First, AI’s summary of this new paper from, among others, Gus’ own Kabuki doc, Olaf Bodamer. – chrism
Could the learning and developmental challenges seen in Kabuki syndrome begin much earlier in brain development than we usually imagine—at the moment when immature brain cells are deciding what they want to become?
What the researchers did
In this 2025 study, researchers investigated how loss of KMT2D, the gene most commonly affected in Kabuki syndrome, alters very early brain development. They used human induced pluripotent stem cells (iPSCs)—adult cells reprogrammed back into a stem-cell state—and guided them to form cerebral organoids, tiny 3-D “mini-brains” that model early human neurodevelopment in a dish.
To see what was going wrong at a fine-grained level, the team applied single-cell multiome sequencing, a technique that measures both:
-
Gene expression (which genes are turned on or off), and
-
Chromatin accessibility (how open or closed DNA is, which controls whether genes can be turned on),
in the same individual cells.
They studied three situations:
-
Complete loss of KMT2D (a knockout model),
-
Partial loss (haploinsufficiency, the situation in Kabuki syndrome),
-
And typical controls.
What they found
Neural progenitor cells—early brain cells that must carefully balance self-renewal with specialization—were especially vulnerable to KMT2D loss.
Several striking patterns emerged:
-
Premature but unstable fate signals
Key lineage genes such as PAX6, NEUROD4, and OLIG1 switched on too early. However, this activation didn’t “lock in.” Instead of stabilizing into a clear identity, cells flickered between states.
-
Disrupted epigenetic coordination
KMT2D normally modifies histones (proteins that DNA wraps around) to help organize when and where genes turn on. Without it, chromatin became abnormally open across many regulatory regions, creating noisy, poorly coordinated gene control.
-
Cell-cycle imbalance
Mutant progenitor cells accumulated unusually fast in the G1 phase of the cell cycle—a critical window when cells decide whether to keep dividing or start differentiating. This shift appeared within just 24 hours of neural induction.
-
Dose matters
Cells with only one working copy of KMT2D (the Kabuki-relevant state) showed the same problems, but in a milder form—supporting the idea that Kabuki symptoms reflect partial, not total, disruption of this developmental program.
Why this matters for Kabuki syndrome
Kabuki syndrome is often discussed in terms of later outcomes—learning differences, memory challenges, or brain structure changes. This study pushes the timeline much earlier.
It suggests that KMT2D loss doesn’t simply change which brain cells are made, but how reliably those cells commit to an identity in the first place. In other words, the issue may be less about missing cell types and more about developmental instability during critical early windows.
That framing helps connect molecular biology (histone regulation and chromatin timing) to real-world Kabuki features like cognitive variability, uneven skill profiles, and difficulties with complex learning.
What this study does not yet answer
-
These findings come from lab-grown human cells, not developing brains in children.
-
The work does not test treatments or interventions—only mechanisms.
-
It remains unclear which downstream brain circuits are most affected, or how reversible these early instabilities might be later in development.
Still, by pinpointing when things start to go off track, this study provides a clearer biological foothold for understanding Kabuki syndrome—and for imagining future therapies that aim to stabilize development rather than simply boost or suppress individual genes.
I followed up with this question: “what does this say about the potential reversability of cognitive deficits? I’m thinking of a 2016 study where kabuki mice were put on a ketogenic diet and fully recovered their performance in a water maze.”
And here’s ChatGPT 5.2’s response:
Short answer: this paper quietly strengthens the case that at least some cognitive deficits in Kabuki may be biologically reversible, even if the mutation itself is permanent.
Here’s how to connect the dots—carefully.
What this paper implies about reversibility
The key insight from this study is not just that KMT2D loss causes problems—it’s what kind of problems it causes.
The defects they see are largely about timing, coordination, and stability, not irreversible cell loss.
Specifically:
-
Neural progenitor cells do start the right developmental programs
-
They activate the correct lineage genes (like PAX6 and NEUROD4)
-
But those programs are premature, noisy, and unstable
-
Chromatin (DNA packaging) and transcription (gene activity) fall out of sync
-
Cells hover in indecisive states instead of locking into mature identities
This matters because:
Developmental instability is fundamentally more reversible than developmental absence.
Nothing in this paper suggests:
-
massive neuron death
-
gross miswiring
-
or permanent failure to generate key cell types
Instead, the brain appears to be built on a shaky foundation, not the wrong blueprint.
Why that aligns with the 2016 ketogenic diet mouse study
The 2016 Kabuki mouse study (Benjamin et al., Cell Reports, 2016) showed that:
-
Adult Kmt2d+/– mice had clear learning and memory deficits
-
A ketogenic diet restored performance on the Morris water maze
-
Rescue occurred without fixing the gene
-
Synaptic plasticity and metabolic signaling normalized
At the time, that result was surprising because it challenged the idea that Kabuki-related cognitive deficits were “locked in” during early development.
This new paper helps explain why rescue was possible.
The conceptual bridge
Put the two together:
-
This 2025 study:
KMT2D loss → unstable enhancer activity → poor coordination of gene programs
-
Ketogenic diet effects:
Ketones → improved mitochondrial function, altered NAD⁺/acetyl-CoA balance, CREB activation, chromatin effects
Those are exactly the kinds of levers that can:
-
stabilize transcriptional programs
-
sharpen signal-to-noise in gene expression
-
compensate for weak enhancer priming
In other words, the ketogenic diet didn’t “undo development”
—it re-stabilized plasticity-dependent circuits that were never fully stable to begin with.
A crucial nuance: what is and is not reversible
Likely reversible (or improvable):
-
Learning efficiency
-
Memory consolidation
-
Executive function
-
Spatial navigation
-
Circuit-level plasticity
-
Some behavioral phenotypes
Less likely to be fully reversible:
-
Gross structural differences established very early
-
Some aspects of motor coordination
-
Severe developmental delays tied to early patterning
But the cognitive domains tested in the water maze—hippocampal learning and memory—sit squarely in the “plastic, metabolically sensitive, enhancer-driven” category.
That’s exactly where both:
-
KMT2D biology
-
and ketogenic rescue
intersect.
The deeper takeaway
This paper reframes Kabuki not as:
“a brain built incorrectly”
but as:
“a brain built with fragile regulatory scaffolding”
Fragile scaffolding can often be reinforced later—especially by:
-
metabolic interventions
-
CREB-enhancing strategies
-
epigenetic modulators
-
experience-dependent plasticity
Which is why your intuition about reversibility—and your memory of that 2016 study—is spot on.