KDM5C Gene: Function, Conditions, and Inheritance

KDM5C is a gene on the X chromosome that encodes an enzyme responsible for removing specific chemical tags from proteins called histones, which package DNA in every cell. When KDM5C doesn’t work properly, the result is a neurodevelopmental condition known as Claes-Jensen syndrome, characterized by intellectual disability, speech delays, and distinctive facial features. Because the gene sits on the X chromosome, it follows an X-linked inheritance pattern, which means males and females are affected differently and often to different degrees.

What the KDM5C Enzyme Actually Does

Your DNA doesn’t float freely in cells. It wraps tightly around histone proteins, and the chemical modifications on those histones act as signals that tell the cell which genes to turn on and which to keep quiet. One such modification is the addition of methyl groups to a specific spot on histone H3, called lysine 4. When that spot carries two or three methyl groups, it generally marks the promoters of genes that are actively being read. KDM5C’s job is to strip those methyl groups away, which helps dial down or fine-tune gene activity at those locations.¹ The enzyme needs iron and a helper molecule called alpha-ketoglutarate to carry out this reaction.²

Think of it like a volume knob for genes. The methyl marks turn gene expression up; KDM5C turns it back down. Without KDM5C, certain genes stay louder than they should, and this is particularly disruptive in the brain. In neurons, KDM5C works alongside another demethylase called KDM1A to maintain a quiet, repressive state on genes that shouldn’t be active in brain cells. When researchers knocked out both enzymes in mouse forebrains, the result was far worse than losing either one alone: non-neuronal genes switched on in neurons that had no business expressing them, and thousands of new regions gained the active methyl marks that KDM5C normally removes.³ The two enzymes cooperate, and when both are gone, the chromatin landscape in neurons falls apart in a synergistic way.

How KDM5C Shapes Neurons

The brain effects of KDM5C loss are not abstract. In mouse models, neurons missing KDM5C have visibly shorter dendrites, the branching extensions that receive signals from other neurons. Spine density, the number of tiny protrusions on dendrites where synapses form, drops substantially. In one study, neurons in the basolateral amygdala of KDM5C-knockout mice had roughly 45% of the spine density seen in normal mice, and the spines that remained were thinner and lacked the mature mushroom-like shape associated with strong synaptic connections.⁴ The basolateral amygdala is critical for processing emotions such as fear and aggression, which helps explain why these behavioral domains are affected in Claes-Jensen syndrome.

At the molecular level, losing KDM5C scrambles the expression of genes that matter for synaptic communication. Genes important for synaptic transmission and neurodevelopment get turned down, while genes involved in neuronal signal transduction get inappropriately turned up. KDM5C also fine-tunes enhancers, the regulatory DNA regions that control when and how strongly nearby genes are expressed. Among those enhancers are regulatory regions that influence genes like Fos, Arc, and Npas4, which are central players in how neurons respond to experience and form memories.⁵ When these enhancers are not properly tuned, the neuron’s ability to respond adaptively to its environment is compromised.

Behavioral testing in knockout mice makes the downstream consequences clear. Mice lacking KDM5C show impaired fear conditioning, meaning they struggle to associate a context or sound with a prior unpleasant experience. They also perform poorly in spatial learning tasks, taking longer to learn the location of a hidden platform in a water maze. These deficits are not explained by sensory problems or physical limitations: the mice perceive pain normally and swim at typical speeds.⁶

Claes-Jensen Syndrome in Males

When a male inherits or develops a pathogenic variant in KDM5C, the consequences are usually severe because males have only one X chromosome. There is no second copy of the gene to compensate. The resulting condition is known as Claes-Jensen syndrome (also referred to as X-linked intellectual disability, Claes-Jensen type). Males with these mutations typically present with intellectual disability, dysmorphic facial features, and neurodevelopmental delays.⁷

The severity varies depending on the specific mutation and where it falls in the gene, but affected males generally share a constellation of features. Language development is delayed, often significantly. Behavioral challenges, including aggression and difficulty with social interactions, are common. Some individuals develop epilepsy, and autism spectrum features tend to appear more often in people whose variants cluster near the tail end of the protein.⁸ Seizures, once considered uncommon in KDM5C-related conditions, have been reported in a growing number of cases, suggesting they may be an underrecognized part of the picture. A recent large cohort study found seizures occurring in individuals across more than 130 affected families.⁹

How Females Are Affected

For a long time, the assumption was that females carrying a single KDM5C variant would be unaffected carriers. They have two X chromosomes, so the working copy on the other X should compensate. In reality, that picture is far too simple. A growing body of clinical reports shows that many heterozygous females do develop symptoms, though the severity ranges widely, from no apparent effects to moderate intellectual disability.

In a study of 19 females carrying pathogenic KDM5C variants, most presented with learning disabilities or intellectual disability (predominantly moderate), and four also had language impairment mainly in expression. Behavioral disturbances were frequent, and endocrine disorders were more common in females than in males with the condition.¹⁰ Four of the 19 women in that cohort were asymptomatic, illustrating the unpredictability. Separate case reports describe girls with moderate developmental delay, serious expressive language delay, short stature, microcephaly, and recognizable facial features.¹¹

Behavioral symptoms are a meaningful part of the female phenotype. In one review, aggressive behavior was reported in over half of affected females, anxiety in close to half, and short stature in about 70%.¹² The point is that “carrier” does not equal “unaffected.” Any clinical evaluation should take the female phenotype seriously, even though its severity is less predictable than in males.

Why KDM5C Partially Escapes X-Inactivation

The reason females can be symptomatic at all, despite having a second X, ties into a quirk of KDM5C’s behavior on the X chromosome. In female cells, one of the two X chromosomes is normally silenced early in development through a process called X-inactivation. This prevents females from having double the dose of most X-linked gene products compared to males. Most genes on the silenced X stay off, but a subset escape this silencing and continue to be expressed from both X chromosomes. KDM5C is one of those escapees.

Recent work has pinned down the DNA elements responsible for this escape. Researchers showed that a roughly 2,600 base-pair region spanning the KDM5C promoter and the first half of the gene’s first intron contains the elements needed to resist silencing. In experimental cells where X-inactivation was triggered, a single copy of this region allowed about 10% of normal expression from the silenced X, while two copies boosted it to around 20%.¹³

This escape is a double-edged sword. In healthy females, it means that both copies of KDM5C contribute to the cell’s enzyme supply, which likely explains why females generally tolerate one broken copy better than males tolerate one. But it also means that a mutant allele on the “silenced” X is not fully silenced. Depending on which X chromosome is inactivated in a given cell, and how completely it is inactivated, a woman’s tissues can end up with varying ratios of working to non-working KDM5C. This randomness in X-inactivation patterns is a major reason why symptoms in females range from absent to moderate.

Inheritance Patterns and How Variants Arise

KDM5C follows X-linked inheritance, which means the gene is passed through the X chromosome. A mother carrying one pathogenic variant has a 50% chance of passing it to each child. Sons who inherit the variant will be affected; daughters who inherit it may or may not show symptoms, as described above.

Not all KDM5C variants are inherited, though. In a combined cohort of 269 individuals across 130 families with identified KDM5C variants, about 47% inherited their variant from their mother, while roughly 17% had de novo variants, meaning the mutation arose spontaneously in the affected person and was not present in either parent. A small fraction, about 1%, inherited their variant from their father, which is unusual for an X-linked gene and occurred in just three individuals from two unrelated families.¹⁴ Paternal inheritance of a pathogenic X-linked variant is rare, because an affected father would typically exhibit the condition himself, but it can happen if the father carries a mosaic variant or a mild variant that doesn’t fully disable the protein.

The high rate of de novo variants matters clinically. It means that a family with no prior history of intellectual disability can still have a child with Claes-Jensen syndrome. Genetic testing of just the affected individual, without parental comparison, may be sufficient to identify the variant, but confirming whether it is de novo requires testing the parents as well.

Diagnostic Challenges

One of the most striking aspects of KDM5C-related diagnosis is how long it can take. A case report from 2025 illustrates this starkly. More than two decades earlier, a family sought answers for three brothers with intellectual disability, developmental delay, behavioral problems, and dysmorphic features. The genetic tools of the time pointed to a variant in a different gene, FGD1, which was present in all three brothers and their carrier mother. For years, this was assumed to explain the condition. It wasn’t until modern genetic screening and episignature analysis, a technique that looks at patterns of DNA methylation characteristic of specific genetic disorders, that the true cause emerged: a novel pathogenic KDM5C variant.¹⁵

This case is a useful reminder that genetic diagnoses made with older technology deserve revisiting, especially when the clinical picture doesn’t perfectly match the supposed diagnosis. Episignature analysis has become a particularly valuable tool for KDM5C variants because they leave a recognizable methylation fingerprint that can confirm pathogenicity even when the variant itself is novel and hasn’t been seen before.

KDM5C Beyond the Brain

Although intellectual disability dominates the clinical picture, KDM5C is not exclusively a brain gene. It is expressed in tissues throughout the body, and its loss has consequences outside the nervous system.

One connection that has received significant attention is to kidney cancer. In clear cell renal cell carcinoma, KDM5C is one of the more frequently mutated genes, and these mutations occur predominantly in male patients, consistent with the gene’s X-linked location. Research has found that losing KDM5C in kidney cancer cells reprograms how those cells handle glycogen, the storage form of sugar, and also suppresses a form of cell death called ferroptosis that normally keeps abnormal cells in check. Restoring KDM5C function in cancer cells significantly reduced glycogen accumulation, suggesting the gene acts as a tumor suppressor in this tissue.¹⁶ Because females have two copies and one typically escapes X-inactivation, they retain a backup copy that may protect against this cancer, which could partly explain why clear cell renal cell carcinoma is more common in men.

A newer finding links KDM5C to cartilage health. When KDM5C is deleted in chondrocytes, the cells that build and maintain cartilage, mitochondrial function deteriorates. The enzyme normally interacts with and supports a protein called HIF-1α that helps cells cope with low-oxygen environments like those found in joint cartilage. Without KDM5C, HIF-1α levels drop, reactive oxygen species increase, and cartilage homeostasis breaks down, accelerating the progression of osteoarthritis in animal models.¹⁷

What KDM5 Does in Simpler Organisms

Humans have four members of the KDM5 family (KDM5A through KDM5D), but fruit flies have just one version, called KDM5 (or sometimes Lid). This makes flies a useful experimental system because there are no closely related family members to compensate when KDM5 is lost. Researchers studying the fly brain found that KDM5 is required during early neural development for proper formation of the mushroom body, a brain structure central to learning and memory. Intriguingly, the mechanism in this context does not seem to depend on the enzyme’s demethylase activity, suggesting KDM5 family members may have structural or scaffolding roles beyond their known catalytic function.¹⁸ The KDM5 family is conserved across a wide range of animals, and variants in the human relatives KDM5A and KDM5B are also associated with intellectual disability, reinforcing that this enzyme family plays a fundamental role in brain development across species.

The Interplay Between KDM5C and KMT2A

If KDM5C removes methyl marks from histones, there must be an enzyme that puts them on. One major “writer” that adds methyl groups to the same spot on histone H3 is KMT2A (also known as MLL1). The relationship between KDM5C and KMT2A is more nuanced than simple opposition. When researchers bred mice that were missing KDM5C and also had reduced KMT2A, some features partially corrected. For instance, the spine density deficit caused by KDM5C loss was not just rescued but actually overshot normal levels in the double-mutant neurons. Dendrite length, which was shortened in KDM5C-knockout neurons, showed a weak restorative effect when KMT2A was also reduced, landing in a range not significantly different from normal.¹⁹

This mutual suppression is conceptually important. It means the balance between methyl-writing and methyl-erasing at H3K4 matters more than the absolute level of either enzyme alone. Tilting the balance too far in either direction disrupts neuronal structure and behavior. The finding also opens a theoretical therapeutic window: if the problem in Claes-Jensen syndrome is excess methylation because the eraser is broken, then partially reducing the writer could restore balance. This has not been tested clinically, but it illustrates how understanding the basic biology can suggest treatment strategies.

Therapeutic Research

No approved therapies currently target KDM5C-related conditions directly, but there are relevant research threads. KDM5 family inhibitors have been developed mainly in the context of cancer, particularly breast cancer. Several of these compounds have shown promise in preclinical models, with some demonstrating anti-tumor activity comparable to existing chemotherapy drugs but with lower toxicity.²⁰ In the cancer setting, the goal is to block KDM5 enzymes that are helping tumors grow, which is the opposite of what would be needed for Claes-Jensen syndrome, where the enzyme is already non-functional.

On the neuroscience side, a KDM5 inhibitor called KDM5-C70 has been used as a research tool to study how neural stem cells differentiate. Treating stem cells with this compound pushed them toward becoming astrocytes rather than neurons, demonstrating that KDM5 enzyme activity is important for suppressing the astrocyte fate in stem cells.²¹ This is far from a clinical application for Claes-Jensen syndrome, but it adds to the evidence that the KDM5 family’s catalytic activity has specific, manipulable effects on cell fate in the nervous system. The broader challenge for any future neurodevelopmental therapy is timing: much of the damage in KDM5C-related conditions likely occurs during brain development in utero and early childhood, which means postnatal treatment would need to address consequences of an already-altered developmental trajectory rather than simply restoring enzyme activity.