Oncofusion

start / graph / gene-tert

Geneunrated confidence 0 of 5: unrated

TERT — telomerase, switched on by the index fusion through HOXA7

also called TERT · telomerase reverse transcriptase · hTERT · EST2

Cancer cells need a way to keep re-lengthening the ends of their chromosomes. This one is switched on by the fusion protein this whole project is about, working through a gene it is already known to control.

The same thing, for a scientist

TERT is the catalytic subunit of telomerase. **This is one of the few entries in this graph whose primary evidence is in the index fusion**: MLL/AF4 suppression in t(4;11)-positive cells diminished telomerase activity and TERT expression, and knockdown of **HOXA7** — a direct MLL/AF4 target — reduced telomerase and TERT to a similar extent, with chromatin immunoprecipitation performed in **SEM** cells. The unbiased selectivity triage separately places TERT among the top ten genes both selectively required in KMT2A-rearranged lines and quiet across normal tissue, with a whole-body median of 0.0 TPM.

Go deeper
A rare case of index-fusion evidence

The partner-provenance audit found that 0 of 24 sampled entries rested on a dedicated human KMT2A::AFF1 experiment, and the unstated-systems sweep found that only 3 of 16 resolved systems touched the index fusion, all through SEM. This entry is on the right side of both counts: the experiments are MLL/AF4 knockdown in t(4;11)-positive cells with ChIP in SEM.

The mechanism is also routed through machinery this graph already holds. MLL/AF4 controls HOXA7; HOXA7 knockdown reproduces the TERT effect; so telomerase sits downstream of the HOXA program rather than beside it. That makes TERT a read-out of fusion activity as much as a target in its own right.

Its normal-tissue profile is the most extreme in the triage: a whole-body median of 0.0 TPM across 54 GTEx tissues, peaking at 3 TPM in testis. Adult somatic tissue is telomerase-silent, which is the textbook fact and the reason telomerase was pursued as a target for thirty years.

Facts and figures
depmap kmt2a selectivity
{
  "chronos_mean_kmt2a_rearranged": -0.3657,
  "chronos_mean_other_leukaemia": -0.0731,
  "difference": -0.2925,
  "pan_essentiality_all_lines": -0.0624,
  "selectivity_rank": 96,
  "rank_denominator": 17728,
  "provenance": "Joined 2026-09-21 from analyses/depmap-kmt2a-selective-scan-full (generated 2026-09-19T10:12:06Z), the project's own genome-wide scan over 11 KMT2A-rearranged and 43 other leukaemia DepMap lines. NOT re-fetched and NOT independent of that scan: every caveat it carries applies here, including that the KMT2A-rearranged arm is a small number of lines drawn from nine individuals, that the rank denominator is 17,728 genes, and that a difference is not a dependency. Added because only 26 of 308 gene entries carried this number, which made any KMT2A-selectivity question answerable for 8% of the corpus.",
  "how_to_read": "A negative difference means MORE required in KMT2A-rearranged lines than in other leukaemia lines. A rank near 17,728 is ANTI-selective. Neither is evidence of a fusion-driven dependency on its own: check pan_essentiality_all_lines before calling anything selective, and note that a paralogue-redundant gene can score near zero for reasons that have nothing to do with this disease - see ana-the-paralogue-redundancy-excuse-is-not-tested-and-my-proxy-decided-the-answer."
}
How this could bite us back
  • Telomerase inhibition is not a benign proposition in a child. The human phenotype for telomerase insufficiency is dyskeratosis congenita, with marrow failure, and the compartments that need TERT are invisible in the GTEx medians that make it look quiet.
  • Any telomerase strategy has a lag problem: telomeres shorten over divisions, so an effect on an acute leukaemia growing in weeks is not the timescale the mechanism operates on.
  • The index-fusion evidence here is a 2010 RNA-interference study whose full text nobody in this project has read.
Still open
  1. Read the full text of PMID 20686504 and record what the ChIP showed at the TERT locus. Whether MLL/AF4 or HOXA7 occupies TERT directly decides if this is regulation or correlation.
  2. Is TERT dependency in the 11-line panel separable from proliferation rate? A dependency that tracks doubling time is not a fusion-specific one.
  3. Does imetelstat have any published activity in KMT2A-rearranged leukaemia? The compound exists and is approved elsewhere, so this is a bounded literature question rather than a new experiment.
Every source behind this entry (2)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    MLL/AF4 suppression in t(4;11)-positive ALL cells diminished telomerase activity and TERT expression; blocking pro-apoptotic caspase activation alongside MLL/AF4 knockdown enhanced rather than relieved the inhibition of TERT expression, arguing the effect is not secondary to apoptosis; and knockdown of HOXA7, a direct transcriptional target of the MLL/AF4 fusion, reduced telomerase and TERT to a similar extent. Chromatin immunoprecipitation was performed in SEM cells.

    PMID 20686504DOI
    What this evidence cannot show

    Abstract read from the NCBI efetch record on 2026-09-20; full text not read, so effect sizes, the number of replicates and what the ChIP actually showed at the TERT locus are not recorded here and must not be quoted from this item. A 2010 knockdown study: RNA interference rather than degron or CRISPR, so off-target and partial-knockdown caveats apply. The paper's design pairs MLL/AF4 with AML1/MTG8 and this item records only the MLL/AF4 arm.

    Evidence that argues against this (2)
    • Telomerase is downstream of proliferation and self-renewal in almost every setting, so a fusion knockdown that stops self-renewal will reduce TERT whether or not the fusion regulates it specifically. The caspase-blocking control addresses apoptosis but not the more general confound that TERT tracks proliferative state, and HOXA7 knockdown reproducing the effect is equally compatible with both readings. Leukemia 2010
    • Thirty years of telomerase-inhibitor development has produced no approved agent in leukaemia, and the one clinical-stage compound, imetelstat, was approved in lower- risk MDS rather than in any acute leukaemia. A target being selectively expressed and mechanistically downstream of the driver has not, historically, been enough. Oncofusion knowledge graph (internal analysis) 2026
  2. E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.analysis

    DepMap gene effect in the 11 Cellosaurus-annotated KMT2A-rearranged leukemia lines is -0.366 against -0.073 in 43 other leukemia lines, a difference of -0.292, ranking 96 of 17,728 protein-coding genes scanned; mean effect across all lines is -0.062, so it is not pan-essential. GTEx v8 gives a whole-body median of 0.0 TPM across 54 tissues, a maximum of 2.7 TPM across brain, heart, blood, liver/kidney and endocrine tissue, and a peak of 3 TPM in Testis.

    What this evidence cannot show

    Selectivity read from the cached 17,728-gene scan on 2026-09-20 and expression queried from the GTEx API the same day; neither was recomputed here. **Eleven cell lines, none infant-derived**, carry the selectivity axis — the graph's standing weak spot — and GTEx is adult post-mortem tissue while the patients are infants. mRNA is not protein and neither is dependency. TERT's near-zero expression across adult tissue is the expected biology rather than a discovery: somatic telomerase silencing is textbook, and the triage is recovering it rather than revealing it.

    Evidence that argues against this (1)
    • TERT's low GTEx median is a poor guide to toxicity for the same reason it is striking: the tissues that do need telomerase — haematopoietic stem cells, intestinal crypts, germline — are small compartments invisible in a bulk tissue median. Telomerase inhibition is expected to hit exactly those, and dyskeratosis congenita is the human phenotype for telomerase insufficiency. Oncofusion knowledge graph (internal analysis) 2026
Connected to