The perfect cut: a CRISPR tale where precision meets power

Any sufficiently advanced technology is indistinguishable from magic.

Background
Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR), a bacterial adaptive immune system in its native form, has revolutionized biotechnology and bioengineering as a programmable, precise and facile method of gene manipulation. With mainstream media and Netflix series like Unnatural Selection and  Human Nature bringing CRISPR into the public eye, its impact has become tangible. Last year, CRISPR therapy for Sickle cell disease (SCD) correction was approved in the UK and US, making genetic intervention for a permanent cure a reality rather than science fiction.
Since CRISPR’s inception as a genome engineering tool, the CRISPR-associated protein 9 (Cas9) from Streptococcus pyogenes (SpCas9) has garnered attention for its high efficiency in editing the human genome, albeit with high promiscuity of DNA interrogation leading to undesired off-target effect. In 2019, we published an article in Proceedings of the National Academy of Sciences (PNAS) characterizing an intrinsically specific Cas9 from Francisella novicida (FnCas9), demonstrating its potential for human genome editing1. We successfully showcased its capability by correcting the SCD disease mutation in Indian patient-derived induced pluripotent stem cells (iPSCs), illustrating its promise for therapeutic gene correction1. Notably, India carries a humongous disease burden of SCD, and this work propelled us towards developing therapeutic solutions for this disease with the motivation of reaching patients one day.
Challenge
The trade-off between the activity and specificity of Cas9s poses serious challenges in protein engineering. With the widely used SpCas9, efforts to reduce off-target effects have often come at the cost of simultaneous reduction of on-target editing efficiency. Similarly, engineering the protospacer adjacent motif (PAM)  to broaden the target scope of SpCas9 variants relaxed the PAM requirement but led to a concomitant increase in off-targeting coupled with attenuated on-target editing efficiency. Furthermore, recent mechanistic studies have shed light on the drawbacks of extreme PAM-flexibility. Discovery of orthologous Cas systems mined from microbial diversity failed to identify enzymes with efficiency comparable to SpCas9, and their complex PAM recognition limits the target range in the human genome. 
Motivation
Clinical success of CRISPR therapies contingents on the efficiency and specificity of Cas enzymes. FnCas9 has a simple 5′-NGG-3′ PAM requirement like SpCas9, and being an intrinsically specific enzyme showed remarkable specificity of DNA interrogation even at the genome-wide level. Interestingly, its DNA interrogation specificity comes from its weak engagement with the off-targets, unlike engineered SpCas9 variants which remain bound in the cleavage incompetent state with high affinity1, 2. This presents a lucrative scenario to engineer FnCas9 and develop engineered FnCas9 variants which can combine high efficiency with its intrinsic specificity.
Success in protein engineering 
Just before the world shut down in early 2020, we teamed up with structural biologist Prof. Osamu Nureki at The University of Tokyo whose group published the structure of FnCas9 in 20163. We engineered the PAM-duplex binding and phosphate-lock loop (PLL) regions of FnCas9 and developed enhanced (en) FnCas9 variants by structure-guided protein engineering. We discovered three enFnCas9 variants (en1, en15 and en31) carrying either single (en1 and en15) or multiple amino acid changes (en31) after screening 49 engineered variants. All three variants showed a >2-fold increase in the rate of DNA cleavage compared to FnCas9  (Fig. 1). enFnCas9 variants showed PAM-flexibility by recognizing 5’-NRG/NGR-3’ PAM in contrast to 5’-NGG-3’ by FnCas9, with en31 showing the highest rate of DNA cleavage. This altered PAM recognition resulted in ~3.5-fold increase in the target range of the human genome  (Fig. 1). Engineering further improved the binding affinity of enFnCas9 variants without sacrificing the single-nucleotide mismatch specificity, thereby resulting in robust FnCas9-based CRISPR diagnostics (CRISPRDx) platforms such as FnCas9 Editor Linked Uniform Detection Assay (FELUDA)4 and Rapid Variant AssaY (RAY)5.

Fig. 1:  Summary of the findings described in the research article.

enFnCas9 as a potent nuclease for genome editing technology 

We showed robust cellular genome editing up to ~90% indel efficiency by enFnCas9 variants, mirroring the efficiency of SpCas9 (Fig. 1). These variants comfortably outperformed the high-fidelity SpCas9 variants both in editing efficiency and specificity even at the highly promiscuous single-mismatch containing off-target. Furthermore, enFnCas9 variants exhibited remarkable genome-wide specificity in human cells, where SpCas9 and PAM-flexible SpRY variants showed up to ~1300-fold and ~700-fold higher off-targeting, respectively, compared to enFnCas9 variants, as seen in Digenome-seq studies aimed at in vitro identification of potential off-targets in the genome  (Fig. 1). Notably, enFnCas9 variants achieved high specificity despite being able to induce robust on-target efficiency in human cells, underscoring its superior properties as a potent genome editor.
enFnCas9 obviates the need of PAM engineering for adenine base editing
We showed the versatility and usefulness of enFnCas9 by demonstrating robust adenine base editing (up to ~70%) by en31-ABEmax8.17d at therapeutically relevant sites in human cells. More importantly, combination of longer spacer containing single guide RNA (sgRNA) and en31-ABEmax8.17d enabled tuning of adenine base editing window which is conventionally predefined w.r.t. Cas orthologs used and the constraints introduced by respective PAM requirements (Fig. 1). This unique ability of en31-ABE coupled with its PAM-flexibility expands the targeting range of pathogenic single-nucleotide polymorphism (SNP) to ~100% in the human genome. 
Therapeutic adenine base editing 
In collaboration with Dr. Indumathi Mariappan’s lab at LVPEI, Hyderabad, India, we demonstrated precise correction of G to A point mutation of RPE65 gene associated with inherited retinal dystrophy, Leber congenital amaurosis type 2 (LCA2) in patient-derived iPSCs (Fig. 1). Furthermore, we showed complete reversal of disease phenotype upon differentiating the base edited iPSCs into retinal progenitors, underscoring its remarkable therapeutic potential.
Publication detour
We could not justify this blog without touching upon the publication journey. This work was reviewed by Nature Biotechnology, and while the reviewers’ comments helped in strengthening our story, the experience was bittersweet. After two rounds of rebuttal, the paper was transferred to Nature Communications, where it was finally accepted in its current form.
Future ahead
The enFnCas9 variants described here have secured a US patent, one of the first patents in the CRISPR space from India6. RNA Biology Lab at CSIR-IGIB has initiated several clinical trials in collaboration with clinical partners to bring affordable gene therapies to India for correcting genetic diseases including SCD, retinal dystrophies, and neurodegenerative disorders7, 8. This study has paved the way for future applications and improvements that might have far-reaching therapeutic applications, especially in the area of genetic disorders.
References

1.    Acharya, S. et al. Francisella novicida Cas9 interrogates genomic DNA with very high specificity and can be used for mammalian genome editing. Proc. Natl. Acad. Sci. U. S. A. 116, 20959–20968 (2019).2.    Chen, J. S. et al. Enhanced proofreading governs CRISPR–Cas9 targeting accuracy. Nature 550, 407–410 (2017).3.    Hirano, H. et al. Structure and Engineering of Francisella novicida Cas9. Cell 164, 950–961 (2016).4.    Azhar, M. et al. Rapid and accurate nucleobase detection using FnCas9 and its application in COVID-19 diagnosis. Biosens. Bioelectron. 183, 113207 (2021).5.    Kumar, M. et al. FnCas9-based CRISPR diagnostic for rapid and accurate detection of major SARS-CoV-2 variants on a paper strip. Elife 10, (2021).6.    Chakraborty, D., Maiti, S. & Acharya, S. Kinetically enhanced engineered FnCas9 and its uses thereof. US Patent (2024).7.    Ghosh, A., Maiti, S. & Chakraborty, D. Emerging opportunities for gene editing therapies in India. Nat. Med. 30, 324–325 (2024).8.    Ledford, H. Hope, despair and CRISPR – the race to save one woman’s life. Nature 630, 284–288 (2024).