The Molecular Biology of ADNP
ADNP-related Helsmoortel-Van der Aa syndrome (HVDAS), often simply called ADNP syndrome, is caused by a change in the ADNP gene. This page explains what the ADNP protein does and what scientists have learned about it, step by step:
1. Cells in a dish, which show what ADNP does inside a single cell.
2. Mice, which reveal how Adnp contributes to development and organ function throughout the body, including the brain, thereby influencing behavior.
3. Human tissue samples, generously donated by individuals with HVDAS and their families, which help understand whether laboratory findings are relevant to people living with the condition. Samples obtained during life and after death can both provide valuable insights.
Each step has its strengths and its limits. Together they give a growing, but still incomplete, picture. Difficult words are explained along the way and collected at the bottom of the page.
What ADNP does: regulating gene activity
Every cell in the body carries the same genes, but different cell types, such as brain cells and muscle cells. In these, different sets of genes, which are made of DNA, need to be active though. Cells therefore carefully control which genes are read and which remain silent. ADNP is one of the proteins involved in this process. It works together with other proteins to regulate how DNA is packaged and how accessible it is to the cell's gene-reading machinery. ADNP is found in the nucleus, the cell's control room where the genes are stored (Mandel et al., 2007; Ostapcuk et al., 2018).
DNA (genes) in the nucleus is not loose and free floating. Instead, it is tightly packaged. Rather like thread wound around a spool. When DNA is tightly packed, the cell's gene-reading machinery has difficulty accessing it. When it is more loosely packed, genes can be more easily read.
ADNP works in a team called ChAHP
ADNP hates to be alone! It is almost always found together with two other proteins, CHD4 and HP1. These three proteins form a team called ChAHP (pronounced "chap"), named after its three members (Ostapcuk et al., 2018). Each member has its own role:
• ADNP is the guide. It recognizes specific "addresses" in the DNA and brings the team there.
• CHD4 is a remodeler. It uses energy to rearrange the proteins around which DNA is wrapped, changing how easily a gene can be read.
• HP1 is a helper that holds on to the packed DNA and to other proteins.
In this way, ChAHP keeps certain genes switched off until the right moment. Studies in mouse cells have shown that this is important when stem cells are triggered to develop into specialized cells (Ostapcuk et al, 2018), such as nerve cells (Clémot-Dupont et al., 2025; Wulfridge et al., 2026).












How well do the mice reflect HVDAS?
HVDAS affects children in many ways, and the range of symptoms varies considerably between individuals. Mouse models capture different aspects of the condition, but no single model reproduces the full diversity of features seen in children. The table compares what each model can tell us.


ADNP and repetitive DNA
A large part of our DNA does not hold recipes for proteins. Much of it consists of repeated pieces of DNA, such as SINEs or ALUs. These originated from ancient "jumping genes" - mobile genetic elements that can copy themselves and spread through the DNA. SINEs can have useful roles in regulating genes, but they can also interfere with normal cell functions if read when they should not be.
Recent studies in mouse cells found that the ChAHP team helps keep SINEs switched off by preventing the cell's reading machinery from landing on them (Schnabl-Baumgartner et al., 2026). A second study showed that the remodeler, CHD4, is essential for this silencing, while HP1 is not required (Ahel et al., 2026). ADNP acts as the captain of the team, bringing its components together and guide them to the SINEs, while CHD4 does the hands-on work of closing it off.
SINEs in mice also attract a protein called CTCF, which helps organize how the long DNA thread folds into a three-dimensional (3D) structure inside the nucleus. Interestingly, ChAHP and CTCF compete for binding at some of these repeats (Kaaij et al., 2019; Wang et al., 2024). Because the 3D folding of DNA influences which genes are active, this competition offers another, more indirect way in which ChAHP may influence gene activity.
Much of what we know about these mechanisms comes from studies in mouse cells. Whether the same processes operate in human cells, and whether they contribute to HVDAS, remain open questions.
ChAHP and its helpers
The ChAHP team has several places where other proteins can attach. Studies have identified additional partners, including POGZ, another protein linked to developmental disorders (Markenscoff-Papadimitriou et al., 2021). These additional partners may act like specialists who help the team with particular tasks.
In human neural stem cells, a 2025 preprint (not peer reviewed) reported that ADNP also works with two proteins called KDM1A and GTF2I (Rizzuti et al., 2025). The study suggests that this partnership may help regulate how neural stem cells multiply and develop into nerve cells.
Researchers are still working out which of these partnerships are stable and long-lasting, and which are more temporary or depend on the cell type or stage of development. They are also investigating what these different partnerships mean for human development and HVDAS.
Step 1: What did cells in a dish teach us?
The main lesson from cell studies is about timing. ADNP helps make sure genes switch on at the right moment. When ADNP is missing or changed, some genes switch on too early and developing nerve cells lose track of what they should become.
Scientists study ADNP in cells grown in a dish. Some are pluripotent stem cells, which can still turn into any type of cell. Others are grown from a patient's own skin or blood cells and turned into nerve cells, or even into mini-brains (organoids): tiny 3D clusters of brain cells that mimic the earliest stages of brain growth. Cells are quick to study and can be changed precisely, but a dish is not a living organism: there is no body, no blood supply and no behavior.
A simplified picture of the timing problem; the lengths are illustrative, not measured
Step 2: What did mice teach us?
Mice reveal what cells in a dish cannot: how changes in Adnp affect the development of the whole animal, the functions of its organs, its physiology and its behaviors. Mouse and human ADNP are more than 90% identical. Studies in mice have been instrumental in advancing our understanding of human development and neuroscience. However, mice and humans are not identical, so mice provide a valuable, but imperfect, model for studying HVDAS (Zamostiano et al., 2001; Larrigan et al., 2026).
Scientists have made several kinds of Adnp mutant mice:
• Mice without any Adnp do not survive: Mice lacking Adnp completely die during embryonic development and show severe developmental abnormalities. This demonstrates that Adnp is essential for normal embryonic development, but this model is much more severe than the genetic changes found in patients with HVDAS. This showed that ADNP is essential for life (Pinhasov et al., 2003).
• Mice with one working copy ("half-dose" mice) model the effects of reduced Adnp dosage, as in a minority of children who have only one working copy of ADNP (Vulih-Shultzman et al., 2007; Hacohen-Kleiman et al., 2018; Cho et al., 2023). This is useful for studying Adnp deficiency, although it does not reproduce every consequence of the variants found in patients with HVDAS.
• Mice carrying the same kind of change as children: a mouse version of the common p.Tyr719* change (Karmon et al., 2022), and a mouse with a frameshift change, p.Leu822Hisfs*6, in the same "hotspot" mutation region found in patients with HVDAS (D'Incal et al., 2026).
• Mice lacking Adnp only in the brain, used to study brain build-up before birth (Clémot-Dupont et al., 2025).
Across these models, established and studied in different laboratories, a few findings come back again and again: the outer layer of the brain (the cortex) is smaller where it was measured, the connections between nerve cells are changed, and the mice show differences in learning, social behavior and anxiety. Consistent with findings in cellular models, the Wnt growth signal was turned down in the brain (D'Incal et al., 2026).
Mouse studies offer valuable insights into HVDAS, although some differences between mice and humans need to be kept in mind. The human brain is more complex than the mouse brain, and behavioral tests, such as finding a hidden platform in a pool of water, provide simplified ways to study abilities related to learning and memory. Laboratory mice are also genetically very similar to one another, whereas children with HVDAS show a wide range of individual characteristics.
A note on Tau
Parents may come across the word Tau when reading about HVDAS. Tau is a protein that helps keep the internal skeleton of nerve cells stable. In ageing-related brain diseases, such as Alzheimer's disease, Tau can clump together into tangles. Tau changes were reported in half-dose mice (Vulih-Shultzman et al., 2007), in p.Tyr718* mice (Karmon et al., 2022) and in the brain of one child with HVDAS (Grigg et al., 2020). These are interesting observations that deserve follow-up. For now, it is best to take the importance of Tau with a grain of salt:
· The findings come from a small number of studies. They have not yet been confirmed by other laboratories or in other mouse models. The human finding comes from a single patient. The authors rightly describe it as a case report, meaning that it documents an observation in one individual. Further studies are needed to determine whether the finding applies to other people with HVDAS.
· ADNP has well-established functions in the nucleus, but it has also been reported to act outside of the nucleus (in the cytoplasm), including in association with microtubule-related proteins. Tau associates with microtubules, which form part of the cell’s internal skeleton. However, the relationship between ADNP and Tau remains unresolved. Crucially, causality has not been demonstrated: it remains unknown whether, and how, ADNP influences Tau function.
• Children with ADNP variants are not known to develop Alzheimer’s disease. However, most people with HVDAS known today are younger than 30. Studies in older individuals are needed to find out whether they become more vulnerable to losing skills, to neurodegeneration (the gradual loss of nerve cells) or to Tau changes.
In conclusion, Tau-related changes may be associated with HVDAS, but whether they contribute to the condition or its symptoms remains unknown.
Step 3: What did a donated human brain teach us?
Cells grown in the laboratory and mouse models are essential for understanding how changes in ADNP affect cells, tissues and development. They allow researchers to investigate disease mechanisms and explore potential therapeutic approaches. But models are not the same as the human condition. Studying human tissue is crucial for checking whether findings from cells and mice also hold true in people with HVDAS.
So far, researchers have had this opportunity only once, thanks to the extraordinary generosity of a family who lost their son to HVDAS. They chose to donate his tissues for research, allowing scientists to study the brain of a child with the condition directly. The researchers are deeply grateful to the family for this generous gift, which has provided the first such insight into the brain of a child with HVDAS.
Two laboratories examined this tissue:
• One study looked at Tau and at gene activity in several organs. They reported Tau changes in the brain and less ADNP activity in the cerebellum, the brain region that coordinates movement (Grigg et al., 2020).
• The other laboratory also looked at the cerebellum specifically and found changes in the chemical tags on DNA and in gene activity. These findings resembled observations from cells and mice, supporting the value of these model systems for understanding HVDAS and showing how studies of human tissue can help validate findings from laboratory models (D'Incal et al., 2024).
Findings from cells and mice that came back in the human brain
• ADNP as a regulator of gene activity. Among genes showing altered DNA methylation (chemical tags on DNA), sequences matching the ADNP-binding motif (as determined in mouse cells) were particularly enriched. This is consistent with a role for ADNP in regulating the human brain's genome, although the study could not directly establish where ADNP was bound in this tissue.
• The Wnt growth signal was turned down, as in nerve cells in a dish and in the frameshift mice.
• The cell's recycling system (autophagy) and its energy factories (mitochondria) showed changes, linking to earlier cell and mouse work. These changes could represent additional characteristic features of HVDAS, although further studies are needed to establish whether they are consistently present in people with the condition.
From cells to mice to the human brain: what was seen at each level
Why we must be careful
This is the tissue of one child, who was also very ill and received strong medicines. Each study compared it with a single control brain: a young adult in one study, and a 9-year-old girl with a different neurological condition in the other. The findings are important clues, but they need to be confirmed with samples from more individuals.
How families can help
One family's generous donation has provided invaluable insights that could not have been obtained from model systems alone. Further donations could help researchers determine which findings are characteristic of HVDAS and which may be specific to an individual. We recognize that thinking about tissue or brain donation can be difficult. Families who would like to learn more can speak with their clinical geneticist or neurologist or contact ADNP Europe. Registering with ADNP Europe can also help families stay informed about research studies they may be able to participate in.
What this means, and what comes next
From cells to mice to one donated brain, a consistent picture is forming: ADNP is a gene switch that helps our cells do the right things at the right time. Some findings, such as changes in Wnt signalling have been reported in cell models, a mouse model and the brain tissue of one child with HVDAS. This convergence is encouraging, although the findings still need to be tested in additional models and human samples, and a direct mechanistic link to ADNP remains to be established. Other findings, such as those related to Tau, require further investigation and independent confirmation.
Understanding what ADNP normally does is an important first step, but we still need to establish how changes in ADNP lead to the diverse features of HVDAS. This is the central goal of ongoing research: to connect molecular changes to the symptoms experienced by children and families, and ultimately to identify new ways to improve care and develop effective treatments.
Key words
References
• Ahel, J., et al. (2026). Remodeling activity of ChAHP restricts transcription factor access to chromatin. Molecular Cell, 86, 2754–2764. doi:10.1016/j.molcel.2026.06.025
• Bennison, S. A., et al. (2023). The cytoplasmic localization of ADNP through 14-3-3 promotes sex-dependent neuronal morphogenesis, cortical connectivity, and calcium signaling. Molecular Psychiatry, 28, 1946–1959. doi:10.1038/s41380-022-01939-3
• Cho, H., et al. (2023). Adnp-mutant mice with cognitive inflexibility, CaMKIIα hyperactivity, and synaptic plasticity deficits. Molecular Psychiatry, 28, 3548–3562. doi:10.1038/s41380-023-02129-5
• Clémot-Dupont, S., et al. (2025). The chromatin remodeler ADNP regulates neurodevelopmental disorder risk genes and neocortical neurogenesis. PNAS, 122, e2405981122. doi:10.1073/pnas.2405981122
• D'Incal, C., et al. (2024). ADNP dysregulates methylation and mitochondrial gene expression in the cerebellum of a Helsmoortel-Van der Aa syndrome autopsy case. Acta Neuropathologica Communications, 12, 62. doi:10.1186/s40478-024-01743-w
• D'Incal, C. P., et al. (2026). An Adnp frameshift variant disrupts Wnt signalling inducing chromatocytoskeletal defects and autism-related behaviour in male mice. eBioMedicine, 128, 106309. doi:10.1016/j.ebiom.2026.106309
• Grigg, I., et al. (2020). Tauopathy in the young autistic brain: novel biomarker and therapeutic target. Translational Psychiatry, 10, 228. doi:10.1038/s41398-020-00904-4
• Hacohen-Kleiman, G., et al. (2018). Activity-dependent neuroprotective protein deficiency models synaptic and developmental phenotypes of autism-like syndrome. Journal of Clinical Investigation, 128, 4956–4969. doi:10.1172/JCI98199
• Kaaij, L. J. T., et al. (2019). The ChAHP complex counteracts chromatin looping at CTCF sites that emerged from SINE expansions in mouse. Cell, 178, 1437–1451. doi:10.1016/j.cell.2019.08.007
• Karmon, G., et al. (2022). Novel ADNP syndrome mice reveal dramatic sex-specific peripheral gene expression with brain synaptic and tau pathologies. Biological Psychiatry, 92, 81–95. doi:10.1016/j.biopsych.2021.09.018
• Larrigan, S., et al. (2026). An allelic series reveals the genetic requirement for Adnp in cortical neurogenesis and learning behavior. bioRxiv (not yet peer-reviewed). doi:10.64898/2026.06.10.731333
• Mandel, S., Rechavi, G., & Gozes, I. (2007). Activity-dependent neuroprotective protein (ADNP) differentially interacts with chromatin to regulate genes essential for embryogenesis. Developmental Biology, 303, 814–824.
• Mandel, S., Spivak-Pohis, I., & Gozes, I. (2008). ADNP differential nucleus/cytoplasm localization in neurons suggests multiple roles in neuronal differentiation and maintenance. Journal of Molecular Neuroscience, 35, 127–141. doi:10.1007/s12031-007-9013-y
• Markenscoff-Papadimitriou, E., et al. (2021). Autism risk gene POGZ promotes chromatin accessibility and expression of clustered synaptic genes. Cell Reports, 37, 110089.
• Ostapcuk, V., et al. (2018). Activity-dependent neuroprotective protein recruits HP1 and CHD4 to control lineage-specifying genes. Nature, 557, 739–743.
• Oz, S., et al. (2014). The NAP motif of activity-dependent neuroprotective protein (ADNP) regulates dendritic spines through microtubule end binding proteins. Molecular Psychiatry, 19, 1115–1124. doi:10.1038/mp.2014.97
• Pinhasov, A., et al. (2003). Activity-dependent neuroprotective protein: a novel gene essential for brain formation. Developmental Brain Research, 144, 83–90.
• Rizzuti, L., et al. (2025). Disruption of ADNP-KDM1A-GTF2I complex drives neural differentiation imbalance in Helsmoortel-Van der Aa syndrome. bioRxiv (not yet peer-reviewed). doi:10.1101/2025.03.06.641037
• Schnabl-Baumgartner, J., et al. (2026). ChAHP silences SINE retrotransposons by inhibiting TFIIIB recruitment. Molecular Cell, 86, 2765–2776. doi:10.1016/j.molcel.2026.06.024
• Sun, X., et al. (2020). ADNP promotes neural differentiation by modulating Wnt/β-catenin signaling. Nature Communications, 11, 2984. doi:10.1038/s41467-020-16799-0
• Vulih-Shultzman, I., et al. (2007). Activity-dependent neuroprotective protein snippet NAP reduces tau hyperphosphorylation and enhances learning in a novel transgenic mouse model. Journal of Pharmacology and Experimental Therapeutics, 323, 438–449. doi:10.1124/jpet.107.129551
• Wang, W., et al. (2024). ADNP modulates SINE B2-derived CTCF-binding sites during blastocyst formation in mice. Genes & Development, 38, 168–188.
• Wulfridge, P., et al. (2026). ADNP regulates chromatin architecture and lineage fidelity during neural differentiation. PLOS Genetics, 22, e1012081.
• Zamostiano, R., et al. (2001). Cloning and characterization of the human activity-dependent neuroprotective protein. Journal of Biological Chemistry, 276, 708–714. doi:10.1074/jbc.M007416200
