNA insert plasmids. A total of 0.33 µg plasmid was used for each transfection together with 5 µl
Lipofectamine 2000 per µg plasmid. The transfection mixture was prepared following the manufactures recommendations. The prepared mixture was added to triplicate wells in a 96-well plate, followed by 24 h incubation and flow cytometry analysis.
3.4. Flow cytometry
Flow cytometry was carried out using MACSQuant® Analyser. DAPI was added to each analysed well to discern viable cells. The investigated properties were cell viability, mCherry and eGFP expression. Shown in figure 3 are the gates used to determine the eGFP expression of each well. The first gate selects the viable cells by investigating DAPI uptake, high intensity in the V1-A :: VioBlue-A channel indicates that the cells are non-viable. Due to DAPI uptake by dead cells with permeable cell membranes. The next two gates exclude cell-clumps, cell-fragments and bubbles from the gating by measuring the side vs forward scatter. The side vs forward scatter of single cells increases in a linear fashion while clumps and
fragments do not. The forth gate also measures side vs forward scatter and gates away particles too large or small to be cells. This is followed by gating of cells with emission of mCherry wavelength. Finally, the eGFP expressing cells are investigated by looking at cells with high emission in the GFP wavelength.
Figure 3. This dotplot is taken from the transfection of HEK293T stop-light cells using RNP2:b-CPP, 1.5 pmol per well and serves as a representative dotplot of eGFP expression quantification. The eGFP positive gate has been set in such a manner as to only truly positive eGFP expressing cells are counted. The strict gating result in missed true positive cells.
10
3.5. Synthesis of sgRNA
Plasmids, 2.2kb, containing a T7 promoter and unique sgRNA were amplified by PCR. Each plasmid contained extensions in rationally chosen location in the sgRNA, this is illustrated in figure 5. The linear template products were used by RNA synthesis using NEB HighScribe T7 kit. The primers used were:
Universal Forward: 5`-TTACGCTGGAGTCTGAGGC-3`, Reverse: 5`-AAAAAGCACCGACTCGGTG-3`, and a reverse primer binding further downstream of the sgRNA, creating an sgRNA tail. The primer binding sites can be seen in figure 5.
a)
b)
Figure 5. a) Insert #1-4 contains the target sequence and differs from each other in the length and location of the rationally added extra nucleotides. Insert #5 lacks both random added nucleotides and target sequence and contains only the sgRNA backbone. b) Shown here is the primer binding sites during PCR amplification, the extended reverse primer binds several bp after the sgRNA.
The original, MiniGene (IDT), plasmid backbone sequence was:
CCCGTGTAAAACGACGGCCAGTTTATCTAGTCAGCTTGATTCTAGCTGATCGTGGACCGGAAGGTGAGCCAGTGAG TTGATTGCAGTCCAGTTACGCTGGAGTCTGAGGCTCGTCCTGAATGATATGCGGCCTCGGCGCGTGATCTTACGGC ATTATACGTATGATCGGTCCACGATCAGCTAGATTATCTAGTCAGCTTGATGTCATAGCTGTTTCCTGAGGCTCAATA CTGACCATTTAAATCATACCTGACCTCCATAGCAGAAAGTCAAAAGCCTCCGACCGGAGGCTTTTGACTTGATCGGC ACGTAAGAGGTTCCAACTTTCACCATAATGAAATAAGATCACTACCGGGCGTATTTTTTGAGTTATCGAGATTTTCA GGAGCTAAGGAAGCTAAAATGAGTATTCAACATTTCCGTGTCGCCCTTATTCCCTTTTTTGCGGCATTTTGCCTTCCT GTTTTTGCTCACCCAGAAACGCTGGTGAAAGTAAAAGATGCTGAAGATCAGTTGGGTGCACGAGTGGGTTACATCG AACTGGATCTCAACAGCGGTAAGATCCTTGAGAGTTTACGCCCCGAAGAACGTTTTCCAATGATGAGCACTTTTAAA GTTCTGCTATGTGGCGCGGTATTATCCCGTATTGACGCCGGGCAAGAGCAACTCGGTCGCCGCATACACTATTCTCA GAATGACTTGGTTGAGTACTCACCAGTCACAGAAAAGCATCTCACGGATGGCATGACAGTAAGAGAATTATGCAGT GCTGCCATAACCATGAGTGATAACACTGCGGCCAACTTACTTCTGGCAACGATCGGAGGACCGAAGGAGCTAACCG CTTTTTTGCACAACATGGGGGATCATGTAACTCGCCTTGATCGTTGGGAACCGGAGCTGAATGAAGCCATACCAAA CGACGAGCGTGACACCACGATGCCTGTAGCAATGGCAACAACGTTGCGCAAACTATTAACTGGCGAACTACTTACT CTAGCTTCCCGGCAACAATTAATAGACTGGATGGAGGCGGATAAAGTTGCAGGATCACTTCTGCGCTCGGCCCTCC CGGCTGGCTGGTTTATTGCTGATAAATCTGGAGCCGGTGAGCGTGGGTCTCGCGGTATCATTGCAGCACTGGGGCC AGATGGTAAGCCCTCCCGCATCGTAGTTATCTACACGACGGGGAGTCAGGCAACTATGGATGAACGAAATAGACA GATCGCTGAGATAGGTGCCTCACTGATTAAGCATTGGTAATGAGGGCCCAAATGTAATCACCTGGCTCACCTTCGG GTGGGCCTTTCTTGAGGACCTAAATGTAATCACCTGGCTCACCTTCGGGTGGGCCTTTCTGCGTTGCTGGCGTTTTTC CATAGGCTCCGCCCCCCTGACGAGCATCACAAAAATCGATGCTCAAGTCAGAGGTGGCGAAACCCGACAGGACTAT AAAGATACCAGGCGTTTCCCCCTGGAAGCTCCCTCGTGCGCTCTCCTGTTCCGACCCTGCCGCTTACCGGATACCTGT CCGCCTTTCTCCCTTCGGGAAGCGTGGCGCTTTCTCATAGCTCACGCTGTAGGTATCTCAGTTCGGTGTAGGTCGTTC GCTCCAAGCTGGGCTGTGTGCACGAACCCCCCGTTCAGCCCGACCGCTGCGCCTTATCCGGTAACTATCGTCTTGAG
11 TCCAACCCGGTAAGACACGACTTATCGCCACTGGCAGCAGCCACTGGTAACAGGATTAGCAGAGCGAGGTATGTAG GCGGTGCTACAGAGTTCTTGAAGTGGTGGCCTAACTACGGCTACACTAGAAGAACAGTATTTGGTATCTGCGCTCT GCTGAAGCCAGTTACCTCGGAAAAAGAGTTGGTAGCTCTTGATCCGGCAAACAAACCACCGCTGGTAGCGGTGGTT TTTTTGTTTGCAAGCAGCAGATTACGCGCAGAAAAAAAGGATCTCAAGAAGATCCTTTGATTTTCTACCGAAGAAAG G. The plasmid backbone can be seen in figure 6.
Figure 6. Ordered IDT MiniGene – Amp plasmid backbone.
The sgRNA insert in plasmid #1-4 all contain the same target sequence and only differ in the length and location of added nucleotides. These nucleotides were added to increase interaction between the negatively charged sgRNA and the positively charged domains of the PepSe.
PCR amplification was done by adding the universal forward and one of the two reverse oligonucleotide primers to a Phusion High-fidelity mastermix according to manufacturer’s recommendations. The thermocycler settings used were: 98°C – 30 s, [98°C – 10 s, 63.7 – 30 s, 72°C 10 s] x 30 cycles. The reverse and extended reverse primer both binds downstream of the sgRNA, binding either on the 3´-end of the sgRNA or a distance downstream thus creating the linear templates for sgRNA1-5:a and 1-5:b. The longer template results in a significant sgRNA tail, since the T7 polymerase used during the in vitro transcription does not dissociate from the template at 3´-end of the sgRNA proper. The PCR product was purified using NucleoSpin® Clean-up kit and eluted in 30 µl water, the concentration was measured using a NanoDrop spectrophotometer NP80 (Implen). The size was confirmed using gel electrophoresis in a 1% SeaKem®
agarose gel with 1xTAE buffer. The purified PCR mixture was used for in vitro transcription following the manufacturers recommended conditions for shorter RNA transcripts. The sgRNA was then purified using chloroform-phenol extraction, chloroform extraction followed by ethanol precipitation. The size of the purified sgRNA was analysed on a 1% agarose gel. The sgRNA4:b failed to be transcribed, and was therefore not used in the experiments described in this thesis.
12
3.6. sgRNA genome editing evaluation
RNPs was created by mixing 330 ng Cas9 with sgRNA1-5:a-b in a molar ratio of 1:1.2 followed by 10 min incubation at RT. Both Cas9 and the sgRNA was dissolved in water. Following the established protocol48, 0.4 µl of RNAiMAX was used for every ≈ 0,6 pmol RNP. A total of 330 ng Cas9 was used for each reaction, resulting of 0,6 pmol RNP into each well in a 96-well plate. Opti-MEM was added after RNP incubation until a total final volume of 16.5 µl was achieved, followed by addition of 13.2 µl Opti-MEM mixed 3.3 µl RNAiMAX which had been incubated for 10 min at RT. This was incubated for 30 min and added to cells, the cells were incubated for 24 h followed by analysis using flow cytometry described in 2.3. Flow cytometry.
3.7. CPP to RNP ratio optimization
Different amounts of PepSe was added to 5.2 pmol RNP2:b, prepared as described in 2.6 sgRNA genome editing evaluation, in a HBG (Hepes buffered glucose) buffer with a final volume of 33 µl. The HBG consisted of 20 mM Hepes and 5 w/v% glucose. The molar ratios tested was: 1:10, 1:30, 1:60, 1:100, 1:120, 1:150, 1:200 molar ratio between RNP and PepSe. The RNP-CPP was incubated 40 min at RT, followed by addition to stop-light HEK293T cells seeded 24 h earlier. A total of ≈ 1.5 pmol was added per well resulting in a concentration of ≈ 0.014 µM RNP. The cells were analysed 24 h later using the flow cytometer, as described in 2.3 flow cytometry, to determine the genome editing efficiency of the different RNP2:b-PepSe ratios.
3.8. Transfection buffer optimization
The buffer optimization for transfection was done by preparing the 5.2 pmol RNP2:b as was done in 2.6 sgRNA genome evaluation, followed by addition of PepSe, in a molar ratio of 1:100 RNP to CPP. Quickly followed by buffer until a final volume of 33 µl was achieved. The PepSe was dissolved in water. The RNP- CPP was incubated 40 min at RT in the buffer, followed by addition to stop-light HEK293T cells seeded 24 h earlier. A total of ≈ 1.5 pmol was added per well resulting in a concentration of ≈ 0.014 µM RNP. The cells were analysed 24 h later using the flow cytometer, as described in 2.3 flow cytometry.
The buffers used were: HBG buffer, 10 w/v% PVP mixed with HBG buffer, HEPES, CaCl2, NaCL, Opti-MEM and water. The PVP solution was made by adding 20 w/v% PVP together with 12.6 w/v% sucrose to water. The PVP transfection solution was prepared by adding 16.5 µl PVP 20 w/v% with 6.5 µl HBG together with CPP in water to the RNP. The HEPES, CaCl2 and NaCl solutions were made by adding 23 µl isotonic buffer together with 5 µl CPP to 5 µl RNP. The solutions using Opti-MEM and water were done by adding 23 µl Opti-MEM or water together with 5 µl CPP to 5 µl RNP. The treatments were added to cells in serum containing media.
13
3.9. sgRNA modification
Extensions of the sgRNA2 and sgRNA5 were made according to the protocol described in 2.5 RNA synthesis. Each reverse primer used bind further downstream of the sgRNA insert, creating the template for an sgRNA tail of increasing length. The binding of the primers to the sgRNA insert can be seen in figure 7. The RNA synthesis was done using the same protocol described in 2.5 RNA synthesis, resulting in the extended sgRNA2:c-e and sgRNA5:c-e. These extended guides, 2:a-e and 5:a-e, were tested with RNAiMAX, using the same protocol as described in 2.6, with the exception of adding ≈ 1.5 pmol RNP per well.
The new guides were transfected using PepSe, ≈ 1.5 pmol RNP per well, with a molar ratio of 100:1 between RNP and CPP, in HBG buffer with 10% PVP. The treatment were otherwise prepared and administered as described earlier, the cells was incubated for 24 h, followed by analysis by flow cytometry.
Figure 7. Primers binding x bp, y bp and z bp from the sgRNA was used to PCR amplify sgRNA templates for the synthesis of sgRNA with extended tails.
3.10. In vitro cleavage assay
3 pmol Cas9 was mixed with sgRNA2:a-e and sgRNA5:a-e in a molar ratio of 1:1.2, followed by addition to 1 µg of target sequence plasmid. This was then incubated for 1h at 37°C followed by a heat inactivation step at 98°C for 30s. Samples were evaluated by gel electrophoresis using a 1% agarose gel in 1xTAE buffer.
14
3.11 Transmission electron microscopy
Electron microscopy was performed to investigate the complex formation between RNP and PepSe. The species analysed was RNP2:b, RNP2:b-CPP and RNP2:b-CPP with 10% PVP. The RNP2:b and RNP2:b-CPP was added to the mesh in a concentration of 0.1 µM while the RNP2:b-CPP with 10% PVP was added in the concentration of 0.01 µM. The samples were prepared following the protocol established in materials and methods 2.6/2.7/2.8, with a molar ratio of 100:1 between CPP and RNP. The samples were applied to a 200 mesh nickel grids coated with Formvar/carbon (Agar Scientific UK), followed by negative staining using an aqueous solution of uranyl acetate (1%) and visualized using the transmission electron microscopy.
3.12 Gel retardation analysis
0,49 pmol RNP was prepared as described in 2.6 sgRNA genome evaluation. RNP was mixed with CPP to achieve a molar ratio of 1:10, 1:20, 1:40, 1:60, 1:100. Added to this was 20% m/v PVP, with the resulting concentration of 10% m/v, HBG was added until the total volume of 20 µl. These samples were loaded, together with pure sgRNA and RNP, to a 1% SeaKem® Agarose gel with 1xTAE buffer.
15
4. Results
4.1. Reporter cell evaluation and RNP-lipid transfection.
The Stop-light HEK293T cells were transfected though lipid transfection with positive and negative control plasmids. The positive control plasmids consisting of two plasmids with Cas9 gene insert respectively a target sgRNA insert. While the negative controls consisted of only Cas9 plasmid, only sgRNA plasmid or Stuffer plasmid which is an empty backbone. The positive control Lipofectamine2000 transfection resulted in expression of eGFP (data not shown). As expected no eGFP expression was detected with the negative controls (data not shown).
It has previously been shown that lipid based transfection reagents can be applied for transfection of Cas9/sgRNA RNP48. Liang et al compared several such lipid reagent and achieved the highest indel creation by RNAiMAX – RNP transfection, therefore RNAiMAX was used for lipid transfection of RNP in this thesis. RNAiMAX was used to transfect the cells with RNP1-5:a-b, which consisted of Cas9 complexed with the sgRNA guides #1-5:a-b. This transfection resulted in eGFP expression in all cases expect 5:a-b, as expected due to sgRNA5 not containing the target sequence. The resulting eGFP expression and
efficiencies can be seen in figure 8, these results imply that the targeting is correct in all target guides and integrity of the sgRNA is adequate for genome cleavage. Failure in the design and synthesis of the sgRNA would result in a percentage of eGFP expressing cells similar to the untreated control. However, as can be seen in figure 8, there is an eGFP expressing population in the treated cells. The quality and size of the sgRNA synthesised were also probed by gel electrophoresis on a 1% SeaKem Agarose gel with 1xTAE buffer. The synthesised sgRNA were of the expected size with acceptable integrity (data not shown). The RNAiMAX mediated RNP treatment resulting in the largest eGFP expressing cell population was
erroneously identified as sgRNA2:b, due to an unexpected fluorescent shift. It is believed certain parts of the cell population become auto fluorescent in the same emission wavelength as eGFP, causing the observed shift. The auto fluorescent population can be seen in figure 8, with an emission intensity higher than the untreated population.
16 Figure 8. RNAiMAX transfection of RNP1-5:a-b. a) The
percentage of cells expressing eGFP and the Median fluorescence intensity (MFI) after 24 h. The transfection was done in serum using RNAiMAX, 0,6 pmol RNP per well. There is a correlation between the treatments resulting in higher eGFP expression and the treatments with high MFI b) Representative dotplots after treatment. Treatment with RNP1.b resulted in this well in 4.69% GFP expressing cells, while RNP5.b result in no eGFP positive cells. The gating was set in a strict manner as to exclude false positive cells.
4.2. Ratio optimization.
Determination of the optimal molar ratio of PepSe to RNP was done to optimize the transfection efficiency. As can be seen in figure 9, the ratio resulting in the largest percentage of edited cells were molar ratio 100:1 PepSe to RNP. The data suggest that the PepSe – RNP can enter mammalian cells and perform genome editing of target
sequences. Albeit at a very low efficiency. The low, <100, molar ratios are speculated to be insufficient to form stable
nanoparticles, while the high, >100, molar ratios are believed to form too strong complexes, unable to dissolve in the endo/-lysosomes.
Investigations using challenge assays could shed further light on the stability of the formed complexes.
Figure 9. The percentage of eGFP expressing cells after 24 h and the corresponding MFI of each molar ratio between PepSe and Cas9 RNP. The buffer used for complexation was HBG.
17
4.3. Complexation buffer impact on genome editing efficiency.
The solution in which the nanoparticles are formed greatly affect the transfection efficiency of the particles as it can affect the size and charge of the formed nanoparticles. This in turn affects the transfection efficiency, therefore, complexation was carried out in different buffers in search of the optimal buffer. The resulting genome editing efficiencies varies greatly, as can be seen in figure 10. The HBG buffer with an added 10% of PVP was the most efficient buffer with ≈ 3% genome editing according to the stop-light assay. PVP is an crosslinker with high polarity, which was believed to be able to interact with the polar groups of the nanoparticles, thus stabilizing them. While CaCl2 has similar levels of efficiency as PVP, the growth of the cells is likely severely inhibited, as can be seen in figure 10. The growth of the cells was evaluated by using the flow cytometry cell count data to calculate the number of cells present in each well. However, this method of evaluating cell growth is far from optimal, and cannot return definitive answers, further analysis with methods developed for this application is needed.
Figure 10. a) The percentage of the eGFP expressing cells together with the average MFI 24 h after each treatment. b) The average cell count in each triplicate, the cell count gives hints as to how the cells react to the transfection agent and buffer. As can be seen here, using CaCl2 as transfection buffer likely inhibits the cell growth to a large extent and is likely toxic to the cells. However, further analysis with methods devoted to cell growth inhibition and toxicity is required.
18
4.4. sgRNA length modification
It was hypothesised that increased sgRNA length in the 3´-end of the sgRNA would increase the accessibility of the sgRNA to the CPP. This was postulated to result in increased complexation which would result in increased nanoparticle formation and transfection efficiency. To that end, sgRNA2:c-e and sgRNA5:c-e was created, with extensions on the 3´-end of the sgRNA. An in vitro cleavage assay was performed to confirm the activity of the longer sgRNAs, all length variations of sgRNA2 cleaved the target sequence while none of the negative control sgRNA5 did. Stop-light cells were transfected with the new extended guides using RNAiMAX and PepSe. The results from the flow cytometry analysis of the
RNAiMAX transfection can be seen in figure 11. As shown, the sgRNA2:a has the highest transfection efficiency using RNAiMAX. However, performing the same transfection using PepSe the highest efficiency was achieved with sgRNA2:b as shown in figure 11.The average cell count in each RNP5:a-e is also shown in figure 11, the amount of cells in each triplicate appears to decrease with increased sgRNA length.
Figure 11. a/b) These two graphs display the average percentage of GFP expressing cells corresponding MFI 24 h after treatment. The wells in graph a) have been transfected using RNAiMAX, while the wells in graph b) have been transfected using PepSe. c/d) These graphs display the average cell number in each well 24 h after treatment of RNP5:a-e using RNAiMAX or CPP as transfection agent. The difference between the untreated control cell count and the treated is generally lower in the PepSe mediated transfection than the RNAiMAX mediated.
19 The auto fluorescent shift is visible during transfection with both sgRNA2a:e and sgRNA5a:e, increasing as the sgRNA length increase, implicating the length of the sgRNA to be the cause for the observed shift in fluorescence and decrease in cell count. The increasing shift can be seen in figure 12.
Figure 12. a/b) Graph a) and b) are representative dotplots of each RNP treatment after 24 h using RNAiMAX, dotplot a), or PepSe, dotplot b). It is possible to see the formation of two populations after treatment, with the population with
increased emission in the FITC-A channel increasing as the sgRNA length increase. This population shift is observable both in RNP2 and RNP5 transfection. Note that the number of dots in not representative of the number of cells in the sample, since different volumes was analysed in each treatment.
20
4.5. Complex investigation
The presence of complexes was evaluated by electrophoresis and transmission electron microscopy. The gel retardation investigation was performed by adding RNP2:b, RNP2:b with varying molar ratio to PepSe and sgRNA2:b to an agarose gel followed by gel electrophoresis. The samples with RNP complexed to CPP was retarded in their movement though the gel, indicating larger particle size. Furthermore, the
complexes had lower fluorescence intensity with higher CPP to RNP ratio reaching the lowest intensity at the highest ratio tested, implying that the dye is excluded from interaction with the sgRNA (data not shown). The presence of complexes was strengthened by EM analysis of RNP, RNP-CPP and RNP-CPP with added 10% PVP, as shown in figure
13. No complexes can be observed when observing only RNP, while complexes formed in the presence of CPP. The addition of PVP appears to contribute to the formation of more defined complexes, as can be seen in genome editing rate doubling with 10 w/v% PVP present during complexation. The complexes formed seemed to be formed from clusters of spherical nanoparticles.
Further images of RNP-CPP complexes in the presence of 10 w/v% PVP can be seen in the appendix figure 15-17.
Figure 13. RNP-CPP-PVP. a) Imaged here is the RNP only, the particles formed under the EM were mostly protein smears with no repeating pattern between each particle. b/c) The RNP complexed to the CPP in an HBG buffer. Here a larger number of particles are seen on the surface of the mesh, furthermore the particles have a common, repeating, pattern. d/e/f) Shown in these images is RNP-CPP complexed in the presence of 10 w/v% PVP. The particles display a clear coherent structure.
21
5. Discussion
This thesis shows that transfection of human cells using Cas9 RNP complexed to CPP in a non-conjugated manner is possible and comparable in genome editing efficiency to RNAiMAX transfection of Cas9/sgRNA.
The results are promising and further optimization will most likely improve the results. This is the first study, to the authors knowledge, of CPP-RNP mediated transfection in a non-covalent manner. Earlier studies have used CPP to transfect the CRISPR/Cas9 system by conjugating the CPPs to the Cas9 and achieved higher indel creation than what is achieved in this thesis. Staalh et al 46 and Ramakrishna et al 47 achieved gene expression of 7% and 7.8% respectively using very similar reporter systems as the stop- light system. However, Staalh used a concentration 70 times higher while Ramakrishna used a
concentration 115 times higher, during three consecutive treatments over 3 days, than the 14 nM used in this thesis. The high concentrations used imply that insufficient genome editing was achieved at lower concentrations. An interesting observation is that both Staalh et al and Ramakrishna et al applied their treatments for no more than 4 h, followed by change of media. This likely implies certain toxicity of the Cas9-CPP conjugates in the concentration ranges used. Therefore, even though the reported genome editing is higher in both previous articles using CPP to transfect RNP, the genome editing per enzyme used is lower. To note is that both Staalh et al and Ramakrishna et al used different cell lines than the HEK293T cells used in this thesis. Therefore, a direct comparison as is done above is not fair and definitive, but the comparison does provide certain hints about the transfection efficiency.
Recent, in-house, unpublished data indicate the presence of micelles in a solution containing PepSe. This data was achieved by increasing the concentration of PepSe in the presence of a Eyosin Y, a dye which becomes fluorescent in hydrophobic environments. It was determined that at sufficient concentrations, hydrophobic environments are formed in solutions of PBS containing PepSe. The suggested origin of these hydrophobic environments is micelle formation of the peptide, with the hydrophobic domains facing inwards and the positively charged groups facing outwards, ready to interact with negatively charged cargo. It is speculated that this also occurs when PepSe is complexed to RNP, and these micelles are the smallest spheres seen during EM imaging. An illustration of the proposed micelle can be seen in figure 14. Investigation of these micelles might provide insight into ways to further improve transfection efficiency.
22 Figure 14. The suggested formed micelle has a hydrophobic core consisting of the hydrophobic domains of the peptide combined with the lipids attached to each peptide. The hydrophobic core is surrounded by the hydrophilic cargo, in this case the RNP. Further characterisation of this possible micelle could contribute to rational experiment designs to increase transfection efficiency.
The buffer, the molar ratio, the sgRNA length all contribute greatly to the gene editing as can be seen in figures: 8-10, this is likely due to the characteristics of the nanoparticle formed. The polarity of the buffer likely contributes to the stability of the nanoparticles by increasing the Gibbs free energy of having PepSe hydrophobic domains exposed to the solution. The unfavorability of having increased free energy likely results in nanoparticles with hydrophobic cores. This increase in the free energy is believed to be insufficient in water for the formation of stable hydrophobic-core containing nanoparticles. Possibly explaining the lack of gene editing when complexes was formed in water. The measured gene editing of nanoparticles formed HBG buffer indicates that the increase in buffer polarity compared to water results in the formation of nanoparticles with hydrophobic cores. The addition of 10 w/v% PVP to the HBG buffer during complexation was done in the belief that the polar crosslinking polymer, which is known to interact strongly with polar substances, would stabilise the nanoparticles and increase transfection efficiency. The addition of PVP doubled gene editing, indicating increased transfection efficiency.
However, the means of this appears to be different from what was believed. Recent, in-house,
unpublished heparin challenge essay data of PepSe-RNA complexes together with 10 w/v% PVP showed that the addition of PVP destabilised the formed complexes (data not shown). The transfection efficiency
23 was increased for PepSe-RNA complexes in 10 w/v% PVP (data not shown), indicating that the less stable nanoparticles have a higher transfection efficiency. It could be that complexation in pure water results in too unstable nanoparticles, unable to reach the cells, while complexation in HBG buffer results in too stable nanoparticles, unable to properly dissolve in endo-/lysosomes. The addition of PVP to HBG buffer possibly brings the stability in the nanoparticles into the goldilocks range of CPP/RNP stability. This is however opposed by the EM images, figure 13, were the nanoparticles formed in the presence of PVP appears more defined and consistent. It could be that PepSe/RNP complexes behave differently from the PepSe/RNA complexes in the presence of PVP. Further studies of the stability and composition of the nanoparticles are warranted. Further work is also necessary to investigate why transfection is only observed during surprisingly high molar ratios of 100:1 CPP to RNP. It should not logically be possible for 100 PepSe peptides to interact with the exposed sgRNA or a single RNP, even with the extended sgRNA.
The second hypothesis investigated, that increased sgRNA length would lead to increased transfection efficiency appears to only be partially true. A short extension, such as is done in sgRNA2:b, appears to increase the transfection efficiency of RNP-CPP compared to the shorter sgRNA2:a. However, extended sgRNA2:c-e results in less editing and more fluorescent shift, which is possibly due to cell apoptosis giving rise to auto fluorescence. The shift can be clearly seen in figure 12, increasing with sgRNA length,
independent on sgRNA target sequence. It is possible that that transfection efficiency is increased by the longer sgRNA, however, this would be hard to observe due to the cell death caused by the longer sgRNA.
Further development is needed to determine if the transfection efficiency is increased, and if so how to decrease the cell reaction to foreign RNA and thus reduce toxicity.
In a study by Liang, X. et al RNAiMAX was used to transfect RNP into HEK293 cells with the achieved indel frequency was 48% after 24 h48. This is far higher than the indel frequency achieved in this master thesis using the same protocol, apart from using more than twice the concentration of RNAiMAX-RNP. The indel frequency is lower even when the lagg time between indel creation and eGFP expression taking into consideration. This likely results in an underestimation of indel frequency. The highest percentage of cells expressing eGFP achieved after 24 h, using RNAiMAX to transfect Cas9 RNP, was ≈ 6%. This likely equals
≈9% indels, due to frameshift of 1 or 2 result in eGFP expression. The low efficiency of the RNAiMAX transfection suggests either an inefficient reporter system or low quality sgRNA. It should therefore be possible to increase RNAiMAX efficiency by exchanging the cell line or synthesising new, high quality sgRNA. This effect could carry over into CPP transfections, resulting in increased efficiency.
The stop-light reporter system works by knocking either of two eGFP genes into frame, giving rise to eGFP expression if indels of +/- 1 or 2 in induced in the target sequence. This means that ≈ 1/3 of all indels are not detected using this reporter system. An alternative system to evaluate indel formation would be PCR amplification of target genes, followed by mismatch analysis using mismatch-
endonucleases. The endonuclease cleaves mismatched DNA strands. One such nuclease is the T7 endonuclease which can recognize and cleave single base pair mismatches, insert and deletions58. This reporter system has successfully been used to evaluate indel frequency of targeted endonucleases and has the advantage of simplicity, robustness and applicability to all cell types and gene targets37,48,58. In summary, this thesis demonstrates a novel, alternative and enhanced CPP mediated delivery of Cas9/sgRNA RNP to human cells in vitro. This delivery method has several advantages over delivery of genomic material encoding the CRISPR/Cas9 system, with the primary advantage being the shortened
24 and reduced presence of the RNP in cells, reducing off-target effect substantially. Furthermore, the non- covalent approach used in this thesis mediates flexibility compared to the covalent approaches
developed.
6. Future perspectives
There are several aspects of RNP transfection using Pep-SEA that was left unexplored in this thesis. It is of importance to investigate further how crosslinking agents can stabilize or destabilize the formed
nanoparticles. In this thesis only PVP was evaluated, which doubled the transfection efficiency, implying the possibility of further crosslinkers conferring higher efficiency.
Further characterisation of the stability, size and toxicity of the nanoparticles are required, this can be done through a heparin competition essays, nanoparticle tracking analysis and cell proliferation assays.
An interesting prospect would the disruption of the larger complexes formed when PVP is added to the solution during complexation, as seen in figure 13. This disruption could possibly be done through
vigorous mixing during RNP-CPP incubation or through the usage of ultrasonication before cell treatment.
It is possible that the smaller spherical units could have an increased transfection efficiency or reduced toxicity. An interesting point to investigate would be to investigate if micelles are formed when RNP is complexed to PepSe.
Another interesting aspect would be the exchange of the target sgRNA to target real genes and evaluate at the indel creation. If successful, then this could be extended to in vivo trials in gene disruption of therapeutic proteins. The creation of a new RNA synthesis protocol were the integrity of the sgRNA has a higher priority would ensure genome editing rates more accurately reflecting transfection efficiencies.
An additional interesting future perspective is the delivery of a DNA template together with the RNP to induce HR. This could be done by having the extended sgRNA contain complimentary sequences for the DNA template, allowing the DNA to be complexed into the nanoparticles together with the RNP. This could facilitate the delivery of the DNA template directly to the DSB, thus increasing the rate of HR.
25
7. Appendix
Figure 15. An overview picture of RNP2:b complexed with PepSe in a 1:100 molar ratio with 10% w/v ratio added PVP. There is a large variation in size of formed nanoparticles and it is unclear which of the particles are the active species, if not all.
26 Figure 16. An magnified image of RNP2:b complexed with PepSe in a 1:100 molar ratio with 10% w/v ratio added PVP. Once again there are larger nanoparticle complexes formed from smaller units.
27 Figure 17. An further magnified image of RNP2:b complexed with PepSe in a 1:100 molar ratio with 10%
w/v ratio added PVP. Shown here is the beginning of larger complexes, when the nanoparticles have just started to collect.
28 References
1. Doudna, J. A. & Charpentier, E. The new frontier of genome engineering with CRISPR-Cas9.
Science (80-. ). 346, (2014).
2. Horvath, P. & Barrangou, R. CRISPR/Cas, the immune system of bacteria and archaea. Science 327, 167–70 (2010).
3. Wiedenheft, B., Sternberg, S. H. & Doudna, J. A. RNA-guided genetic silencing systems in bacteria and archaea. Nature 482, 331–8 (2012).
4. Jiang, W., Bikard, D., Cox, D., Zhang, F. & Marraffini, L. A. RNA-guided editing of bacterial genomes using CRISPR-Cas systems. Nat. Biotechnol. 31, 233–9 (2013).
5. Bortesi, L. & Fischer, R. The CRISPR/Cas9 system for plant genome editing and beyond.
Biotechnol. Adv. 33, 41–52 (2015).
6. Shan, Q. et al. Targeted genome modification of crop plants using a CRISPR-Cas system. Nat.
Biotechnol. 31, 686–688 (2013).
7. Friedland, A. E. et al. Heritable genome editing in C. elegans via a CRISPR-Cas9 system. Nat.
Methods 10, 741–3 (2013).
8. Hwang, W. Y. et al. Efficient genome editing in zebrafish using a CRISPR-Cas system. Nat.
Biotechnol. 31, 227–9 (2013).
9. Dickinson, D. J., Ward, J. D., Reiner, D. J. & Goldstein, B. Engineering the Caenorhabditis elegans genome using Cas9-triggered homologous recombination. Nat. Methods 10, 1028–34 (2013).
10. Gratz, S. J. et al. Genome Engineering of Drosophila with the CRISPR RNA-Guided Cas9 Nuclease. Genetics 194, 1029–1035 (2013).
11. Cho, S. W., Kim, S., Kim, J. M. & Kim, J.-S. Targeted genome engineering in human cells with the Cas9 RNA-guided endonuclease. Nat. Biotechnol. 31, 230–2 (2013).
12. Cong, L. et al. Multiplex genome engineering using CRISPR/Cas systems. Science 339, 819–23 (2013).
13. Jinek, M. et al. RNA-programmed genome editing in human cells. Elife 2, e00471 (2013).
14. Mali, P. et al. RNA-guided human genome engineering via Cas9. Science 339, 823–6 (2013).
15. Nelson, C. E. et al. In vivo genome editing improves muscle function in a mouse model of Duchenne muscular dystrophy. Science (80-. ). 351, 403–407 (2016).
16. Long, C. et al. Postnatal genome editing partially restores dystrophin expression in a mouse model of muscular dystrophy. Science 351, 400–3 (2016).
17. Tabebordbar, M. et al. In vivo gene editing in dystrophic mouse muscle and muscle stem cells. Science (80-. ). 351, 407–411 (2016).
18. Bibikova, M., Beumer, K., Trautman, J. K. & Carroll, D. Enhancing Gene Targeting with Designed Zinc Finger Nucleases. Science (80-. ). 300, 764–764 (2003).
19. Porteus, M. H. & Baltimore, D. Chimeric Nucleases Stimulate Gene Targeting in Human Cells.
Science (80-. ). 300, 763–763 (2003).
29 20. Miller, J. C. et al. A TALE nuclease architecture for efficient genome editing. Nat. Biotechnol.
29, 143–148 (2011).
21. Joung, J. K. & Sander, J. D. TALENs: a widely applicable technology for targeted genome editing. Nat. Rev. Mol. Cell Biol. 14, 49–55 (2013).
22. Kelton, W. J., Pesch, T., Matile, S. & Reddy, S. T. Surveying the Delivery Methods of
CRISPR/Cas9 for <I>ex vivo</I> Mammalian Cell Engineering. Chim. Int. J. Chem.
70, 439–442 (2016).
23. Kim, S., Kim, D., Cho, S. W., Kim, J. & Kim, J.-S. Highly efficient RNA-guided genome editing in human cells via delivery of purified Cas9 ribonucleoproteins. Genome Res. 24, 1012–9 (2014).
24. Nishimasu, H. et al. Crystal Structure of Cas9 in Complex with Guide RNA and Target DNA. Cell 156, 935–949 (2014).
25. Esvelt, K. M. et al. Orthogonal Cas9 proteins for RNA-guided gene regulation and editing. Nat.
Methods 10, 1116–1121 (2013).
26. Davis, A. J. & Chen, D. J. DNA double strand break repair via non-homologous end-joining.
Transl. Cancer Res. 2, 130–143 (2013).
27. Li, X. & Heyer, W.-D. Homologous recombination in DNA repair and DNA damage tolerance.
Cell Res. 18, 99–113 (2008).
28. Bibikova, M., Golic, M., Golic, K. G. & Carroll, D. Targeted chromosomal cleavage and mutagenesis in Drosophila using zinc-finger nucleases. Genetics 161, 1169–75 (2002).
29. Burma, S., Chen, B. P. C. & Chen, D. J. Role of non-homologous end joining (NHEJ) in maintaining genomic integrity. DNA Repair (Amst). 5, 1042–1048 (2006).
30. Cho, S. W. et al. Analysis of off-target effects of CRISPR/Cas-derived RNA-guided endonucleases and nickases. Genome Res. 24, 132–41 (2014).
31. Qi, L. S. et al. Repurposing CRISPR as an RNA-Guided Platform for Sequence-Specific Control of Gene Expression. Cell 152, 1173–1183 (2013).
32. Bikard, D. et al. Programmable repression and activation of bacterial gene expression using an engineered CRISPR-Cas system. Nucleic Acids Res. 41, 7429–7437 (2013).
33. Gilbert, L. A. et al. CRISPR-Mediated Modular RNA-Guided Regulation of Transcription in Eukaryotes. Cell 154, 442–451 (2013).
34. Zhao, Y. et al. Sequence-specific inhibition of microRNA via CRISPR/CRISPRi system. Sci. Rep.
4, 3943 (2015).
35. Flotte, T. R. Size does matter: overcoming the adeno-associated virus packaging limit. Respir.
Res. 1, 16–8 (2000).
36. Mingozzi, F. & High, K. A. Immune responses to AAV vectors: overcoming barriers to successful gene therapy. Blood 122, 23–36 (2013).
37. Armeanu, S. et al. Optimization of Nonviral Gene Transfer of Vascular Smooth Muscle Cells in Vitro and in Vivo. Mol. Ther. 1, 366–375 (2000).
38. Dokka, S., Toledo, D., Shi, X., Castranova, V. & Rojanasakul, Y. Oxygen Radical-Mediated Pulmonary Toxicity Induced by Some Cationic Liposomes. Pharm. Res. 17, 521–525 (2000).
39. Gabriel, R. et al. An unbiased genome-wide analysis of zinc-finger nuclease specificity. Nat.
30 Biotechnol. 29, 816–823 (2011).
40. Gaj, T., Guo, J., Kato, Y., Sirk, S. J. & Barbas, C. F. Targeted gene knockout by direct delivery of zinc-finger nuclease proteins. Nat. Methods 9, 805–807 (2012).
41. Sun, L., Wu, J., Du, F., Chen, X. & Chen, Z. J. Cyclic GMP-AMP Synthase Is a Cytosolic DNA Sensor That Activates the Type I Interferon Pathway. Science (80-. ). 339, 786–791 (2013).
42. Wagner, H. Toll meets bacterial CpG-DNA. Immunity 14, 499–502 (2001).
43. Ishida, K., Gee, P. & Hotta, A. Minimizing off-Target Mutagenesis Risks Caused by Programmable Nucleases. Int. J. Mol. Sci. 16, 24751–71 (2015).
44. Karikó, K. et al. Incorporation of Pseudouridine Into mRNA Yields Superior Nonimmunogenic Vector With Increased Translational Capacity and Biological Stability. Mol. Ther. 16, 1833–
1840 (2008).
45. Shen, B. et al. Generation of gene-modified mice via Cas9/RNA-mediated gene targeting. Cell Res. 23, 720–723 (2013).
46. Staahl, B. T. et al. Efficient genome editing in the mouse brain by local delivery of engineered Cas9 ribonucleoprotein complexes. Nat. Biotechnol. (2017). doi:10.1038/nbt.3806
47. Ramakrishna, S. et al. Gene disruption by cell-penetrating peptide-mediated delivery of Cas9 protein and guide RNA. Genome Res. 24, 1020–7 (2014).
48. Liang, X. et al. Rapid and highly efficient mammalian cell engineering via Cas9 protein transfection. J. Biotechnol. 208, 44–53 (2015).
49. Lehto, T., Ezzat, K., Wood, M. J. A. & EL Andaloussi, S. Peptides for nucleic acid delivery. Adv.
Drug Deliv. Rev. 106, 172–182 (2016).
50. Nakase, I., Tanaka, G. & Futaki, S. Cell-penetrating peptides (CPPs) as a vector for the delivery of siRNAs into cells. Mol. Biosyst. 9, 855 (2013).
51. Regberg, J., Eriksson, J. N. K. & Langel, U. Cell-penetrating peptides: from cell cultures to in vivo applications. Front. Biosci. (Elite Ed). 5, 509–16 (2013).
52. Järver, P. et al. Peptide-mediated Cell and In Vivo Delivery of Antisense Oligonucleotides and siRNA. Mol. Ther. Nucleic Acids 1, e27 (2012).
53. Ezzat, K. et al. PepFect 14, a novel cell-penetrating peptide for oligonucleotide delivery in solution and as solid formulation. Nucleic Acids Res. 39, 5284–98 (2011).
54. Liu, J., Gaj, T., Patterson, J. T., Sirk, S. J. & Barbas III, C. F. Cell-Penetrating Peptide-Mediated Delivery of TALEN Proteins via Bioconjugation for Genome Engineering. PLoS One 9, e85755 (2014).
55. Lehto, T., Ezzat, K., Wood, M. J. A. & EL Andaloussi, S. Peptides for nucleic acid delivery. Adv.
Drug Deliv. Rev. 106, 172–182 (2016).
56. El-Sayed, A., Futaki, S. & Harashima, H. Delivery of Macromolecules Using Arginine-Rich Cell- Penetrating Peptides: Ways to Overcome Endosomal Entrapment. AAPS J. 11, 13–22 (2009).
57. Cleal, K., He, L., D. Watson, P. & T. Jones, A. Endocytosis, Intracellular Traffic and Fate of Cell Penetrating Peptide Based Conjugates and Nanoparticles. Curr. Pharm. Des. 19, 2878–2894 (2013).
58. Huang, M. C., Cheong, W. C., Lim, L. S. & Li, M.-H. A simple, high sensitivity mutation
31 screening using Ampligase mediated T7 endonuclease I and Surveyor nuclease with
microfluidic capillary electrophoresis. Electrophoresis 33, 788–796 (2012).
www.kth.se
