Scientists in Switzerland have used an enhanced CRISPR-Cas gene editing technology to correct the gene mutation that causes the metabolic disorder phenylketonuria (PKU), in mice, and restore blood levels of the amino acid phenylalanine, to normal. Reporting on the technology in Nature Medicine, the researchers say their results offer “compelling evidence” that the method can be used to reverse the disease. "This approach has great potential for application in humans," claims study head Gerald Schwank, Ph.D., a professor at the ETH in Zurich.
In the reported studies the new technology enabled mRNA correction rates of up to 63%, which the authors say suggests “applicability to a large number of genetic diseases.” Their published paper is titled, “Treatment of a metabolic liver disease by in vivo genome base editing in adult mice.”
Phenylketonuria is an autosomal recessive disorder that causes a deficiency in the liver enzyme phenylalanine hydroxylase (PAH), which normally metabolizes the amino acid phenylalanine. People with PKU inherit a defective copy of the Pah gene from each parent, and the disorder means that they require a special diet to prevent build-up of phenylalanine in the body. Untreated infants can suffer from severe retardation, microcephaly, and seizures, and so some countries carry out routine screening for PKU in newborn babies.
The ETH team has now developed an approach to curing the disease that uses gene editing to correct the gene mutation. The most widely used systems for in vivo genome editing are based on the CRISPR gene editing enzyme Cas9 to introduce site-specific double-stranded DNA (dsDNA) breaks at target sites on the chromosome, which can then be repaired by the cell's homology-directed repair (HDR) mechanisms. However, the authors explain, this approach doesn't work well in very slowly dividing cells and tissues. “HDR in nondividing cells is highly inefficient …” they write. “Therapeutic application of CRISPR-associated nucleases to target genetic diseases in slowly proliferating tissues is therefore restricted to a small group of disorders for which either knockout of a gene is sufficient or the precise correction of a mutation confers a selective growth advantage to the edited cell.”
In contrast, the strategy of base editing enables genome editing without forming dsDNA breaks and without relying on HDR. Instead, base editors comprising a deaminase enzyme fused to a catalytically dead Cas9 enzyme (dCas9) can convert a C-G to T-A base pair, or vice versa. “Importantly, these deaminases are single-strand-specific,” the authors note. “We reasoned that base editors allow precise correction of disease-causing mutations in nondividing hepatocytes at rates sufficient to cure a disease phenotype.”
To test their hypothesis the researchers developed a CRISPR-Cas-associated base editing technology to correct the homozygous Pah mutation in a mouse model of PKU. The system, packaged into adeno-associated viral (AAV) vectors, effectively comprises the enzyme citidine deaminase fused to the dCas9 enzyme. The construct binds to the gene locus containing the Pah mutation and open the DNA strands. The deaminase enzyme then converts the incorrect C-G base pair in the Pah gene into the correct T-A base pair.
The team first tested the efficiency of their base editing approach in a cultured hepatocyte reporter cell line, and then in primary liver cells, before moving on to in vivo evaluation of gene correction system in adult mice, which received the AAV constructs via tail vein injections.
The treatment resulted in the animals' abnormally high blood levels of phenylalanine dropping 20-fold to normal physiological levels within six weeks. Encouragingly, post-treatment blood levels of the amino acid then remained normal for the full 26 weeks of the investigation. The growth-retarded animals also gained weight after the gene editing therapy, and their hypopigmented, paler fur also darkened to about the same shade as that of normal control animals, within weeks of treatment.
The researches reported that up to 60% of all copies of the defective gene in the mouse livers were corrected, together with Pah mRNA correction rates of up to 63% after 14 weeks. “Restoration of PAH enzyme activity (1.7–22.8% of wild-type enzyme activity) was confirmed in whole liver lysates and correlated to correction rates on mRNA and genomic DNA,” they write. There was also no evidence of any off-target effects, or of significant DNA damage. “… we found no indication of excessive DNA damage or cell proliferation after prolonged exposure to low levels of base editors,” the team states. “Our data suggest that base editors in combination with highly specific guide RNAs have a low risk for generating off-target mutations, even when expressed over longer periods of time.”
Prior methods of genome editing have had far less success at correcting target mutations in living animals, and the gene correction rate has been at best a few percent, Dr. Schwank comments. “Here we’ve achieved several fold higher editing rates—nobody has managed that so far.”
"The use of a base editor was the key to our success," explains first study author Lukas Villiger, a Ph.D. student at ETH. Prior to using the base editing technology, the ETH researchers had been working with traditional CRISPR/Cas approaches, but in 2016 switched to the base editing approach first reported by its developers at Massachusetts Institute of Technology (MIT). However, as Villiger acknowledges, "even with the new base editors, the path still didn't follow a straight line—we had to tinker around quite a bit." The researchers say that the biggest surprise was that the new system has proven to be so much more effective than the traditional CRISPR/Cas toolbox.
The ETH team is now looking for funding to carry out additional preclinical trials in other animals, including pigs. They say that follow-up studies will also be needed to test whether the AAV vector might cause any side effects, and to check more extensively for any off-target mutations. “The human liver consists of several billion cells, Dr. Schwank notes. “In none of them do we want to induce any mutations that could cause cancers.”