Nuchal translucency appears on ultrasound as a dark, anechoic space between the skin of the fetal neck and the soft tissues beneath it. It is a fleeting feature of normal development: between roughly 11 and 14 weeks of gestation, the lymphatic vessels of the neck gradually forge connections with the jugular venous system, and before those channels are fully open, a small amount of lymphatic fluid pools in the cervical region. In most fetuses, the accumulation drains away after week 14. But when the measured thickness exceeds 2.5 millimeters, clinicians treat it as a red flag for adverse outcomes. Increased nuchal translucency has been linked to a sweeping spectrum of problems, from chromosomal abnormalities and structural birth defects to single-gene syndromes such as Noonan syndrome, skeletal dysplasias, and an assortment of neurodevelopmental disorders. Until now, however, Pitt-Hopkins syndrome was nowhere on that list.

Pitt-Hopkins syndrome is a rare and devastating condition caused, in the overwhelming majority of cases, by mutations in TCF4, a gene that encodes transcription factor 4—a DNA-binding protein that switches other genes on and off during development. The disorder follows an autosomal dominant pattern, meaning a single faulty copy is enough to cause disease, and it announces itself after birth with profound intellectual disability, distinctive craniofacial abnormalities, characteristic facial features, and generalized low muscle tone. Some affected children develop epilepsy, myopia, and episodes of deep, rapid hyperventilation. Several hundred individuals worldwide have now received the diagnosis, carrying a wide range of TCF4 defects: missense and nonsense changes, small insertions and deletions that scramble the reading frame, larger structural rearrangements, and splicing variants. Crucially, because the syndrome is usually recognized in children rather than fetuses, increased nuchal translucency had never been reported in connection with it—until this case.

The story began in March 2023, when a woman who had undergone a frozen embryo transfer at Zhejiang Provincial People’s Hospital became pregnant with dichorionic diamniotic twins, each developing in a separate placenta and a separate amniotic sac. First-trimester ultrasound told two very different stories. Fetus A measured 1.9 millimeters of nuchal translucency, comfortably within the normal range, and showed no other prenatal abnormalities. Fetus B measured 3.5 millimeters, well past the 2.5-millimeter warning threshold. With no family history to guide them and both parents apparently healthy, the clinical team collected amniotic fluid from fetus B along with blood samples from both mother and father, then performed trio-based clinical exome sequencing to compare all three genomes at once. The study was approved by the hospital’s medical ethics committee, and the parents provided written informed consent.

Clinical exome sequencing is a targeted deep-sequencing strategy that homes in on the protein-coding regions of the genome—the fraction of DNA that harbors most known disease-causing variants—along with 20 base pairs of flanking intronic sequence on each side of every exon. In this case, the assay covered 7,099 clinically relevant human genes. The team fragmented and repaired fetal DNA, ligated sequencing adapters, captured the target regions with a clinical exome kit, and sequenced the libraries on an Illumina NextSeq 550DX platform, aligning the resulting reads to the human reference genome hg19 (GRCh37) with the Burrows-Wheeler Aligner. Candidate variants were then filtered against public databases including dbSNP, ExAC, HGMD, ClinVar, OMIM, and gnomAD. Copy-number analysis, performed both on the exome data and with a dedicated CNV-seq assay, came back clean, ruling out large duplications and deletions. What remained was a single striking finding: a heterozygous variant, c.1146+3A>T, lodged in intron 14 of the TCF4 gene (transcript NM_001083962.1)—present in the fetus and in neither parent. Sanger sequencing confirmed the call: the mutation had arisen de novo, in the fetus alone.

A single-letter change three base pairs into an intron might seem trivial, but in splicing, position is everything. The variant disrupts the donor splice site—the molecular signal that tells the cell’s splicing machinery where an exon ends and the intron begins. To prove the point, the researchers built a minigene assay. They amplified a 1,404-base-pair segment of the relevant TCF4 region from the fetus’s amniotic fluid DNA and, in parallel, from a normal control, then inserted each fragment into a pcMINI reporter plasmid using the restriction enzymes KpnI and NotI. The two constructs—one wild-type, one mutant—were introduced into HEK-293T kidney cells and HeLa cervical cancer cells with Lipofectamine 3000, and 48 hours later the cells’ RNA was extracted, reverse-transcribed, and read out. The difference was unmistakable. Cells receiving the wild-type construct produced a transcript with exon 14 neatly spliced into place. Cells receiving the mutant construct produced a shorter product in which exon 14 had been skipped entirely, stitched directly from the preceding exon to the following one.

The consequences cascade from there. Losing exon 14 deletes 77 nucleotides of coding sequence, shifting the reading frame and generating a premature stop codon in exon 15—a change annotated p.Gly358Lysfs*4. The result is a truncated TCF4 protein of just 360 amino acids, stripped of its C-terminal region, including the conserved basic helix–loop–helix domain the protein needs to grip DNA and dimerize with partner proteins. That domain, encoded across exons 9 through 19, is precisely the region where Pitt-Hopkins mutations do their worst damage, wrecking protein stability, DNA binding, and transcriptional activation. Applying the joint interpretation guidelines of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology, the team classified the variant as pathogenic, stacking five independent lines of evidence: a different nucleotide substitution at the same donor splice site had already been reported as a de novo variant in a Pitt-Hopkins patient and judged likely pathogenic; the variant arose de novo with both parents confirmed as non-carriers; it is absent from the gnomAD population database; three separate splicing-prediction tools flagged the disruption of the highly conserved donor site of exon 14, a non-phase exon; and the minigene experiment delivered direct functional proof of exon skipping.

The case slots into a wider pattern that prenatal specialists have been assembling. In a cohort study of 503 twin pregnancies, increased nuchal translucency in at least one fetus was associated with a significantly higher risk of adverse outcomes than when both fetuses measured normally. More sobering still, among 1,658 fetuses with increased translucency followed through 1,309 live births, roughly one percent of infants still showed neurodevelopmental delay or abnormalities even after chromosomal abnormalities and pathogenic copy-number variants had been excluded—hinting that hidden single-nucleotide variants account for a stubborn residual risk. Recent reports have been closing that gap one gene at a time: a de novo SMAD4 variant in a fetus whose measurement reached 5.7 millimeters, later explained as multi-system Myhre syndrome, and compound heterozygous NUP107 variants in two fetuses with increased translucency who later showed renal anomalies and neurodevelopmental impairment. The new TCF4 finding extends that expanding map. Notably, a change at this very splice site, c.1146+3A>G, was documented in an affected child back in 2008, and across the nine previously reported patients carrying TCF4 splicing variants, every case displayed the core features of the syndrome: global developmental delay, severe language impairment, hypotonia, facial dysmorphism, and intellectual disability.

The discovery also underscores how far TCF4’s influence reaches beyond the developing brain. The gene is broadly expressed and participates in multiple developmental pathways; an entirely different TCF4 lesion—a CTG trinucleotide repeat expansion known as CTG18.1 lodged within one of its introns—causes Fuchs endothelial corneal dystrophy, a common cause of progressive vision loss. Many of Pitt-Hopkins’ apparently non-neurological features, including abnormal breathing patterns and constipation, may in fact reflect autonomic nervous system dysfunction rather than isolated defects of unrelated organs. And the syndrome’s presentation can be startlingly variable: one family carrying a missense change in exon 18 (c.1849G>A, resulting in p.Val617Ile) showed only mild, nonspecific neurodevelopmental impairment rather than the full classical picture. For clinicians and genetic counselors, the message is that TCF4-related disease is not a monolithic entity but a spectrum—one more reason any prenatal clue, however subtle, deserves attention.

The twins’ story did not end in the sequencing lab. When the team followed up during the children’s first and second years of life, fetus A—now a toddler—was healthy and developing normally, exactly as predicted. Fetus B, by contrast, showed widening of the brain’s lateral ventricles along with delayed growth and development, hallmark features of Pitt-Hopkins syndrome that strengthened the prenatal suspicion. Because the disorder typically unfolds across childhood, the researchers expect additional symptoms to emerge over time and plan close monitoring of the child. Taken together—the de novo occurrence of the variant, the minigene proof of exon 14 skipping, the truncated protein shorn of its DNA-binding domain, and the postnatal neurological signs—the evidence points firmly to c.1146+3A>T as a genuine, pathogenic Pitt-Hopkins mutation. And it elevates a routine ultrasound measurement into something far more consequential: a potential early intrauterine sign of TCF4-related disease. For any fetus with unexplained increased nuchal translucency, the researchers conclude, Pitt-Hopkins syndrome and related TCF4 disorders should now enter the differential diagnosis.

Clinically, DCM features markedly enlarged heart chambers and impaired pumping, whereas LVNC is defined by spongy, honeycomb-like myocardium. Until now, patients carrying RBM20 variants have largely received standard heart-failure care, despite the possibility that this molecular subgroup might respond to more tailored therapies.

The researchers examined two families with RBM20 mutations that differ by just a single protein “letter” at the same critical site. In the DCM family, an arginine is replaced by tryptophan; in the LVNC family, the same arginine is replaced by leucine. The subtlety of the difference makes the contrasting phenotypes especially striking.

To explain how one residue shift could remodel cardiac physiology, the team probed calcium handling in patient-derived material. They found distinct calcium-system failures: LVNC models showed heightened calcium-cycle activity and an energetically costly calcium regime, while DCM models behaved as if internal calcium stores leak, disrupting rhythmic contraction–relaxation coupling.

The study used pluripotent stem cell technology to recreate disease-relevant heart muscle in vitro. Induced pluripotent stem cells (iPSCs) were differentiated into beating cardiomyocytes, and the researchers further generated organoid-like spherical models as well as engineered micro–heart muscle tissues to better mimic tissue-scale behavior.

To establish causality rather than correlation, CRISPR/Cas9 was applied as “gene scissors” to correct defective RBM20 in diseased cells. The team then reintroduced the DCM mutation into repaired LVNC backgrounds, demonstrating that the specific RBM20 change at that single site is sufficient to generate the divergent calcium phenotypes.

From these mechanistic insights, the researchers mapped potential “molecular players” that may guide precision drug targeting. Among the candidates, verapamil—an established calcium-effect blocker used clinically for arrhythmias—showed partial improvement of contractile performance in laboratory experiments.

The authors emphasize that translating these findings to patients will require extensive preclinical validation. Still, the results outline a roadmap for individualized therapies, where genetic diagnosis could inform which calcium-related pathway to modulate in RBM20-associated disease.

Subject of Research: Cells
Article Title: RBM20 variants disrupt Ca2+ handling and metabolism in dilated and non-compaction cardiomyopathy stem cell models
News Publication Date: 14-Jul-2026
Web References: http://dx.doi.org/10.1038/s41392-026-02838-7
References: 10.1038/s41392-026-02838-7
Image Credits: Katrin Streckfuß-Bömeke / University of Würzburg

Keywords: cardiomyopathy, RBM20, calcium homeostasis, LVNC, DCM, iPSCs, CRISPR/Cas9, stem cell models, verapamil