Genome editing is usually a two-step process. First, stuff cells with DNA that encodes for genome-editing proteins. Second, let the DNA insert itself into the cell’s DNA and begin expressing the genome-editing proteins, which then, finally, begin altering the genome.

The process is fraught with difficulties, particularly in the first step. Many DNA delivery strategies cannot be used in animal or human patients. One such strategy, the use of viral vectors to inject DNA into cells, can raise complicating long-term safety issues.

It may be possible, however, to skip the first step, say researchers at Harvard University. According to these researchers, common cationic lipid nucleic acid transfection reagents can potently deliver proteins that are fused to negatively supercharged proteins. The researchers found that this approach worked when they used proteins that contain natural anionic domains or that natively bind to anionic nucleic acids.

This result appeared October 30 in Nature Biotechnology, in an article entitled, “Cationic lipid-mediated delivery of proteins enables efficient protein-based genome editing in vitro and in vivo.”

“This approach mediates the potent delivery of nM concentrations of Cre recombinase, TALE- and Cas9-based transcription activators, and Cas9:sgRNA nuclease complexes into cultured human cells in media containing 10% serum,” wrote the authors. “Delivery of unmodified Cas9:sgRNA complexes resulted in up to 80% genome modification with substantially higher specificity compared to DNA transfection.”

The researchers were led by Professor of Chemistry and Chemical Biology David Liu, Ph.D. Dr. Liu’s group, which included Drs. John Zuris and David Thompson, developed a system that uses commercially available molecules called cationic lipids—essentially long, greasy molecules that carry a positive charge at one end—to efficiently introduce genome-editing proteins into cells. The researchers even demonstrated that the technology can be used to modify genes in living animals.

Working with Zheng-Yi Chen, an Associate Professor of Otology and Laryngology at Harvard Medical School and researcher at Massachusetts Eye and Ear Infirmary, Dr. Liu and colleagues used the newly developed system to modify genes in specialized “hair cells” in the inner ear of mice. Hair cell damage, either from environmental or genetic factors, is a common cause of hearing loss.

“We had the very simple idea to use the same commercially available cationic lipids researchers use to deliver DNA and RNA to deliver proteins. But instead of using super-positively charged proteins, we use super-negatively charged proteins, which resemble nucleic acids in their highly negatively charged state.” Dr. Liu said. “The potency of delivering proteins that are associated with highly negatively charged molecules using cationic lipids is approximately 1,000 times greater than delivering proteins using positively charged proteins or peptides.”

Importantly, the team’s experiments showed that the new system, when applied to the delivery of genome-editing proteins, results in target gene modification that is at least as efficient as the best results they observed from the delivery of DNA encoding genome-editing proteins. But Dr. Liu and coworkers showed that the specificity of genome editing—how accurately the targeted genes are modified versus modification of other sites in the human genome—was much higher from protein delivery instead of DNA delivery.

This outcome was what the researchers hoped to see. “Following DNA delivery, the encoded proteins can be expressed in difficult-to-regulate amounts for long periods of time,” Liu said. “There has always been a mismatch between DNA delivery and the desired outcome of genome editing. In genome editing, the mission is to fix one or two copies of a gene. After a genome-editing protein finishes that mission, you want it to go away, because the only things it can do after that point are undesired and possibly harmful.

“So protein delivery, which is transient and short-lived, seemed to be a better match than DNA delivery for most genome-editing applications.”