AP2M1. The genetic landscape of neurodevelopmental disorders is shaped as much by what we see as by what we don’t. In 2019, we discovered dominant-negative mutations in AP2M1 as a novel cause of neurodevelopmental disorders with epilepsy, but it also left a lingering question. Judging from large population databases, AP2M1 has all the features of a haploinsufficient gene rather than a dominant-negative disease mechanism. However, protein-truncating variants or deletions in AP2M1 have never been described. In a recent study, we demonstrate that a small deletion on chromosome 3 offers a compelling insight into this mystery. By narrowing down the candidate region of the 3q27.1 deletion syndrome to a 430 kb region, AP2M1 remerges as the most likely candidate. But it also reopens a broader question: Where are the missing haploinsufficient genes?
Figure 1. Schematic representation of overlapping 3q26.3-3q27.2 deletions, and genes involved. The smallest region of overlap (light blue) is 0.43 Mb, encompassing 20 protein-coding genes including AP2M1. The AP2M1 gene encodes a presynaptic adapter protein involved in initiating the first step of synaptic vesicle recycling. The recurrent AP2M1 p.Arg170Trp variant had previously been identified to impair synaptic vesicle recycling and cause genetic epilepsies with clinical features of myoclonic astatic epilepsies (MAE). However, prior to the publication by Gear et al., 2025, haploinsufficiency in AP2M1 had not been known as a disease mechanism (adapted from Gear et al., 2025 under a under the terms of the Creative Commons Attribution and Supplemental data by Helbig et al., 2019).
Missing haploinsufficiency. AP2M1 has a pLI of 0.99, making it among the top candidates for developmental disorders due to haploinsufficiency. And yet, until now, no truncating variants had been definitively linked to human phenotypes. Even though it has never been evaluated systematically, genes constrained against LOF, but for which only gain-of-function or dominant-negative disease mechanisms are known, may be more common than we admit.
Building the case. The DNM1 gene provides a second example of genes with missing haploinsufficiency phenotypes. Despite its LOF intolerance, all pathogenic variants identified so far are dominant-negative alleles, mainly affecting the GTPase domain. In addition, ATP1A3, the causative gene of alternating hemiplegia of childhood (AHC), has a pLI of 1, but haploinsufficiency phenotypes remains speculative at best. So where are individuals with heterozygous nonsense variants? Do these variants result in embryonic lethality? Or does haploinsufficiency in these genes result in clinical presentations that are too mild or non-specific to be captured in current diagnostic pipelines? AP2M1 offers some surprising insight into this question.
3q37. In a recent publication by Gear and collaborators, we addressed this question by narrowing down the interval of the 3q27.1 deletion syndrome. The 3q26.33–q27.2 region had long been associated with a complex genetic syndrome characterized by developmental delay, microcephaly, intrauterine growth restriction, and characteristic facial features. But until now, the causative genes within the microdeletion remained broad and elusive. By including seven previously unreported individuals with 3q27.1 deletions, we were able to narrow down the critical interval to a 430 kb region at 3q27.1, which contains just 20 protein-coding genes.
AP2M1 emerging. Within this 430 kb interval, AP2M1 emerges as the strongest candidate gene, being one of the few genes highly intolerant to loss-of-function (LOF). In addition, the study reports on an individual with a de novo loss-of-function variant in AP2M1 and a neurodevelopmental phenotype. While other genes within the deleted interval—such as CHRD, THPO, and DVL3—may contribute to specific aspects like cardiac anomalies or thrombocytopenia, AP2M1 appears to be the primary neurodevelopmental driver.
Two faces and missing genes. The AP2M1 story adds nuance to the case of missing haploinsufficient genes such as DNM1 and ATP1A3. Haploinsufficiency does cause disease, but the phenotype differs from what we see with dominant-negative or gain-of-function variants. The AP2M1 story is a reminder that allelic series—the spectrum of phenotypes resulting from different types of variants in the same gene—are more common than we think, and that constraint metrics like pLI may be pointing us toward disease genes that have not yet had their full phenotypic story told.
What you need to know. In a recent publication we narrow down the critical region of a long-know microdeletion syndrome with the 3q27.1 region and encounter a familiar gene: AP2M1, a causative gene for genetic epilepsies in which only recurrent missense variants with dominant-negative mechanism had been identified despite strong evidence for possible haploinsufficient phenotypes from population databases. Accordingly, narrowing down the critical region to AP2M1 is more than just than refining a microdeletion syndrome – it is a lens into the hidden burden of underrecognized haploinsufficiency and a call to look more closely at genes whose LOF constraint has outpaced our ability to detect their associated phenotypes.
