November 15, 2017 (Vol. 37, No. 20)
Tool Suppliers Are Developing Arrayed Libraries and Synthetic gRNAs for More Precise, Targeted Genome Manipulations
CRISPR, a tool advancing the field of functional genomics, changed the game with knockouts instead of knockdown, and facilitated more robust phenotyping mapping. As the technology continues to mature, tool suppliers are pulling their weight and offering variations of new guide RNAs (gRNAs) and libraries.
Many of these new tools were discussed at Global Technology Community’s 7th Genomics and Big Data Summit in Coronado, CA, in September 2017 and at the CRISPR & Precision Genome Editing conference in Berlin in November 2017.
Detecting gene-editing outcomes is of paramount importance. Next-generation sequencing (NGS) is one method; NGS provides all the information needed to know what is happening at a gene locus. But preparation of libraries—as well as genome analysis—take time and expertise.
Another approach, droplet digital PCR (ddPCR), provides very sensitive and precise quantitation of homology-directed repair (HDR), point mutagenesis, and nonhomologous end-joining (NHEJ)—without the need for markers or artificial reporters. The method is relatively quick and inexpensive compared with NGS.
ddPCR and NGS are not mutually exclusive; ddPCR is an alternative for quick turnarounds, especially if a laboratory is focused on a few genes and designing gene-specific probes for deep probing.
Chromosomal double-strand breaks trigger DNA repair, and restoration can occur via two cellular pathways: HDR and NHEJ. Although CRISPR/Cas9 can induce both HDR and NHEJ, NHEJ predominates. HDR is based on recombination between the genomic DNA and the template DNA, and can induce very precise single base-pair resolution editing. And while NHEJ-mediated repair of Cas9-generated breaks is generally effective, it is also prone to errors.1 CRISPR could be used in more diverse ways if HDR could be induced more efficiently, perhaps, as studies have shown, through the knockout of NHEJ-related factors.2
Yuichiro (Ichi) Miyaoka, Ph.D., project leader, Regenerative Medicine Program, Tokyo Metropolitan Institute of Medical Science, was the first to use ddPCR to detect CRISPR genome-editing outcomes. His current research focuses on finding the best conditions for HDR induction. ddPCR is used to systematically compare different germinating conditions for inducing NHEJ and HDR activity, as it can simultaneously detect both events. Many other detection methods are specific for either HDR or NHEJ.
New Arrayed Libraries
Short interfering RNA (siRNA), used for years for functional genomic screening, has helped to establish best practices paving the way for the CRISPR paradigm.
The permanent genome-level edit in CRISPR/Cas9 has its own unique challenges when compared to the transient silencing in RNA interference (RNAi). The same target may be altered in nearby genomic regions, resulting in unintentional silencing. As opposed to RNAi off-targets, unintentional silencing of one gene sequence may occur, but that gene may be antisense to a noncoding region with regulatory activity, or the two DNA strands may have different roles or the edit may disrupt an intronic noncoding region.
Arrayed CRISPR library offerings are relatively new and allow a higher level of interrogation of the biology compared with the limitations of plus/minus in a lentiviral pooled screen.
Arrayed libraries of CRISPR RNA (crRNA), such as Edit-R crRNA libraries (Dharmacon—A Horizon Discovery Company), cover the whole human genome—18,000 genes—and are also offered in predefined catalog collections or custom cherry-pick builds. The approach incorporates chemically modified synthetic guide RNAs (gRNAs) that provide a two-part guide RNA system similar to the endogenous CRISPR/Cas9 mechanism. The two-part gRNA is more practical and cost-effective for high-throughput chemical synthesis, so thousands can be made in a week—and scale up to larger amounts is much more straightforward.
Four crRNAs per gene offer high statistical power and a straightforward hit-ranking system. The crRNAs can also be pooled into a single reagent for screening applications. A validated gRNA selection algorithm ensures functionality and specificity (Figure 1).
Best practices for CRISPR screening include reconfirming hits following the primary screen by repeating the experiment on a smaller scale to highlight both weak and strong hits, and using different gRNAs per gene to determine if similar phenotypes result. Next is ensuring that the phenotype correlates with the genomic-level alteration, typically tested with a PCR-based mismatch detection assay.
“We also recommend going back to RNAi to see how well gene silencing corresponds to gene knockout. Although the phenotype may be slightly different, if RNAi results agree with CRISPR results, it provides confirmation that the results are due to loss of function,” explains Louise Baskin, senior product manager at Dharmacon.
The company is also working on synthetic and lentiviral tools for CRISPR activation (CRISPRa), the activation of transcription by gRNAs that target sequences near the transcriptional start site.
Lentivirus can transduce many relevant cell lines including nondividing cells. The Invitrogen LentiArray CRISPR Libraries also target the whole human genome (Figure 2). The library consists of Cas9 and gRNA particles along with positive and negative controls. Gene sets or custom picks are available as well.
“gRNAs are designed to primarily 5 prime coding exons of a target gene using our CRISPR gRNA design tool in order to maximize knockout efficiency and to minimize off-target effects. We have four highly ranking guides per gene; cleavage efficiency averages around 60% or greater for a more robust knockout and reproducible phenotype,” discussed Jonathan Chesnut, Ph.D., senior director, synthetic biology R&D, Thermo Fisher Scientific.
“Lentiviral titers are guaranteed to at least 1 x 106 TU/mL, and are tested using an antibiotic resistance assay to count actual infectious particles.”
Internal experiments, with a CellSensor (Thermo Fisher Scientific) line containing a reporter system driven by an NF-?B transcription factor response element, looked at the effects of knocking out kinases known to be active in that pathway and showed very low assay-to-assay variability.3 In another arrayed full-kinome screen, hits from a CRISPR and an siRNA knockdown screen were compared.3 siRNA picked up the main players; however, the CRISPR screen identified a dozen other known involved genes.
Simone Treiger Sredni, M.D., Ph.D., associate professor, Pediatric Neurosurgery, Northwestern University Feinberg School of Medicine, Ann and Robert H. Lurie Children’s Hospital of Chicago, used a subset of the human kinome library for her work on malignant rhabdoid tumors, deadly embryonal tumors that occur in infants and toddlers. The tumors are associated with poor survival outcomes and typically don’t respond well to currently available therapies.4
Dr. Sredni and colleagues mutated 160 kinases in a well-characterized rhabdoid tumor cell line using the kinome library subset, and then the investigators measured the clones’ growth rates measured over a period of months. In 2–3% of the cell lines, growth rate slowed significantly. Further work determined that the mutation of Polo-like kinase 4 (PLK4) by CRISPR/Cas9 produced a higher impairment of cell proliferation, opening up the door for possible treatment options.
Synthetic Single-Guide RNAs
More research is needed to fully understand how toxic CRISPR components can be to living cells, and the type of immune responses they may elicit in the different therapeutic approaches—such as human in vivo, human ex vivo, human germline, or antiviral- and antimicrobial-based therapies. In addition, the likelihood of off-target effects needs to be thoroughly assessed; some off-targets may be averted by using transient CRISPR therapeutics.
As CRISPR evolves, gRNAs have evolved. Two-piece complexes consisting of crRNA and trans-activating crRNA (tracrRNA) are based on the native bacterial system. Single guide RNAs (sgRNAs)—hybrid, single-molecule guides developed several years ago—have also been shown to be accurate, reproducible, and stable.
Before the availability of synthetic sgRNAs, many labs produced CRISPR guides using an approach known as in vitro transcription (IVT), whereby enzymes are used to convert a DNA template into an RNA molecule. This time-consuming approach yields highly variable products requiring significant purification. Recently, new technological advancements have allowed for the synthesis of full sgRNAs (Figure 3).
“A significant advantage that the synthetic sgRNAs have over the IVT-derived sgRNAs is that they can be chemically modified to provide stability and protection to the guide. This has enabled extremely high editing efficiencies in primary T cells and stem cells, where unprotected IVT-derived guides typically fail. Now available at a practical scale, price, and turnaround time, synthetic sgRNA is likely to become the de facto standard for CRISPR gRNAs,” said Kevin Holden, Ph.D., head of synthetic biology, Synthego.
The first clinical trial utilizing CRISPR, aimed at programming immune cells to target and kill lung cancer tumor cells, began in China in late 2016. Currently, most CRISPR trials are in the preclinical phase. Given that many of these therapeutic approaches will rely on a transient CRISPR system, it is highly likely that they will incorporate either purified Cas9 nuclease, or Cas9 mRNA and a synthetically generated sgRNA.
Standardizing Custom Library Design
Often, mixing disciplines can have a beneficial effect. Oxford Genetics was founded to bring a level of standardization to genetic engineering, an approach typically seen in the engineering sciences. The use of automation and scalable processes increase predictability and robustness while lowering cost and human error.
According to Ryan Cawood, Ph.D., chief executive officer, Oxford Genetics, the goal is to minimize steps in complex library design to improve turnaround time. Individual pieces of the CRISPR process are locked down, validated individually (to ensure that they work independently), and then validated in combination with other pieces.
Libraries are built using high-throughput automation into a modular plug-and-play background, like a car chassis, and can be modified in a number of ways, such as docking sites, expression systems, viral vectors, etc., providing extra flexibility if the research direction changes.
Custom pooled and arrayed CRISPR libraries are are developed in client-led efforts. Pooled libraries can contain up to 50,000 gRNAs in a tube, and arrayed libraries can reach 1000 gRNAs in one well of a 96-well plate. Custom CRISPR screening services are expected to be automated in the next six months.
The company focuses its R&D on adding economic value to the pharmaceutical development workflow: discovery, development, manufacturing, and delivery. One avenue is using CRISPR to research and reengineer viruses for manufacturing use. One example is a reengineered virus that goes through its life cycle repressing itself from producing its own proteins and redirecting itself toward producing the protein of interest. Other foci include using CRISPR knockouts to facilitate adeno-associated virus and lentivirus production.
Although the company is not involved in therapeutic development, Dr. Cawood believes that CRISPR will play a big part in the future of regenerative medicine. CRISPR, he says, will kick-start stem cells, reawaken the growth potential that exists within the body, and facilitate genetically engineered replacement organs.
1. C.A. Waters et al., “The Fidelity of the Ligation Step Determines How Ends Are Resolved during Nonhomologous End-Joining,” Nat Commun. (5), 4286 (July 3, 2014).
2. L. Zhu et al., “CRISPR/Cas9-Mediated Knockout of Factors in Nonhomologous End-Joining Pathway Enhances Gene Targeting in Silkworm Cells,” Sci. Rep. 5, article number: 18103 (2015), doi:10.1038/srep18103.
3. J. Chesnut, “Unraveling Biology and Identifying Targets with Functional Genomics Approaches Supported by LentiArray CRISPR Library Screens,” presentation at the 7th Genomics & Big Data Summit 2017 in San Diego, CA (September 27, 2017).
4. S.T. Sredni et al., “A Functional Screening of the Kinome Identifies the Polo-Like Kinase 4 as a Potential Therapeutic Target for Malignant Rhabdoid Tumors, and Possibly, other Embryonal Tumors of the Brain,” Pediatr. Blood Cancer, doi: 10.1002/pbc.26551.