Some genes make protein. This one does not, and that is exactly why it keeps slipping through the cracks. A study published this week in Genetics in Medicine pulled together 60 people from clinics and research centers around the world, all carrying disease-causing changes in a tiny gene called RNU4ATAC. Forty-two of them had never been described in the medical literature before. The picture that emerges is of a condition that is both wildly variable in how it presents and unusually easy for modern DNA testing to overlook.
RNU4ATAC does not code for a protein. It produces a small RNA that helps run the minor spliceosome, a piece of cellular machinery that trims a rare class of introns out of other genes' messages. When the gene is broken, that trimming stalls. The result is a group of overlapping syndromes with names like microcephalic osteodysplastic primordial dwarfism type 1 and Roifman syndrome, which doctors now group under the umbrella term RNU4ATAC-opathy.
One gene, many faces
The cohort makes the range plain. Most people assessed shared a core set of features: an unusually small head, short stature, skeletal differences, developmental delay, changes in the brain, skin problems, and immune deficiency. But severity swung from mild to profound. Some individuals had diabetes. Some had holoprosencephaly, a serious malformation of the developing brain. And some were missing features that the textbooks call typical, which is the kind of thing that sends a diagnosis sideways for years.
Across the group the researchers cataloged 33 distinct variants in the gene. Thirteen of those had not been reported before. That is a lot of genetic diversity for a gene only about 130 nucleotides long, and it helps explain why two children with the same diagnosis can look so different in clinic.
When the DNA test isn't the problem, the reading is
Here is the part that should matter to any lab running genetic tests. RNU4ATAC variants are usually captured by clinical exome sequencing. The bases are right there in the data. The trouble is that analysis pipelines are built around protein-coding genes, and a noncoding gene like this one often gets filtered out or simply never examined. The disease-causing change sits in the file, unread.
To sharpen the diagnosis, the team turned to RNA sequencing, which was available for seven of the affected individuals. Every one of them showed the same telltale signature: minor introns that should have been spliced out were being retained instead. That is the molecular fingerprint of a stalled minor spliceosome, and it lines up with what a broken RNU4ATAC is expected to do. In six people, the RNA evidence was strong enough to reclassify variants that had been labeled "of uncertain significance," bumping them up to likely pathogenic. For a family, that shift can be the difference between an open question and an answer.
What the study leaves open
This is a descriptive cohort, not a treatment trial. It does not offer a therapy, and it does not tell us how common RNU4ATAC-opathy actually is in the general population, since the group was assembled from people already flagged as affected. RNA sequencing was done for only a handful of participants, so its diagnostic value here rests on a small sample. And with 42 new cases added at once, the reported spectrum of the condition will almost certainly keep shifting as more people are found.
Still, the practical message is hard to argue with. The authors urge labs to make sure RNU4ATAC and other noncoding genes are actually assessed by their analysis software, not just sequenced and set aside. For a child whose small head and delayed milestones have gone unexplained through round after round of testing, the diagnosis may have been sitting in the data the whole time. Someone just needed to look in the right place.
Comments