Reduced functional dosage
Gene-level haploinsufficiency is established.
GENE & BIOLOGY
ASH1L regulates chromatin through more than one molecular action.
Human genetics establishes that reduced ASH1L function can cause autosomal-dominant syndromic neurodevelopmental disorder. Structural and experimental studies show a large chromatin regulator whose recruitment, reader, catalytic, and downstream effects depend on molecular partners, genomic locus, cell identity, and developmental state.
E1Established human geneticsClinGen validity and dosage
E2Direct molecular evidenceStructure, biochemistry, or a defined molecular assay
E3Model-system evidenceAn engineered cell perturbation, species, lineage, stimulus, or developmental system
XTestable hypothesisA prediction that still needs a causal experiment
THE BIOLOGICAL ARGUMENT
A multi-domain protein, a context-dependent chromatin output, and a human phenotype that must be measured over time.
The map separates what is established in people from what has been measured in proteins, cells, and model systems. It also shows where the causal chain stops.
Evidence can move across scale only when the model, perturbation, lineage, time course, and rescue remain attached.
Recruit · sense · write
Multiple molecular jobs must be measured separately.
Identity changes the output
Neural state, lineage, locus, and stimulus matter.
Phenotype across time
Clinical chronology identifies questions; it does not finish the mechanism.
ALLELEWhat RNA and protein remain?
CONTEXTWhich partner, locus, cell, and developmental state?
DYNAMICSWhat opens, stabilizes, closes, and recovers?
CAUSALITYDoes correction or ASH1L rescue normalize the readout?
HUMAN-GENETIC ANCHOR
The gene–disease relationship is established. Variant-level function is not interchangeable.
ClinGen gene-validity and dosage curations answer different questions. They support ASH1L as a definitive disease gene and loss of one functional copy as a disease mechanism; they do not turn every reported molecular finding into the same functional allele.
Autosomal-dominant syndromic intellectual disability / neurodevelopmental disorder.
ClinGen classification · February 23, 2023Sufficient evidence for haploinsufficiency; no current evidence for triplosensitivity.
ClinGen dosage evaluation · June 5, 2017Nonsense-mediated decay, residual protein, catalytic activity, chromatin engagement, localization, and rescue remain allele-level questions.
Sequence-defined and non-sequence contexts stay analytically separate.HOW ASH1L WORKS
ASH1L is not one mark and not one switch.
Its C-terminal machinery can recruit the protein to nucleosomes, recognize local chromatin, regulate an autoinhibited catalytic region, and produce context-dependent outputs.
DIRECT MOLECULAR EVIDENCE
Recruit
MRG15 can enhance ASH1L nucleosome recruitment and catalytic activity. Structural studies also support relief of catalytic autoinhibition in studied complexes.
DIRECT STRUCTURE–FUNCTION EVIDENCE
Sense
The PHD finger recognizes H3K4me2/3. The 2025 structure–function study also reported DNA-binding functions in neighboring C-terminal modules and a PHD–BAH unit that supports nucleosome engagement.
DIRECT BIOCHEMICAL EVIDENCE
Write
The strongest catalytic assignment is as an H3K36 mono-/dimethyltransferase in defined systems. Specific DNA-repair experiments also report ASH1L–MRG15-dependent H3K4me3 and FACT deposition.
THE CENTRAL BIOLOGICAL RULE
Interpret context
Partner, substrate, genomic locus, cell identity, developmental stage, and stimulus determine the measured output. ASH1L can support activation, maintenance, repression, or repair in different experimental contexts.
REFERENCE PROTEIN
A long, sparsely annotated N-terminus precedes a multi-domain C-terminal machine.
The first roughly 2,000 amino acids remain poorly characterized. Most experimentally mapped catalytic, reader, and chromatin-engagement modules cluster toward the C-terminus. That asymmetry describes the research record—not biological importance.
Long N-terminal regionAWS 2087–2138SET 2141–2261Bromo 2438–2543PHD 2581–2623BAH 2660–2794
APPROXIMATELY AA 1–2,000
The long N-terminal region is still poorly characterized.
A 2025 structure–function study noted that functional information was not available for the first roughly 2,000 amino acids. “Unannotated” is a research gap—not evidence that this region is unimportant or clinically interchangeable.
Review the 2025 structure–function study ↗DOMAIN-RICH C-TERMINAL REGION
Most experimentally mapped domains cluster toward the C-terminus.
The AWS/SET catalytic region is followed by bromodomain, PHD, and BAH modules involved in catalytic control and chromatin engagement. Structural knowledge is much deeper here, but position alone still cannot predict a person’s phenotype.
Open the molecular & structural collection →The display uses the 2,969-aa UniProt canonical scale. RefSeq NP_060959.2 is 2,964 aa; every variant must be transcript- and isoform-crosswalked before positional analysis.
Catalytic activity and autoinhibitory control.
Isolated-domain behavior is not full-length function.DNA binding was reported in the 2025 structure–function study.
Do not assume canonical acetyl-lysine reading without direct evidence.Recognition of H3K4me2/3 in structural and biochemical systems.
Reader function is distinct from catalytic product.DNA/linker-DNA engagement and integration with the neighboring PHD module.
Coordinates depend on reference sequence and annotation source.WHAT THE EXPERIMENTS ACTUALLY SHOW
Each model earns a specific conclusion—and no more.
The same word, “ASH1L,” can refer to human heterozygous disease, complete gene knockout, an isolated catalytic region, a mouse lineage, or a fly ortholog. Those perturbations cannot be silently exchanged.
Evidence contextMeasured evidenceWhat it establishesWhat it does not establish
E1Human genetics
Definitive autosomal-dominant gene–disease relationship; sufficient evidence for haploinsufficiency.
Reduced functional ASH1L dosage can cause syndromic neurodevelopmental disease.
Classify every missense, splice, CNV, complex, or familial finding—or predict severity.
E2Protein structure & biochemistry
Autoinhibitory catalytic control, MRG15-dependent activation or recruitment, H3K4me2/3 recognition, DNA engagement, and H3K36 methyltransferase activity.
ASH1L is a multi-domain chromatin regulator with separable recruitment, reader, and catalytic functions.
Show how a specific heterozygous human allele behaves in a relevant cell.
E3Human neural CRISPR models
ASH1L was perturbed in neural progenitors, immature and mature glutamatergic neurons, and mature GABAergic neurons in the 2026 pooled study.
Downstream consequences must be resolved by neural cell state. An ASH1L-specific arrayed assay found increased Ki-67 signal in neural progenitors.
Turn cross-gene convergence into an ASH1L-specific result. Reported ASH1L Seahorse parameters were not significantly changed.
E3Developmental and tissue models
Selected mouse, fly, and lineage-specific systems connect Ash1l to differentiation, neuronal activity, muscle fusion, epidermal repair, bone remodeling, cardiac development, and immune signaling.
Concrete tissues, cell types, stimuli, and readouts worth testing.
Establish the frequency, cause, or management of a corresponding feature in people.
HUMAN OBSERVATIONConnected human cohort
Molecular context, fifteen clinical lanes, chronology, load windows, formal records, reported formal content, and parent observation are mapped separately.
Which human signals need prospective measurement and comparison.
Prove that a biologically plausible pathway caused any person’s finding.
2026 neural-model precision: the pooled study’s broad convergence results span multiple neurodevelopmental genes. For ASH1L specifically, the published arrayed work supports increased Ki-67 signal in neural progenitors, while the reported Seahorse respiration parameters were not significantly altered. Complete knockout in modeled cells is not a heterozygous patient-allele dosage series.Read Fernandez Garcia et al. ↗
CHROMATIN × CONTEXT
H3K36–Polycomb antagonism is real biology. A universal ASH1L→Polycomb disease pathway is not established.
H3K36me2/3 can inhibit PRC2-mediated H3K27 methylation in biochemical systems, and Ash1l can oppose Polycomb repression at selected developmental loci. Other experiments show that Ash1 can counter Polycomb even when H3K36 methylation is unavailable, leaving room for additional catalytic, recruitment, or non-histone mechanisms.
Mark-level antagonism
H3K36 methylation can directly inhibit PRC2-mediated H3K27 methylation. This is a biochemical relationship, not a patient-level measurement.
Yuan et al. ↗Selected developmental genes
Mouse differentiation and tissue models connect Ash1l to Polycomb opposition and activation thresholds at selected loci.
Miyazaki et al. ↗Not always H3K36-dependent
Drosophila experiments show Polycomb counteraction can persist without H3K36 methylation, arguing against one obligatory downstream route.
Dorafshan et al. ↗MOTIVATED BY THE BIOLOGY
Selected gene programs may open, stabilize, or close at the wrong time in a particular cell state.
NOT DEMONSTRATED IN PEOPLE
- This does not show universal loss of bulk H3K36 methylation
- Genome-wide H3K27me3 spread
- Broad failed bivalent-domain resolution in patient cells
- One “failure to close” mechanism across organ systems
DECISIVE EXPERIMENT
A timing hypothesis requires a time course.
A single baseline sample cannot show delayed opening, premature closure, prolonged activation, or slow recovery. The experiment must follow the same allele and lineage through a defined transition and test whether correction or ASH1L rescue restores the trajectory.
- 01
Baseline
Stable molecular and functional reference
- 02
Open
Defined signal, transition, or differentiation step
- 03
Peak
Maximum chromatin, RNA, and functional response
- 04
Close
Feedback, termination, or stabilization phase
- 05
Recover
Return to baseline, adaptation, or new steady state
Chromatin
ASH1L occupancy · H3K36me1/2 · H3K27me3 · H3K4me3 · accessibility
Transcription
Immediate response · lineage programs · feedback brakes · RNA processing
Cell function
Maturation · excitability · barrier integrity · metabolism · repair
Clinical bridge
Prospectively defined state, trigger, duration, recovery, and objective endpoint
Measure residual function first.
Transcript use, nonsense-mediated decay, RNA and full-length protein abundance, localization, catalytic activity, and chromatin engagement.
Use isogenic, lineage-matched comparisons.
Repeat the same perturbation and recovery curve in relevant neural and non-neural cells instead of pooling incompatible contexts.
Require normalization, not association alone.
Correction or ASH1L rescue should restore the molecular trajectory and the paired functional readout in independent experiments.
WHAT WOULD WEAKEN THE MODEL?
- No reproducible allele-dependent chromatin or RNA trajectory
- No lineage or state specificity across repeated experiments
- No normalization after genetic correction or ASH1L rescue
Scientific naming boundary: “state-timing” or “epigenetic synchrony” can describe a research framework. Neither is an established diagnosis, a demonstrated single mechanism across organ systems, or a basis for treatment selection. A broad “failure to close” account remains Tier X until patient-allele, isogenic, time-resolved, rescue-supported experiments demonstrate it.
FROM BIOLOGY TO STUDY DESIGN
Biological plausibility defines an assay—not a clinical attribution.
Broad expression and model-system findings identify tissues, pathways, and measurements for study. Human attribution still requires case-level evidence, appropriate comparators, prospective timing, and a design capable of separating ASH1L effects from competing explanations.