Skip to main content
Premium Trial:

Request an Annual Quote

Yale-Led Gene Editing Study Uses PNAs to Correct Hemoglobin Gene in Human CD34 Cells, Mice


NEW YORK (GenomeWeb) – A new study led by researchers at Yale University reports gene editing using a technology that leverages a cell's endogenous DNA repair machinery. Delivered by sugar polymer nanoparticles in an intravenous injection, these peptide nucleic acids could form the basis for gene therapy that doesn't rely on a foreign nuclease.

Peptide nucleic acids (PNAs) are a "funny fusion" of a molecule, Yale Professor Peter Glazer told GenomeWeb. Not quite a protein, not quite an oligonucleotide, they're a hybrid with a polyamide backbone but legs made of nucleotide bases. But they're turning out to be extremely useful in a number of applications. PNAs bind extremely tightly to nucleic acids, making them useful for applications like silencing micro RNAs or as an RNA clamp in PCR. They can even bind DNA via Hoogsteen base pairing to create a triple helix.

Now, scientists led by Glazer and co-first authors Raman Bahal and Nicole McNeer, also at Yale, have demonstrated the utility of PNAs in gene editing. Joined by collaborators at Carnegie Mellon University and the University of Massachusetts Medical School, the researchers used specially designed PNAs to mediate editing of the hemoglobin gene in both mouse and human cells.

These PNAs contain a tail that allows them to bind DNA twice, first via strand invasion and then again to form a triple helix that contorts the genome and triggers endogenous repair pathways. The cell's own nucleases open up the target site for repair, which can be aided with an oligo template to correct the mutation via homology-directed repair.

In a mouse model of beta thalassemia driven by a splicing defect, the researchers reported editing efficiencies of about 7 percent. "We essentially cured the mice of anemia by a simple IV infusion," Glazer said. "When we submitted our manuscript revisions, we had followed those mice for 140 days after treatment. They still had elevated blood hemoglobins in the normal range for a mouse." In human primary hematopoietic stem cells edited ex vivo, the team was able to reach efficiencies of about 5 percent, approaching theoretical levels of clinical utility.

Glazer and his colleagues published their results today in Nature Communications.

It's not the first use of PNAs to edit clinically relevant gene mutations — Glazer has also published a strategy to edit the cystic fibrosis mutation — but the paper is the first to suggest it could be a promising platform for gene therapy. The paper showcases several technological leaps that greatly improve the prospects of PNAs in gene editing, Glazer said.

He's been working on triple helix-driven editing since the 1990s. "At first the efficiency of the system was really low," Glazer said. Editing happened, but barely above the level of detection. "With regular DNA, the ability to get into the cell and bind the target was modest." Beyond that, target needed to be situated in between long runs of purines and pyramidines to form a triple helix. Those limitations held back the technology for 10 to 15 years, but when he began using PNAs with their polyamide backbone, the work started to take off. "That difference in backbone confers the ability to bind DNA much better than DNA binds to itself," he said.

Not only can PNA bind DNA, it can bind a PNA-DNA double helix to form a triple helix. Substituting a side chain at the gamma position of the PNA backbone helps boost triplex formation by forcing the PNA into a helical position to start out with, like a cord for a phone handset.

If the gamma-PNA clamps around a single strand of genomic DNA, it strongly provokes repair, eventually leading to the double strand break needed for genome editing. "Once you've made it that far, it falls into the same thing that CRISPR does," Glazer said.

While the Nature Communications paper doesn't delve into a head-to-head comparison with CRISPR/Cas9- or zinc finger nuclease-based genome editing, it naturally invites it. That's according to both Glazer and Neville Sanjana of the New York Genome Center, who has worked extensively with CRISPR/Cas9 and has also co-authored a study on editing the fetal globin gene.

"I think an important comparison is the recent beta-globin correction in hematopoietic progenitors using CRISPR," led by Jacob Corn of the University of California, Berkeley and Dana Carroll at the University of Utah, Sanjana said in an email. "This was just published a few weeks ago and goes after a similar goal. Notably, both techniques seem to achieve similar amounts of gene correction. It's fascinating that both groups get approximately the same editing efficiency. Overall, I think it is great to see different approaches to therapeutic gene editing." Sanjana was not involved in either study.

Corn and Carroll are just the latest researchers to report Cas9-based editing strategies for hemoglobin gene mutations. Prior to that, reports from groups at Dana Farber Cancer Institute, St. Jude Children's Research Hospital, and UCSF have also described using CRISPR to edit hemoglobinopathy-inducing mutations.

And Sangamo Biosciences is conducting preclinical studies for hemoglobinopathies using zinc-finger nucleases.

But Glazer's technology appears to be close to something not even the most advanced CRISPR-based methods can sniff at: the ability to edit blood cells in vivo. To deliver the PNAs to cells, Glazer enlisted Mark Saltzman, also at Yale, who has developed a sugar polymer nanoparticle called poly(lactic-co-glycolic acid) (PLGA). It's already used to deliver the US Food and Drug Administration-approved prostate cancer drug Lupron (leuprolide acetate for depot suspension).

Packaging the PNAs and DNA oligo donor molecules in these PLGAs leads to "gentle and effective delivery into mouse and human cells," Glazer said. Previous work had shown PLGAs could deliver PNAs into primary blood stem cells with no detectable toxicity.

It's the combination of efficiency and deliverability that make gamma PNAs especially interesting to Glazer. Not only can they edit cells in culture, but they can be injected into the whole mouse intravenously. "There's no rigmarole to take out bone marrow," Glazer said.

In one experiment reported in the new paper, the researchers injected mice four times over eight days. "That was sufficient to achieve gene editing of the beta globin gene to completely reverse the anemia." The injections led to gene editing at the target mutation in about 7 percent of primary blood stem cells. The researchers also treated mouse cells in culture, reaching efficiencies of up to 15 percent.

In another experiment, the researchers used PNAs to edit human CD34 cells from healthy donors. "We treated those in culture and got about 5 percent gene editing [efficiency] in a single treatment," Glazer said, a level that could be clinically meaningful.

And because PNAs bind so strongly to DNA, they showed few off-target effects. "When we looked at sites somewhat similar to the target site, we saw editing that was 100,000-fold lower than the frequency of editing at the target site."

While doing an unbiased analysis of off-target editing is a high bar to clear, it's the next step to really compare PNAs to other genome editing technologies, Sanjana said.

Moreover, the level of editing efficiency needed to diminish or erase disease still needs to be established. "Although both [the Glazer-led and Corn-led] studies are impressive, it is unclear if these gene repair rates are high enough for a long-term therapeutic effect," Sanjana said.

A major drawback to the technology is the cost of reagents. "It's one of the things that has limited adoption of PNAs," Glazer said. A single mouse experiment could cost as much as $5,000, he estimated. Generic PNAs are made by protein synthesis and are sold by several research products companies. But the gamma PNAs are custom made in collaboration with Danith Ly of Carnegie Mellon and not commercially available.

But if they were to be taken into clinical research and production properly scaled by a pharmaceutical company, the cost could be similar to small interfering RNAs or antisense RNAs, he said.

Glazer and four co-authors have applied for a patent related to the study, but declined to disclose information on the commercialization of that patent or any other PNA-related patents, which Glazer also holds.

"I hope our work will lead to more interest," Glazer said. He's already begun working on using PNAs to edit the sickle cell mutation in mice and any single-gene genetic disease driven by a point mutation could be a viable target.

"In many ways it overlaps in where you might go with CRISPR," he said.