LONDON – Scientists in China have developed a chemical method to precisely edit single bases in DNA, and they used the technique in human embryos to correct a mutation in the beta globin gene that is a major cause of inherited blood disorder beta thalassemia.

Amongst mutations that cause beta thalassemia, HBB-28, in which an adenine base is replaced by a guanine base, is one of the most common in patients in China and Southeast Asia.

The researchers directly corrected that defect by changing the matching base on the opposite strand of DNA from cytosine to thymine. Because thymine pairs with adenine, that led to a conversion of the cytosine-guanine base pair to a thymine-adenine base pair.

The base editor used elements of a CRISPR/Cas9 genome editing construct to target the gene defect, but rather than causing a double-strand break in the DNA, it delivered an enzyme that converted cytosine to uridine. The uridine was subsequently converted to thymine by normal cellular processes.

The research was led by Puping Liang at the Laboratory of Gene Engineering at Sun Yat-Sen University in Guangzhou, whose group in 2015 published the first paper describing the use of CRISPR/Cas9 genome editing in human embryos.

The latest work, published in Protein & Cell, was prompted by problems including low efficiency, mosaicism, off-target cleavage and unintended homologous recombination that Liang encountered when attempting to use CRISPR/Cas9 to correct the beta globin gene in human embryos.

Efficient base editing at single base resolution has been reported in plant, yeast, human cells and mouse zygotes, and Liang and colleagues theorized it could provide a more efficient and specific alternative to CRISPR/Cas9 in human embryos.

Although base editing produced better results than Liang achieved with CRISPR/Cas9, the mutation was corrected in fewer than one-quarter of the cells of the embryos.

Matthew Cobb, a professor of zoology at Manchester University, suggested that the real potential of the technique was not in changing the germline of embryos but in using base editing to correct the defect in a patient's blood stem cells. "If this technique can be made more effective . . . and it was focused on blood stem cells, real progress could be made," Cobb said.

In fact, Liang and his colleagues did hone the base-editing construct over the course of their research, which began by testing the technique in a cell model of beta thalassemia. Here, the rogue guanine base was converted to adenine with 46.7 percent efficiency.

However, a shortcoming was that a second guanine base close to the one being targeted, also was converted.

The researchers moved on to attempt to base edit fibroblasts from a beta thalassemia patient. In total, 17.8 percent of cells were precisely edited, a level which they said is likely to increase expression of beta globin and provide some benefits.

However, fibroblasts do not express beta globin, and it was not possible to assess if normal expression was restored in those cells in which the mutation was corrected.

Rather than using an animal model to see if there is an effect on beta globin production, the researchers moved directly to test base editing in human embryos. Those embryos were generated by cloning methods, in which donated oocytes were enucleated and then fertilized with the nuclei of homozygous lymphocytes from the blood of a patient with beta thalassemia.

That allowed the researchers to assess if one or both mutant alleles was converted to a normal allele, and to show that unlike CRISPR/Cas9, base editing is able to correct a defect without having a normal gene as a template.

Just before transfer of the lymphocyte nuclei, the oocytes were injected with the base editor, which had been modified from the earlier experiments to correct the mutant guanine and not the adjacent guanine.

In 16.7 percent of cells of the resulting embryos both copies of the allele were corrected; in 6.3 percent one copy was corrected. As a demonstration that there is scope to improve the precision of base editing, the adjacent guanine was not converted to adenine.

Although the embryos were mosaic, the researchers said, "Collectively, this study demonstrated the feasibility of curing genetic disease in human somatic cells and embryos by the base editor system."

This is a "powerful study," said Helen O'Neill, program director for reproductive science at University College London. But she said, "It remains to be seen whether the efficiency can be improved upon." In addition, "more work is needed to assess the precision of base-editing technology, using genomewide assays."

While Darren Griffin, professor of genetics at Kent University, agreed the presence or absence of off-target effects needs more examination, he said making precise chemical alterations to DNA bases represents a "highly significant advance." But he added, "It is important not to get carried away about its widespread [clinical] utility."