ABOVE: Ribbon and surface representation models of the phenylalanine hydroxylase enzyme

In a study published yesterday (August 5) in Scienceresearchers identify human and mouse long noncoding RNAs, whose loss mimics the metabolic disorder phenylketonuria (PKU) in human cells and mice. Injecting the human long noncoding RNA into a mouse model of the disease improved the animals’ symptoms, potentially pointing to new opportunities to develop treatments for human patients.

Scientists and clinicians have historically categorized PKU as a problem with the gene that encodes the enzyme phenylalanine hydroxylase (PAH), which converts the amino acid phenylalanine into tyrosine. Without this enzyme, phenylalanine consumed as part of a normal diet in foods such as milk, eggs, cheese, eggs, and meat builds up, causing seizures, intellectual disabilities, and psychiatric disorders. But DNA sequencing has demonstrated that sometimes people with PKU don’t have a mutation in the gene for PAH, calling into question PKU’s status as a single-gene disorder. The new study suggests the depletion of certain noncoding RNAs could also cause the condition, and that boosting the levels of those RNAs could relieve symptoms.

“This just shows that even after such a long time of studying what we thought was simple disease, we’re still learning more and more about other factors than just the mutations in this specific gene,” says John Mitchell, a pediatric endocrinologist at McGill University Health Center in Canada who was not involved in the study. “We haven’t been able to really describe the genotype-phenotype relationship for this particular disease,” he adds. “Having new areas to pursue for treatment is going to be important.”

See “The RNA Age: A Primer

In the new paper, University of Texas MD Anderson Cancer Center’s Liuqing Yang and colleagues write that nearly all mutations in the human genome happen in noncoding DNA. Based on this idea and on work showing that PKU sometimes occurs in the absence of mutations to the gene for PAH, they sought to identify potentially relevant long noncoding RNAs by searching for ones with differing expression levels in mouse livers—the main site of phenylalanine metabolism—at two points in time: on embryonic day 18.5, when PAH activity has not yet reached its peak, and in adult mice.

Yang and his collaborators identified two versions of a noncoding RNA that were upregulated in adult mouse livers, suggesting they may play a role in phenylalanine metabolism. The researchers then introduced a mutation to disrupt the first splice site in the noncoding RNA, which depleted both forms. The mice with reduced noncoding RNA were slow growing, had higher levels of phenylalanine in their serum, and displayed diminished PAH activity, similar to people with PKU. Furthermore, as these mice aged, they experienced seizures and early death. The team used an antibody to isolate PAH and determined that the noncoding RNA they found is associated with the enzyme, so they named it Pair (PAH-activating long noncoding RNA).

Next, they looked in donated human liver tissue and found that human PAH also has a long noncoding RNA—called HULC (hepatocellular carcinoma upregulated long noncoding RNA)—associated with it. In a cell-free system, the authors showed that both Pair and HULC interact with PAH and likely tweak its shape to enable the enzyme to better bind to phenylalanine. In human hepatocytes missing HULC, adding the long noncoding RNA back improved the ability of PAH to convert phenylalanine to tyrosine, suggesting insufficient levels of the RNA may play a role in the pathogenesis of PKU.

In an email to The Scientist, Nenad Blau, a consultant in biochemical genetics at University Children’s Hospital Zurich in Switzerland and coauthor on the paper, notes that HULC deficiencies have yet to be detected in patients with abnormally high phenylalanine levelsStill, he says the work highlights that “PAH genotyping is essential for the diagnosis and management” of such disorders.

David Dimmock, senior medical director of Rady Children’s Institute for Genomic Medicine in California who did not participate in the work, says the RNA results are “not something that I intuitively would have expected,” but that more research is needed. “It will be fascinating to see whether or not we can find changes in this long noncoding RNA or others that actually do explain some of the phenotypic variability” seen in PKU in people, he says. “The next logical question is, if there is variability in the long noncoding RNAs, is there therapeutic potential?”

See “New Treatments for Phenylketonuria Aim to Loosen Reins on Strict Diet

In an effort to address that sort of question, Yang, Blau and their colleagues delivered a liver-targeted HULC mimic to mice with the PAH mutation seen in the majority of PKU patients in the US. Control mice that received an injection of a scrambled RNA had high serum levels of phenylalanine, but treated mice had comparatively low levels of phenylalanine and high levels of tyrosine. Then the researchers gave HULC mimic alongside an established drug—a synthetic version of a protein known to support PAH function—that works in some patients with mild cases of PAH deficiency, but has been shown before to be ineffective in this mouse model and in patients with full blown PKU. The combined treatment had an additive effect in the mice, lowering serum phenylalanine levels even more than HULC mimic alone. According to Blau, the results indicate that HULC mimics may be a viable treatment option for some patients with PKU.

“It’s certainly a tantalizing thought that maybe this would be a way of ramping up endogenous activity in individuals with certain genetic changes,” Dimmock says of the experiments. However, before applying a related strategy therapeutically in people, he notes that research would need to address whether it’s possible to manipulate one enzyme without having negative effects that would harm the patient. And while Yang, Blau, and their colleagues have identified these RNA candidates that appear to function in similar ways, he adds that another challenge with long noncoding RNAs “seems to be how do you jump the species gap, because your model systems may or may not actually be accurate mimics of what’s going to happen in a human.”