LAMP5 (BAD-LAMP) - a direct MLL-fusion target gene, a proposed immunotherapy antigen, and a brain protein
also called BAD-LAMP · brain and dendritic cell-associated LAMP · LAMP-5 · UNC46 · C20orf103
LAMP5 is a membrane protein switched on directly by the KMT2A fusion, and leukaemia cells need it to grow, which is why it has been proposed as a target for immune therapy. It is also one of the most brain-enriched proteins in the body, concentrated in the basal ganglia.
The same thing, for a scientist
LAMP5 (UniProt **Q9UJQ1**, LAMP5_HUMAN, **280 aa**, ENSG00000125869) is a single-pass type I membrane glycoprotein with an extracellular topological domain spanning residues **30-235** [E4]. It reached this graph through a published surfaceome screen: filtering leukaemia expression repositories through the in silico human surfaceome identified LAMP5 as highly and specifically expressed in MLL-rearranged leukaemia and a **direct target of the oncogenic MLL-fusion protein**, whose depletion inhibits leukaemia growth in vitro and in vivo and which routes innate-immune signalling away from type-I interferon and toward NF-kB [E1]. This graph already holds that mechanism as `cell-state-innate-immune-lamp5`; what it has not held until now is the gene itself, or its normal-tissue distribution. In the three KMT2A::AFF1 lines LAMP5 sits at SEM 120.1, RS4;11 284.4 and MV4-11 372.4 nTPM against 8.41 across all 1,206 HPA cell lines - a very large leukaemia signal [E2]. It is **not** AFF1-specific: 2.68x a B-precursor comparator and only **1.18x** the KMT2A-rearranged-other-partner arm, which is what the source paper's own 'lineage-independent MLL-r signature' framing predicts.
This is a prediction, not a measurement. No one has seen this protein in this shape. The colours say how much the software trusts itself, and that is worth reading before anything else — the orange stretches are places the model is effectively saying this part has no fixed shape.
I have never read one of these pictures before
A protein is one long chain of amino acids that folds up. The ribbon traces that chain from one end to the other; it is not a wire, it is a simplification of where the backbone runs. Drag to rotate, scroll to zoom.
Domains are the stretches of chain that do a particular job, and they are what the coloured blocks above the picture mark out. Click one and the view flies to it. In this project the important question about any domain is usually whether the fusion still has it — the labels say lost where it does not.
Anything in yellow sticks is a bound small molecule, which is usually the drug.
The normal-tissue distribution, in two atlases
| atlas | peak normal tissue | value | whole-body median |
|---|---|---|---|
| HPA consensus (51 tissues) | basal ganglia | 297.8 nTPM | - |
| GTEx v8 (54 tissues) | Brain_Nucleus_accumbens_basal_ganglia | 323.89 TPM | 1.21 TPM |
| GTEx v8 | Brain_Caudate_basal_ganglia | 134.31 TPM | |
| GTEx v8 | Brain_Putamen_basal_ganglia | 85.09 TPM | |
| GTEx v8 | Brain_Frontal_Cortex_BA9 | 48.51 TPM |
HPA and GTEx agree independently and the agreement is on the same structure 23. Outside the brain LAMP5 is close to absent - the GTEx whole-body median is 1.21 TPM - so it is an unusually clean target if the brain does not count, and the treated population is infants whose basal ganglia are developing. Normal lymphohaematopoietic tissue is low: bone marrow 2.1, lymph node 5.3, tonsil 4.9 nTPM 2.
The gene's own name records this: BAD-LAMP, brain and dendritic cell-associated LAMP. The brain expression is not a surprise finding, which raises the question of whether it was weighed.
| uniprot | Q9UJQ1 |
|---|---|
| length AA | 280 |
| ensembl | ENSG00000125869 |
| alphafold ID | AF-Q9UJQ1-F1 |
| hpa protein class | Predicted membrane proteins |
| all 1206 cell lines mean nTPM | 8.41 |
| normal basal ganglia nTPM | 297.8 |
| normal bone marrow nTPM | 2.1 |
| gtex v8 median of 54 tissues | 1.21 |
| shown for fusion | MLL-rearranged as a partner-agnostic class (PMID 33910331); nothing in KMT2A::AFF1 specifically. This beat measures 1.18x over the KMT2A-other-partner arm, i.e. no AFF1 specificity. |
domains
[
{
"name": "Extracellular topological domain",
"start": 30,
"end": 235
}
]index arm nTPM — the numbers behind the chart
{
"SEM": 120.1,
"RS4;11": 284.4,
"MV4-11": 372.4
}gtex v8 peak
{
"tissue": "Brain_Nucleus_accumbens_basal_ganglia",
"tpm": 323.89
}- LAMP5 is among the most brain-enriched proteins in the body and this is an infant disease. Any LAMP5-directed modality that reaches the CNS meets a developing basal ganglion, and CNS-directed therapy is already routine in KMT2A-rearranged infant ALL.
- LAMP5 is a direct fusion target that the leukaemia depends on for growth - which means selective pressure against it produces LAMP5-low escape variants that may simply grow more slowly rather than not at all.
- The same paper shows LAMP5 loss increases type-I interferon signalling. An antigen whose removal turns on interferon could be doing part of a cell therapy's work for it, or could be inducing an inflammatory state in a child already at risk of cytokine release syndrome.
- How much LAMP5 is at the plasma membrane at steady state, as opposed to in endosomes and the ERGIC? UniProt lists six membrane compartments and a CAR needs the first one.
- Is LAMP5 protein detectable on neurons in a developing infant basal ganglion, or is the mRNA transcribed into a protein that stays in synaptic vesicles? The distinction decides the toxicity argument entirely.
- Does LAMP5 expression differ between KMT2A partner fusions in patient material? This beat's 1.18x partner ratio is from five cell lines and is the weakest number in the entry.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
LAMP5 was identified by filtering leukaemia expression repositories through the in silico human surfaceome as highly and specifically expressed in MLL-rearranged leukaemia; it is a direct target of the oncogenic MLL fusion protein; its depletion inhibits leukaemia growth in vitro and in vivo; and it modulates innate-immune signalling, with downregulation inhibiting NF-kB and increasing type-1 interferon activation.
LAMP-5 is an essential inflammatory-signaling regulator and novel immunotherapy target for mixed lineage leukemia-rearranged acute leukemia. · Haematologica 2022What this evidence cannot show
MLL-rearranged as a class; the signature is explicitly lineage-independent and partner-agnostic, so nothing in it is specific to KMT2A::AFF1. Abstract read in full, figures not read on this beat.
Evidence that argues against this (1)
- This beat's independent measurement puts LAMP5 at only 1.18x the KMT2A-rearranged-other-partner arm, consistent with a pan-MLL rather than an AFF1 antigen, and at 2.68x a B-precursor comparator, below the 4x fusion-enrichment threshold used here. Oncofusion Knowledge Graph 2026
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
LAMP5 is 120.1 nTPM in SEM, 284.4 in RS4;11 and 372.4 in MV4-11 against 8.41 across all 1,206 HPA cell lines, and reaches 297.8 nTPM in normal basal ganglia while sitting at 2.1 in bone marrow and 5.3 in lymph node.
Human Protein Atlas, RNA cell line nTPM (rna_celline.tsv), 1,206 human cell lines; HPA release v25.1, file Last-Modified 2025-11-05, fetched 2026-09-23 · Human Protein Atlas 2026What this evidence cannot show
One nTPM per cell line, no replicates. Adult bulk brain tissue for the normal value; the infant basal ganglia is not measured by either atlas and is developmentally active.
Evidence that argues against this (1)
- High neural mRNA does not establish that LAMP5 protein is surface-exposed and antibody-accessible on neurons behind an intact blood-brain barrier; LAMP5 is substantially an intracellular-vesicle protein by UniProt's own annotation. UniProt 2026
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
GTEx v8 independently places LAMP5 at 323.89 TPM in nucleus accumbens (basal ganglia), 134.31 in caudate and 85.09 in putamen, with a whole-body median across 54 tissues of 1.21 TPM.
What this evidence cannot show
Adult post-mortem tissue; GTEx v8 contains no bone marrow so it cannot address the treated compartment.
Evidence that argues against this (1)
- A whole-body median of 1.21 TPM means LAMP5 is exceptionally quiet everywhere except brain, so a modality that cannot cross the blood-brain barrier might have an excellent therapeutic index; the brain value is a conditional objection, not an absolute one. GTEx Portal 2026
- E4established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
LAMP5 is UniProt Q9UJQ1 (LAMP5_HUMAN), 280 amino acids, annotated Cell membrane with an extracellular topological domain at residues 30-235 and additional endosomal, ERGIC and synaptic-vesicle membrane locations.
What this evidence cannot show
The topological domain is ECO:0000255 (sequence-analysis inference), not experimental. Several of the location comments are ECO:0000250 by-similarity from the mouse orthologue rather than human measurements.
Evidence that argues against this (1)
- UniProt lists Cell membrane alongside five intracellular membrane compartments and notes recycling from the endocytic recycling compartment. A protein that spends most of its time in vesicles may present very little epitope at steady state, which cuts against it as a CAR target independently of the brain issue. UniProt 2026
normal-tissue adjudication of the gene entry
LAMP-5 routes innate-immune signalling from type-I interferon toward NF-kB; the cell state this graph already held
- ← regulatesKMT2A::AFF1 (MLL-AF4), t(4;11)(q21;q23)
LAMP5 is a direct target of the oncogenic MLL fusion protein (partner-agnostic MLL-r signature)
- expressed in →KMT2A-rearranged acute lymphoblastic leukaemia
highly and specifically expressed in MLL-rearranged leukaemia; SEM 120.1, RS4;11 284.4, MV4-11 372.4 nTPM against 8.41 across 1,206 lines
the CNS exposure that the normal-tissue axis reveals