- Completely Synthetic Block Co-Polymer Micelles Are Alternatives to Lipid Nanoparticles for Intracellular Delivery
- Reporter Gene STOP Codon Successfully Knocked Out in Mice
Introduction
CRISPR-based genome editing tools show tremendous potential as therapeutic agents for the treatment of genetic diseases and viral infections, as discussed in previous Zone blogs and the June 2021 TriLink Research Spotlight titled, A Novel Base-Editing Strategy Promises to Treat Sickle Cell Disease. Nevertheless, safe and efficient delivery of CRISPR tools to target tissues continues to attract growing attention (see Chart). A wide variety of technical approaches are being explored, the main one being the formulation of nano-sized particles loaded with different types of CRISPR-related cargo.
In May 2021, Duan et al. published a review titled Nanoparticle Delivery of CRISPR/Cas9 for Genome Editing. The review can be consulted as a lead reference to gain a better understanding of the Cas9 endonuclease-mediated cutting of DNA, as well as of representative delivery strategies. This blog, which features co-encapsulation of Cas9 mRNA and single-guide RNA (sgRNA) in polyplex micelles for genome editing in the mouse brain, uses a relatively unique formulation strategy pioneered by über-famous polymer scientist Prof. Kazunori Kataoka, pictured here.
Kataoka is one of the world's leading scientists in bio-related and bio-compatible polymers. He has made several seminal contributions in polymer self-assembly and developed several breakthrough technologies, including the use of polymer micelles and complexes of ionic block copolymers for controlled drug delivery. The following section provides a brief synopsis of these ionic block copolymers. It will be followed by highlights of the featured application for co-delivery of TriLink's CleanCap® Cas9 mRNA and chemically modified sgRNA in mouse brain.
Ionic Block Copolymer Micelles
As the term implies, a copolymer is a polymer mixture comprised of two different monomeric repeating units designated as A and B. In contrast to an alternating copolymer (A-B)n structure, a block copolymer structure is comprised of repeating-unit "blocks," such as [(AA…)n-(BB..)m>
, in which the values of n and m can be varied to control the properties of the composite material. If A and/or B are monomers with amino groups, then protonation at physiological pH affords a positively charged cationic block copolymer. These positively charged ionic blocks can then spontaneously associate with anionic polymers, such as DNA and RNA, to form aggregates (aka polyplexes) for nucleic acid delivery.
What about micelles? The standard definition of a micelle is an aggregate (or supramolecular assembly) of surfactant molecules dispersed in a liquid, forming a colloidal (aka nanoparticle) suspension. In water, a typical micelle forms an aggregate with the hydrophilic "head" regions facing the surrounding solvent, sequestering the hydrophobic single-tail regions in the micelle center.
Following the emergence of hydrophilic poly(ethylene glycol) (PEG) conjugation to improve the pharmacology of proteins, reviewed elsewhere by its pioneer Frank F. Davis (1921–2021), "pegylation" of biologically active macromolecules became a widely employed strategy in drug design. Currently, this is an important aspect of PEG-lipid nanoparticle delivery of mRNA vaccines against COVID-19 (reviewed by Schoenmaker et al., 2021).
In many years of systematic investigations, Kataoka utilized the advantageous properties of PEG and the DNA/RNA-binding properties of cationic polymers to synthesize PEG-polypeptide block copolymers for drug and gene delivery, as reviewed elsewhere. This general approach involves an extensive body of synthetic organic and polymer chemistry; however, for this blog, it is helpful to focus on only a few aspects, which we will discuss in the following paragraphs.
The study led by Kataoka featured in this blog (referred to as Abbasi et al. from here on out) explored methods to deliver genome editing components using polyplex micelles (PMs) prepared from a PEG-polycation block copolymer, which forms a "core–shell" structure comprised of condensed RNA in the inner core with PEG as an outer shell. Specifically, these PMs were prepared using a pegylated polycation core of poly(N′-(N-(2-aminoethyl)-2-aminoethyl) aspartamide, termed PEG-PAsp(DET), as depicted here. NMR analysis of this synthetic block polymer indicated that the PEG portion (red brackets) had an average m ~ 300, while the PAsp(DET) portion (green brackets) had an average n ~ 78.
Importantly, previous investigations led by Kataoka (see Itaka et al., 2010) not only demonstrated that gene delivery via plasmid DNA transfection using PAsp(DET) is efficient, but also that PAsp (DET) is biodegradable, non-toxic to nonhuman primates, and facilitates endosomal escape, all of which are key factors required for clinical utility of mRNA delivery systems. Moreover, the PEG shell inhibits mRNA recognition by Toll-like receptors, allowing mRNA to be introduced with minimal inflammatory responses while enhancing the "percolation" of micelles into treated tissue (see Kataoka and coworkers, 2009).
Study Rationale
According to Abbasi et al., the search for an optimal non-viral vector is still underway when it comes to safety and efficacy. Notably, the issue of how to best co-deliver Cas9 mRNA (~4.5 kb) and sgRNA (~0.1 kb), which are substantially (~45-fold) different in size, remains poorly investigated. In this study, Abbasi et al. explored methods to address this issue of delivering the disparately sized mRNA and sgRNA genome editing components. They did so using polyplex micelles (PMs) prepared from biodegradable PEG-PAsp(DET) block copolymer.
Ideally, as depicted here, this strategy for simultaneous co-delivery formulation would have a core comprised of cationic PAsp(DET) (orange) bound to anionic RNAs (black), namely Cas9 mRNA and sgRNA, surrounded by a hydrophilic PEG (blue) shell termed PMCas9/sgRNA. Efficacy of co-delivery of these genome editing components by this formulation would be compared with the use of two separate control formulations comprised of PEG-Asp(DET) carrying only mRNA or only sgRNA, termed PMCas9 and PMsgRNA, respectively.
For in vivo genome editing in mice using these formulations, Abbasi et al. selected the brain as the target organ because many neurological disorders, including Huntington's disease, fragile X syndrome, and Alzheimer's disease, have a strong genetic component, which allows for targeting by CRISPR Cas9 mRNA/sgRNA. As will be discussed below, this co-encapsulation of Cas9 mRNA and sgRNA in PMs successfully induced efficient genome editing in neurons, astrocytes, and microglia following direct injection into the cerebral cortices of mice.
Additionally, co-loading Cas9 mRNA and sgRNA into a single PM exhibited markedly greater improvements in sgRNA stability than a PM loading sgRNA alone. Although previous studies have utilized large, biologically inert structural polyanions to help condense short nucleic acids into polyplexes, this new strategy is unique, as it enables the stabilization of sgRNA using Cas9 mRNA, where both components are functional and required for genome editing.
Cas9 mRNA and sgRNA Materials
A number of previous Zone blogs have discussed the advantages of using TriLink's CleanCap® technology for optimizing the yield, purity, and translational efficiency of in vitro translated (IVT) mRNA, an ability that an expert review labeled a "revolution." As detailed in the protocol recently published by Henderson et al., co-transcriptional IVT mRNA synthesis using a CleanCap® reagent is relatively straightforward, but a number of commonly employed mRNAs, including Cas9 mRNA, are also readily available as stocked CleanCap® mRNAs from TriLink.
Abbasi et al. used CleanCap® Cas9 mRNA encoding a version of the Streptococcus pyogenes SF370 Cas9 protein. This version of Cas9 mRNA has unmodified bases and encodes the Cas9 protein with an N and C terminal nuclear localization signal (NLS). Incorporating two NLS signals within the mRNA increases the frequency of delivery to the nucleus, thus increasing the rate of DNA cleavage. Additionally, a C terminal HA (hemagglutinin) epitope can be utilized as a tag to aid detection, isolation, and purification of the Cas9 protein. This mRNA is capped using CleanCap®, which results in the naturally occurring Cap 1 structure with high capping efficiency. It is polyadenylated and optimized for mammalian systems while also mimicking a fully processed mature mRNA. A newer version of CleanCap® Cas9 mRNA that is uridine-depleted, i.e., fully modified with 5-methoxyuridine (5moU), is now available.
As discussed in a TriLink eNews item available here, researchers at Stanford University using TriLink Cas9 mRNA reported that chemical modification of CRISPR RNA guide strands could enhance CRISPR-Cas9-mediated genome editing efficiency in human primary cells. Briefly, they tested several types of chemically modified sgRNA, including the 2'-O-methyl 3' phosphorothioate constructs depicted here, which performed well.
In the study by Abbasi et al., a chemically modified Ai9 sgRNA with 2′ -O-methyl 3′ phosphorothioate moieties introduced in the 3 terminal linkages at both the 5′ and 3′ ends was used, as represented by this 5' → 3' sequence:
[mA>
(ps)[mA>
(ps)[mG>
(ps)UAAAACCUCUACAAAUGGUUUUAGAGCUAGAAAUAGCAAGUUAAAAUAAGGCUAGUCC
GUUAUCAACUUGAAAAAGUGGCACCGAGUCGGUGC[mU>
(ps)[mU>
(ps)[mU>
(ps)U
Results and Discussion
Preparation of PMs Loaded with Cas9 mRNA and sgRNA: formulation of Cas9 mRNA and sgRNA in PMs was carried out in two ways: (i) PEG-PAsp(DET) block copolymer was added to the solution containing either Cas9mRNA or sgRNA to load Cas9mRNA (PMCas9) and sgRNA (PMsgRNA) separately, or (ii) the block copolymer was added to the mixture of Cas9mRNA and sgRNA (PMCas9/sgRNA). In both procedures, the weight ratio of Cas9mRNA and sgRNA was adjusted to 1:1.
Dynamic light scattering (DLS) measurements revealed a unimodal size distribution of PMCas9, with an average particle size of 72.2 ± 0.7 nm. However, PMsgRNA had an average particle size of 149.5 ± 3.1 nm and showed two peaks in the size distribution graph, indicating the presence of some larger secondary aggregates. On the other hand, PMCas9/sgRNA showed a monodisperse average size distribution of 64.9 ± 5.1 nm. Various FRET measurements using fluorescently labeled Cas9 mRNA (see Label IT™) and sgRNA demonstrated their co-encapsulation as a single core in PMCas9/sgRNA. All formulations showed near-neutral zeta-potential, an indicator of successful complexation and PEG shielding resulting in charge neutralization.
Stability of sgRNA and Cas9 mRNA in PMs: according to Abbasi et al., because PMs undergo extensive dilution upon in vivo administration, their stability following dilution dictates their efficacy. In light of this, the researchers evaluated RNA release upon dilution using non-denaturing agarose gel electrophoresis. PMCas9 was stable upon dilution, with no bands corresponding to free Cas9 mRNA detected throughout the tested dilution range of 1–100 μg/mL. In contrast, PMsgRNA released sgRNA upon dilution over the same range. The leakage of sgRNA after dilution was attributed to the weak interaction between PAsp(DET) and the relatively short sgRNA vs. Cas9 mRNA. However, when co-encapsulated with Cas9 mRNA (PMCas9/sgRNA), sgRNA was not released, as no bands corresponding to free sgRNA were observed even at a sgRNA concentration of 2 μg/mL.
Following in vivo administration, diluted PMs could also be subject to enzymatic attack. To determine nuclease stability of RNA under diluted conditions, quantitative real-time PCR (qRT-PCR) was used to test the integrity of Cas9 mRNA and sgRNA after incubation in 50% fetal bovine serum for 30 min at 37 °C. The researchers concluded that after incubation, the remaining amount of Cas9 mRNA in PMCas9 was comparable to that in PMCas9/sgRNA, indicating that Cas9 mRNA was stably encapsulated in both PMs after dilution. In contrast, sgRNA in PMsgRNA was extensively degraded, while that in PMCas9/sgRNA was more resistant to nucleases. This difference was attributed to the more stable packaging of sgRNA in PMCas9/sgRNA than in PMsgRNA, in which sgRNA release led to its enzymatic degradation.
Genome Editing in Mouse Brain: Next, Abbasi et al. evaluated the ability of PMs to induce in vivo genome editing. The polycation segment used in this study comprises PAsp (DET), which has protonation degrees of 0.51 and 0.82 at pH values of 7.4 and 5.5, respectively. Following endocytosis by the target tissue cells, the protonated amines at endosomal pH are expected to aid in the escape of Cas9 mRNA and sgRNA to the cytosol. The Cas9 endonuclease translated from the delivered mRNA exerts a specific double-strand break (DSB) in the genome in a region near the Protospacer Adjacent Motif (PAM) sequence, which is recognized by sgRNA through conventional Watson-Crick base pairing.
The resulting DSB is subsequently repaired viaendogenous repair pathways such as non-homologous end joining (NHEJ), making CRISPR/Cas9 a useful tool to knock out genes. In this study, Abbasi et al. designed a sgRNA to knock out the STOP cassette of tdTomato (td = tandem dimer) expression in Ai9 transgenic mice. This exceptionally bright red fluorescent protein is six times brighter than enhanced green fluorescent protein (EGFP) and is akin to mCherry (shown here), which makes it ideal for live animal imaging studies. Following genome editing, this reporter mouse model thus provides a robust red fluorescent signal to test the activities of PMs in vivo.
To determine whether PMs were a feasible option in the treatment of human neurodegenerative diseases, their ability to induce in vivo genome editing was evaluated following direct injection into the mouse brain. Since ideal genome editing modalities require only a one-time injection to obtain a permanent cure, an invasive but single injection of CRISPR/Cas9 in the brain parenchyma could potentially treat intractable brain disorders in humans. PMs were directly injected into the frontal lobe of the cerebral cortex in Ai9 mice. Tissue sectioning then occurred at 42 μm intervals on day 3 post-injection to observe tdTomato expression.
The sections exhibiting the most efficient tdTomato expression in each mouse for each treatment group were used for quantitative analysis. PMCas9/sgRNA showed more efficient genome editing compared to a mixture of PMCas9 and PMsgRNA, as indicated by a more widespread tdTomato expression area near the injection site in PMCas9/sgRNA. The low efficiency of genome editing after the delivery of a mixture of PMCas9 and PMsgRNA was attributed to either the low stability of PMsgRNA or the different tissue distribution profiles of Cas9 mRNA and sgRNA.
To rule out the latter possibility, the distribution of PMs in the brain tissue was observed using Cy3-labeled Cas9mRNA and Cy5-labeled sgRNA, prepared either by co-encapsulation (PMCas9/sgRNA) or separate formulation (PMCas9 + PMsgRNA). The Cy3 and Cy5 signals from Cas9 mRNA and sgRNA, respectively, co-localized in the brain tissue with both formulation methods. Thus, according to Abbasi et al., even though both Cas9 mRNA and sgRNA localized in the same tissue region following administration of separate PM formulations, their failure to induce efficient genome editing could be explained by the poor stability of PMsgRNA upon dilution in the brain tissue.
By contrast, efficient genome editing by PMCas9/sgRNA in the area surrounding the injection site was attributed to the rapid uptake of PMCas9/sgRNA into the cells surrounding the injection site shortly after injection. The possibility of such rapid cellular uptake in vitro was investigated using a neuronal cell line. PMCas9/sgRNA co-encapsulating Cy3-labeled Cas9 mRNA and Cy5-labeled sgRNA were applied to the hypothalamic neuronal cell line, GT1–7 before intracellular fluorescence signals were quantified by flow cytometry.
GT1–7 cells showed significant uptake of both Cas9 mRNA and sgRNA 30 min after addition of PMCas9/sgRNA , suggesting that uptake of PMCas9/sgRNA in brain tissue is rapid shortly after injection. Although the pathways involved in the cellular uptake of PMs into neurons and other brain cells remain unknown, Abbasi et al. noted that previous reports also showed efficient uptake of PEGylated nanoparticles through caveolae- and/or clathrin-mediated endocytosis. In addition, the efficiency of PMCas9/sgRNA in genome editing was compared to that of a non-PEGylated polyplex (polyplexCas9/sgRNA) prepared from Cas9mRNA, sgRNA, and PAsp(DET) homopolymer, which is the same polycation segment of the block copolymer used for the preparation of PMs in this study.
Importantly, the tdTomato signal observed following treatment with polyplexCas9/ sgRNA was less widespread compared to that seen with PMCas9/sgRNA. This could be due to the poor diffusion of polyplexCas9/sgRNA in the brain tissue, as indicated by the limited distribution of Cy3-labeled Cas9 mRNA or Cy5-labeled sgRNA in brain sections when co-encapsulated in this polyplex. Poor diffusion of polyplexCas9/ sgRNA in the brain could be attributed to the lack of PEG, cationic surface charge, or larger particle size. These reasons are consistent with the investigators' previous report, which demonstrated that PEGylated PMs loaded with pDNA in injected tumors percolated better than cationic polyplexes. In addition, non-PEGylated polyplexes aggregate in physiologic fluids, which can further compromise their diffusivity, cellular uptake, or subsequent intracellular events such as RNA release.
Profile of genome editing after injection of PMCas9/sgRNA: further analyses were performed to evaluate the detailed profile of genome editing using PMCas9/sgRNA. First, the distribution of genome editing at the injection site was corroborated by 3D imaging, constructed by compiling sequential 2D images obtained using two-photon confocal laser-scanning microscopy. PMCas9/sgRNA showed a cluster of red fluorescence that expanded around 200 μm in the brain tissue.
Finally, Abbasi et al. identified the type of genome-edited cells to assess the therapeutic potential of PMCas9/sgRNA in specific brain cells. Immunohistochemical analysis revealed that PMCas9/sgRNA induced genome editing in neurons, microglia, and astrocytes in the brain of Ai9 mice, as indicated by co-localization of tdTomato signal and the antibody-labeled specific cell markers.
Concluding Comments
The observations mentioned above highlight the advantages of PM-mediated co-delivery of IVT CleanCap® Cas9 mRNA and chemically synthesized sgRNA mRNA into a wide range of brain cell types in vivo for CRISPR-based genome editing with expanded therapeutic utility.
Although Abbasi et al. achieved this by knockout of the reporter protein tdTomato as a mouse model, the results support possible extension to proteins associated with human brain diseases. For example, Abbasi et al. suggest the therapeutic potential of PMs for the treatment of neurodegenerative diseases by genome editing in the neuron itself. The conversion of astrocytes or glial cells into neurons by knocking down specific genes provides a powerful approach to treating neurodegeneration by replacing lost neurons, they say.
In conclusion, Abbasi et al. state that "[t>
o the best of [their>
knowledge, this is the first report demonstrating genome editing in various brain cell types using RNA-based delivery of CRISPR/Cas9, potentiating the clinical rationale of RNA-based genome editing to treat intractable brain diseases in the future."
The Zone hopes that these novel findings, which were enabled by non-lipid PMs comprised of completely synthetic PEG-protein block copolymers, IVT mRNA, and sgRNA, will inspire others to pursue clinical development of therapies for brain diseases and other human maladies.
Your comments are welcomed, as usual.
Please feel free to share this blog with your colleagues or on social media.
