Best Practices for Utilizing CRISPR Advances

Advances in CRISPR technology have led to increased applications, as well as improvements in the speed, ease, accuracy, precision, and efficiency of gene editing. Biocompare recently hosted a Bench Tips webinar where senior post-doctoral fellows discussed how they were utilizing these innovations in editing techniques and reagents in their labs. They shared their best practices and technical know-how and below are some learnings that were captured. The complete "Practical Tips to Optimize and Improve Gene Editing" webinar, including Q&A with audience, is available on demand.

The prokaryotic adaptive immune response is mediated by Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) and CRISPR-associated proteins (Cas), referred to as the CRISPR-Cas system. When the cell is exposed to a foreign genetic material, fragments of that DNA get incorporated into the prokaryotic CRISPR array as spacer sequences between the repeats. Later when the cell encounters the same infection, the CRISPR array gets processed into individual crRNAs that assemble with Cas proteins and target the complementary invading DNA or RNA for cleavage, thereby providing immunity.

New RNA editing tools to minimize off-target effects

CRISPR-Cas nucleases can be easily programmed with different crRNA to target any complementary sequence, which has turned out to be very useful for gene editing. Cas9 is the most commonly used nuclease that targets DNA, leading to permanent double-strand breaks. Similarly, Cas13 is a RNA-targeting nuclease that cleaves single-stranded RNA, leading to its degradation by exonucleases. Like RNA interference (RNAi), Cas13 editing causes a transient, not a permanent, RNA knockdown.

However, one of the main drawbacks of Cas13 editing is that it cleaves non-specifically at a region distal to the target binding site. This often causes cellular toxicity due to widespread RNA degradation. David Colognori, Ph.D., a postdoctoral fellow in the laboratory of Nobel Prize winner Jennifer Doudna at the University of California, Berkeley, is looking to develop better CRISPR-based tools for RNA editing. He is working with the Csm complex, a multiple subunit RNA-targeting nuclease, for precise RNA targeting in eukaryotes. He explained that unlike Cas13, Csm makes several cuts but only within the region of target complementarity. “The big question is off-target effects, and this is where Csm outperforms both shRNAs (short hairpin RNAs) and Cas13 for all targets tested.”

In his studies, Cas13 shows ten times the number of off-target effects compared to Csm. With Csm, it is possible to multiplex several crRNAs against different targets and there is no observable toxicity. Csm can also enter the nucleus and knock down nuclear RNAs. “Our initial approach was to deliver all of the components to cells on separate plasmids, but later we sought to consolidate everything into a single vector, which required some optimization,” said Colognori. The team tested three nuclear non-coding RNAs and eight cytoplasmic mRNAs. “Surprisingly, the knockdown was highly efficient, with at least one crRNA per target giving 90% knockdown.” The knockdown persisted for several days, peaking two to three days post-transfection and waned thereafter, consistent with the transient delivery of Csm by transfection.

Base editing to improve specificity

Base editors were first designed in Dr. David Liu’s lab at the Broad Institute by modifying Cas9 into a dead Cas9 or in subsequent versions a nickase Cas9, so it could no longer make DNA breaks. By attaching a DNA-editing module to the Cas9, they were able to turn it into a guided DNA mutagenizer. For instance, the nucleoside base cytosine (C) could be converted into a thymine (T) by attaching a cytidine deadminase to Cas9. Similarly, an adenine base editor could switch adenine (A) to guanine (G). These CRISPR/Cas9 base editors could mutagenize DNA, as well as RNA and proteins. Specificity was further increased using a variant of Cas9 that only required NG PAM instead of NGG PAM for binding. Since guanine occurs frequently in the genome, this NG-Cas9 could be used to edit several regions.

“Broad specificity NG-Cas9, combined with either C->T or A->G base editors, allows you to densely mutagenize your target,” noted Benjamin Lampson, M.D., Ph.D., an instructor at the Dana-Farber Cancer Institute and a postdoctoral fellow in the lab of Nobel Laureate Dr. William Kaelin, who is using base editing to find drug-resistant mutants. “In other words, this approach greatly expanded both the type of mutations and the locations of mutations that could be introduced. You could essentially tile a gene.”

One of the limitations of using base editing is that one can’t exactly predict which residues will be mutagenized. For instance, if there are multiple cytosines near the PAM site where the base editor can reach, it could mutate any combination of them. “Unlike CRISPR/Cas9 knockout screens, it is not always clear exactly why a “hit” single-guide RNA (sgRNA) scored,” Lampson said. “So, what we did was take the individual sgRNA and reintroduce it into the cells with the base editor, and then deep sequenced the region of the genome to ask exactly what mutations were introduced.” Algorithms to predict mutations introduced by base editors are good, but not perfect. “Predictions are usually accurate, although we do occasionally see more edits than what is predicted.” According to Lampson, the efficiency of base editing varies based on guide and cell line. Hence, positive selection screens are best suited for base editing because they are not hindered as much by the noise that comes from ineffective guides.

Base editing to replicate disease etiology

Heterogenicity in cancer mutations is immense and complex. A significant portion of the mutations are nonsense mutation (sequence change to a stop codon) or missense mutations (produces an amino acid that is different from the original amino acid), and even single nucleotide variants (SNVs) lead to important functional consequences in tumorigenesis and drug sensitivity. “Heterogenicity is poorly represented in the model systems that we have today,” explained Adrian Vega, Ph.D., a postdoctoral researcher at Weill Cornell Medicine in the laboratory of Dr. Lukas Dow.

“We need technologies to engineer missense mutations to better recapitulate human biology.” The Dow lab has generated an inducible base editing (iBE) mouse model that enables efficient creation of targeted nonsense and missense mutations to better understand the immune microenvironment in the context of tumor development. “The mouse model carrying an optimized, inducible cytosine base editor enables temporal control of base editing in a wide variety of murine tissues and is capable of driving highly efficient base editing both in vitro and in vivo,” added Vega.

The mouse model helps overcome the problem of exogenous delivery of base editors, which tend to be big proteins and offers other advantages. Cas9 proteins can be immunogenic and activate adaptive immune response, which is a problem when studying tumor-immune interactions or immunotherapies. “In our mouse model the transgene is integrated in the genome and hence, there is no immune response,” noted Vega. Also, when the base editor is highly expressed and constitutively activated, there can be DNA and RNA off-target effects and toxicity. “In our mouse model there is only one copy of the base editor, and we can turn it on and off to minimize these effects.”

The team was also able to generate organoids from the iBE mice. Using these organoids, they can potentially introduce a gRNA to create a new mutation while the tumors are growing, to recapitulate what happens in humans. “This shows that we can potentially model in vivo sequential editing, which has been challenging to do,” Vega explained. Using editing technologies for generating in vivo cancer models harboring common cancer-associated mutations, and generating complex in vitro tumor organoids to facilitate multiplex editing signifies a big step toward precision medicine.