NLGN4X belongs to the neuroligin family, a group of postsynaptic adhesion proteins that sit on the receiving side of a synapse and help bind the presynaptic and postsynaptic compartments together. Neuroligins were first identified in the 1990s as synaptic cell-adhesion molecules capable of inducing the formation of functional connections between neurons, and they have since become some of the most intensively studied synaptic proteins in neuroscience. NLGN4X, located on the X chromosome, is of particular clinical interest because loss-of-function mutations in the gene have been reported in individuals with autism spectrum disorder, intellectual disability, and related conditions. Because the gene sits on the X chromosome, and because a close paralog called NLGN4Y exists on the Y chromosome, questions about sex-linked patterns of inheritance have surrounded NLGN4X for years, adding a further layer of relevance to debates about why neurodevelopmental conditions often affect males and females differently.

At the level of the synapse, the biology of neuroligin-4X is well described. The protein is anchored in the postsynaptic membrane, where its extracellular domain reaches across the synaptic cleft and engages with presynaptic neurexins, partner proteins that help organize the release of neurotransmitters. Through this trans-synaptic bridge, neuroligin-4X contributes to the balance between excitatory and inhibitory signaling, a balance that many investigators regard as one of the recurring weak points in autism biology. Pathogenic variants in NLGN4X can destabilize the protein, reduce its trafficking to the cell surface, or otherwise compromise its ability to cluster with presynaptic partners, and each of these defects can translate into altered synaptic strength or altered synaptic specification. In animal and cellular models, disruption of neuroligin function has repeatedly been shown to shift the excitatory-to-inhibitory ratio, a change thought to degrade the fidelity of neural circuits involved in language, social processing, and sensory integration.

The new analysis, however, pushes the story beyond the synapse. By assembling evidence around pathogenic NLGN4X variants and their cellular consequences, the study highlights that the gene also appears to participate in cortical development, the protracted process by which neural progenitor cells proliferate, migrate to their correct positions, and differentiate into the layered architecture of the cerebral cortex. This is a striking reframing. The cortex is built over weeks to months, and its formation depends on the precisely timed behavior of radial glial cells, intermediate progenitors, and migrating neurons. If NLGN4X variants compromise not only mature synapses but also these earlier developmental programs, then the clinical outcomes associated with the gene may reflect a compound insult: circuits that are miswired because neurons ended up in the wrong places, and circuits that still misfire even where the wiring is correct, because the synapses themselves are deficient.

Technically, the distinction matters because it changes the kind of experiments and therapies one would consider. If a variant’s primary effect were purely synaptic, then approaches aimed at restoring the excitatory-inhibitory balance, for example through pharmacological modulation of inhibitory signaling, might in principle compensate for the deficit. But if the same variant also perturbs progenitor proliferation or neuronal migration during corticogenesis, the window for intervention may open much earlier, and the targets may lie in developmental signaling pathways rather than in neurotransmitter receptors. The dual-role framing also helps explain why the severity of symptoms linked to NLGN4X mutations can be so variable. Two individuals carrying variants that affect the same protein domain might experience different outcomes depending on how strongly each developmental process is disrupted, and on compensatory mechanisms that differ from one brain to another.

The study also speaks to an ongoing debate about genotype-phenotype relationships in neurodevelopmental disorders. NLGN4X variants have been associated with a wide clinical spectrum, ranging from severe intellectual disability with limited language to milder presentations that include autism features without global cognitive impairment. Some of this variability has been attributed to the position and biochemical nature of each mutation, and laboratory work has shown that different missense variants can produce different degrees of protein instability and trafficking failure. By connecting synaptic phenotypes with cortical developmental phenotypes, the new findings offer a conceptual scaffold for this variability: clinical severity may track not with a single disrupted function but with the combined burden placed on multiple biological processes by a single genetic lesion. This kind of multidimensional view is increasingly common in the field, as large-scale sequencing has revealed that many neurodevelopmental risk genes are pleiotropic, influencing several cellular systems at once.

There is also a dimensional angle to the findings that will interest researchers working on sex differences in brain disorders. Because NLGN4X sits on the X chromosome, males with damaging variants have only one functional copy, while females carry a second copy on their other X chromosome, subject to X-inactivation. In females, the random nature of X-inactivation means that a proportion of neurons will express the healthy allele and a proportion will express the mutated one, potentially producing a mosaic of affected and unaffected cells. How such mosaicism plays out across cortical development is an open question, but the dual-role model gives it new significance: a mosaic brain might simultaneously contain patches of neurons with migration or differentiation deficits and synapses of uneven quality, generating a heterogeneous pattern of circuit function that could contribute to the often subtler clinical presentations seen in carrier females.

From a methods standpoint, work in this area typically combines patient-derived cells, engineered cellular models, and computational analyses of gene expression to link variants to mechanisms. Induced pluripotent stem cell lines from affected individuals can be differentiated into cortical neurons, allowing investigators to compare synapse formation, neurite outgrowth, and electrophysiological properties between mutant and corrected lines. More recent single-cell transcriptomic approaches have made it possible to ask whether NLGN4X expression patterns track particular cortical cell types or developmental stages, a strategy that can reveal whether the gene’s role in early cortical formation is direct or secondary. Protein interaction studies complement these approaches by mapping how specific variants disturb binding to neurexins and other synaptic partners, and structural work has clarified how alterations in particular domains compromise the stability of the neuroligin-neurexin complex. The synthesis presented in the new paper draws on this methodological ecosystem to argue that the synaptic and developmental roles of NLGN4X are not separate stories but facets of a single, integrated biological function.

The broader implication is that the list of autism-associated synaptic genes may need to be re-read through a developmental lens. Genes such as those encoding other neuroligins, neurexins, and SHANK scaffolding proteins are conventionally categorized by their synaptic functions, yet many of them are expressed during early brain development as well. If pathogenic variants in these genes perturb corticogenesis, then therapeutic strategies aimed only at synapses may address the downstream consequences of a problem whose origins lie months earlier in development. Conversely, understanding the developmental roles of these genes could open avenues for early detection, since markers of altered cortical development might in principle be identified before behavioral symptoms appear. For families affected by NLGN4X-related conditions, the practical value of the current study lies less in immediate treatment options than in sharpening the biological map that future therapies will depend on. By showing that one gene can simultaneously govern the formation of the cortex and the function of the synapses within it, the research underscores a central lesson of modern neuroscience: to understand a brain that struggles to connect, one must sometimes go back to the very beginning of its construction.

Subject of Research: How pathogenic NLGN4X variants affect both synaptic connectivity and cortical development

Article Title: Pathogenic NLGN4X variants reveal dual roles in synaptic connectivity and cortical development

Article References: Hong, E., Reyes Mendez, M., Guissart, C., Zaafrane-Khachnaoui, K., Kapoor, A., Brzdąk, P., Badger, J. D., II, Lu, W., & Roche, K. W. (2026). Pathogenic NLGN4X variants reveal dual roles in synaptic connectivity and cortical development. Translational Psychiatry. https://doi.org/10.1038/s41398-026-04464-x

Image Credits: AI Generated

DOI: 10.1038/s41398-026-04464-x

Keywords: NLGN4X, neuroligin, autism spectrum disorder, synaptic connectivity, cortical development, intellectual disability, X chromosome, neurexin, excitatory-inhibitory balance, neurodevelopmental disorders, corticogenesis, genotype-phenotype

The research, led by Xiaoli Xie, Delin Tan, Jingyun Mo and colleagues under the supervision of Huiqing Jiang, Tingting Li and Jia Wang, set out to address a conspicuous gap in cancer genomics. While alterations in three-dimensional chromosome organization—topologically associating domains, chromatin loops and A/B compartments—are well documented in many tumors, the behavior of the X chromosome itself during carcinogenesis has attracted remarkably little attention. Yet the X chromosome carries more than 800 genes, many involved in cell proliferation, metabolism and immune regulation, and liver disease is notorious for its pronounced sex bias: men face a substantially higher risk of hepatocellular carcinoma than women, a pattern the team suspected might be written into the architecture of the sex chromosomes.

To follow genomic architecture across disease progression, the researchers built what they describe as a 4D nucleome landscape, combining chromosome conformation capture technologies such as Hi-C with ChIP-seq, RNA-seq, whole-genome sequencing and DNA methylation profiling, sampled at multiple time points as chemically induced liver disease unfolded in mice. Using F1 hybrid mice, in which the two parental X chromosomes can be distinguished by allele-specific sequence variants, the team could separate the folding of the single active X chromosome in males from the paired active and inactive X chromosomes in females. This allelic resolution proved critical, because it exposed a striking asymmetry: the conformational changes and gene activation occurring on the male active X chromosome during hepatocarcinogenesis were far more pronounced than those on the female active X chromosome, mirroring the male bias seen in human liver disease.

The most surprising finding emerged when the team turned to human hepatic tissue. In normal liver samples from both men and women, the active X chromosome is not organized as a single continuous domain but is split into two large mega-domains separated by a defined boundary located near the 90-megabase region of the chromosome. This bipartite configuration is specific to the liver—it was not observed in other tissues such as colon and lung—and, notably, it appears largely independent of the polyploid state of hepatocytes, which commonly carry doubled or quadrupled genome copies without ill effect. The bipartite active X chromosome, in other words, is a liver-specific architectural feature, maintained even in cells with unusual ploidy.

That stability, however, collapses early in disease. Analyzing fibrotic, cirrhotic and tumoral liver tissue from patients of both sexes, the researchers found that the mega-domain boundary weakens and the two halves of the active X chromosome merge into a single, unified domain. The merging is not merely a passive byproduct of genomic chaos in tumors. When the two halves fuse, the strengthened long-range contacts between them occur predominantly between A-type, active compartments, and these new A–A interactions are associated with the activation of X-linked genes. Genes on the X chromosome were among the most disproportionately upregulated in hepatocellular carcinoma samples, a pattern confirmed in data from The Cancer Genome Atlas, where the ratio of upregulated to total genes on the X chromosome ranked at or near the top in male tumors.

To establish causality rather than correlation, the team employed CRISPR–Cas3, a genome-editing system capable of deleting megabase-scale stretches of DNA that standard Cas9 cannot handle. By excising the boundary element that separates the two mega-domains of the active X chromosome, the researchers recreated, in isolation, the architectural event that accompanies liver cancer. The result was unambiguous: disruption of the bipartite conformation alone was sufficient to trigger a premalignant transition. In liver organoids, boundary deletion transformed normally cystic structures into thickened, compact spheroids reminiscent of early tumor growth, accompanied by the activation of X-linked genes. The boundary, it seems, functions as a tumor-suppressive architectural element whose loss unlocks oncogenic expression programs.

Having demonstrated that structural dissolution drives malignant change, the researchers then asked which of the activated X-linked genes actually matter for liver cancer. Using complementary CRISPR-based screening approaches—CRISPRa to boost the expression of genes that were weakly expressed, and CRISPRi to silence genes that were highly expressed—the team winnowed the long list of upregulated X-linked genes down to a set of functional drivers in hepatocellular carcinoma, including metabolic and stress-response genes such as G6PD and members of the ARMCX family. The identification of these effective genes provides a bridge between an abstract chromosomal conformation and concrete molecular targets that could, in principle, be therapeutically interrogated.

The study’s technical scope is considerable. The raw data, encompassing Hi-C, ChIP-seq, RNA-seq, whole-genome sequencing and bisulfite sequencing from both mouse and human samples, have been deposited in public archives, and all analysis code has been released through GitHub and Zenodo. Genome-wide, the researchers also documented the broader erosion of liver genome organization during carcinogenesis: topologically associating domains weakened, CTCF and H3K27ac signals at boundaries diminished, chromatin loops faded, and DNA methylation increased. Against this backdrop of global architectural decay, the specific, boundary-dependent dissolution of the bipartite active X chromosome stands out as an unusually discrete and mechanistically tractable event—one of the earliest and most consistent architectural signatures of hepatocarcinogenesis in both mice and humans.

The findings also carry implications for the long-standing puzzle of sex bias in liver cancer. Because males possess only a single X chromosome, every hepatocyte depends entirely on the architectural integrity of that one active X for proper dosage regulation of X-linked genes. The more pronounced conformational disruption and gene activation observed on the male X during disease progression offers a structural explanation for why male livers appear more vulnerable to malignant transformation, complementing earlier hypotheses about tumor-suppressor genes that escape X inactivation. Intriguingly, the work resonates with prior studies showing that the inactive X chromosome also folds into a bipartite structure in mice—one organized around the macroH2A-rich DXZ4-like boundary—suggesting that bipartite architecture may be a general feature of X chromosome biology whose functions differ profoundly between the active and inactive states.

For the field of cancer genomics, the message is that the fourth dimension of the genome—its behavior across time and disease states—can reveal vulnerabilities invisible to static snapshots. A liver-specific, megabase-scale boundary on the active X chromosome emerges from this study as both an early casualty of hepatocarcinogenesis and a functional guardian against it. If confirmed and extended, monitoring the integrity of this bipartite structure, or the X-linked genes it restrains, could eventually inform risk stratification in patients with chronic liver disease, while the boundary itself and the genes it controls offer fresh mechanistic entry points into one of the world’s deadliest cancers.

Subject of Research: Dissolution of the bipartite active X chromosome structure during hepatocellular carcinoma development

Article Title: 4D nucleome landscape reveals dissolution of bipartite active X chromosome in liver cancer development

Article References: 4D nucleome landscape reveals dissolution of bipartite active X chromosome in liver cancer development. (n.d.). https://doi.org/10.1038/s41556-026-02063-y

Image Credits: AI Generated

DOI: 10.1038/s41556-026-02063-y

Keywords: X chromosome, 4D nucleome, hepatocellular carcinoma, genome architecture, Hi-C, bipartite mega-domain, liver fibrosis, CRISPR-Cas3, X-linked gene activation, sex bias in cancer, chromatin loops, premalignant transformation