CRISPR-Cas enzymes, like Cas9, have transformed genome editing by making the process faster and more precise. However, the limitations in flexibility of Cas9 and other Cas enzymes becomes immediately clear when trying to make an exact change in DNA. This constraint became familiar to me when working with collaborators to design strategies to install precise changes, including generating specific SNPs to model various diseases. Often, a Cas9 target site could not be designed near the base pair of interest, and there would not be an efficient path towards making the mutation.
The feature of Cas9 and other CRISPR-Cas enzymes that dictates target site design is their requirement to recognize a DNA sequence called a protospacer adjacent motif (PAM). The necessity of the Cas protein to bind a PAM stems from their origins in prokaryotic adaptive immune systems, where the PAM allows the enzyme to distinguish between a genomic site harboring a stored “memory” of a previous phage infection and a target site in the DNA of an invading phage. Cas9 from Streptococcus pyogenes (SpCas9) recognizes an NGG PAM sequence (where N is any nucleotide), enabling targeting in roughly 1 in 8 base pairs of DNA. In the context of genome editing, the PAM requirement can limit targeting precision and manifests clearly for applications like base editing (Figure 1). To enable genome editing with greater flexibility, altering the PAM requirement of Cas enzymes through protein engineering has been one of our major interests.
To alter the PAM requirement of SpCas9, we needed a method to determine how amino acid changes to the SpCas9 protein corresponded to changes in PAM recognition. Both the accuracy and throughput of the method would be essential; the assay had to recapitulate PAM requirements in human genome editing applications and the throughput would determine the size of the mutation space we could explore. We developed the high-throughput PAM determination assay (HT-PAMDA) to permit the scalable characterization of the PAM requirements of Cas enzymes and enable protein engineering projects that were previously infeasible(1). HT-PAMDA is based on in vitro cleavage of plasmid libraries harboring a target site and flanking randomized sequence in place of the PAM by Cas9 or other Cas enzymes. The composition of PAMs in the library are determined by next generation sequencing at multiple timepoints to monitor the depletion of targetable sequences over time (Figure 2). While so-called “depletion” approaches are common for characterizing PAM requirements, HT-PAMDA is unique in its design to enable rapid and scalable characterization while reflecting enzyme performance specifically in human genome editing applications.
Precise control of in vitro reaction conditions is critical to tune depletion assays to reflect enzyme performance in a particular setting. Several approaches, including the low-throughput PAMDA predecessor to HT-PAMDA, rely on laborious protein purification to carefully control enzyme input and tune the reaction conditions(2). Gao et al. demonstrated the use transient transfection and cell lysis to produce unpurified Cas enzymes in human cell lysate, an approach amenable to higher throughput but lacking precise control of Cas enzyme concentration(3). HT-PAMDA improves on this strategy by coupling Cas enzyme expression to a fluorescent reporter and enabling fluorescence-based enzyme normalization in human cell lysates. This solution eliminated the bottleneck of obtaining concentration-normalized enzyme for characterization. Combining this in vitro workflow with a library preparation and analysis pipeline designed for sample multiplexing, HT-PAMDA enables the rapid characterization of the PAM requirements of hundreds of Cas enzymes simultaneously. With HT-PAMDA, we were able to quickly evaluate the impact of a greater number of mutations and mutation combinations in a Cas enzyme, promising a deeper understanding of enzyme function and the development of enzymes with improved properties.
In our first application of the method, we set out to leverage the scalability of HT-PAMDA to reduce the PAM requirement of SpCas9 through rational mutagenesis. We iteratively evaluated over 100 SpCas9 mutants, testing and generating hypotheses based on the assay’s information-rich readout. Characterization of the impact of single amino acid changes on PAM recognition enabled a detailed understanding of the engineering trajectories of the multiple-mutant derivative enzymes. Ultimately, we developed SpG and SpRY, variants of SpCas9 capable of targeting NGN and NRN>NYN PAMs, respectively, nearly eliminating the PAM requirement of SpCas9 and enabling single base pair resolution editing(1). The throughput of the HT-PAMDA method was critical in the development of SpG and SpRY, which contain 6 and 11 amino acid changes, respectively, relative to the wild-type SpCas9 protein. Only through exploration of this large mutation space were we able to identify these combinations of positions and mutation identities yielding enzymes with the desired properties.
A detailed description of the HT-PAMDA protocol can be found in our Nature Protocols paper(4). Beyond the procedure, we discuss experimental design considerations from modifications for different Cas enzymes to practical matters including timing and cost. Additionally, we elaborate on best practices for interpretation and visualization of the results of HT-PAMDA experiments, such as different methods to depict PAM preference (Figure 3).
We anticipate that HT-PAMDA and similar approaches will be a useful set of methods for the development of the next generation of genome editing technologies. In our description of the HT-PAMDA protocol, we describe how the approach can be adapted to characterize the properties of alternative genome editing modalities, including C-to-T and A-to-G base editors. Further, we anticipate simple modifications to the procedure will enable rapid assessment of properties beyond PAM requirements, such as enzyme specificity, or guide RNA features. Finally, in addition to protein engineering, methods to rapidly characterize critical properties of putative genome editing tools have utility in the characterization of novel Cas orthologs. Together, the scalability of HT-PAMDA will offer new insight into important properties of CRISPR enzymes, unlocking a deeper understanding of sequence-function relationships.
- Walton, R. T., Christie, K. A., Whittaker, M. N. & Kleinstiver, B. P. Unconstrained genome targeting with near-PAMless engineered CRISPR-Cas9 variants. Science 368, 290–296 (2020).
- Kleinstiver, B. P. et al. Engineered CRISPR-Cas12a variants with increased activities and improved targeting ranges for gene, epigenetic and base editing. Nat. Biotechnol. 37, 276–282 (2019).
- Gao, L. et al. Engineered Cpf1 variants with altered PAM specificities. Nat. Biotechnol. 35, 789–792 (2017).
- Walton, R. T., Hsu, J. Y., Joung, J. K., & Kleinstiver, B. P. Scalable characterization of the PAM requirements of CRISPR-Cas enzymes using HT-PAMDA. Nat. Prot. (2021).