Cardiomyocyte - the non-haematopoietic cell that two of this graph's drug targets are required in, which makes it the graph's clearest bystander-organ liability
also called cardiac myocyte · heart muscle cell · ventricular cardiomyocyte · embryonic cardiac ventricle
Heart muscle cells are not blood cells and have no part in leukaemia, but two proteins this project wants to drug turn out to be needed in them. That makes the heart the place where these drugs are most likely to do harm that has nothing to do with the disease.
The same thing, for a scientist
Q69 recorded a vocabulary gap for cardiomyocyte because HDGFL2's only cell type with a loss-of-function phenotype is the embryonic cardiac ventricle. Writing the node found a second and much more load-bearing case in the same compartment. **Cardiac-specific knockout of Dot1L in mice causes increased mortality with chamber dilation, increased cardiomyocyte cell death, systolic dysfunction and conduction abnormalities - a phenocopy of human dilated cardiomyopathy - through DOT1L-dependent transcription of Dystrophin and consequent stability of the dystrophin-glycoprotein complex, with miniDmd expression largely rescuing the phenotype** [E1]. DOT1L is the target of pinometostat, a drug this graph carries for KMT2A-rearranged leukaemia. The HDGFL2 result is the second: Hdgfrp2-knockout mice develop normally and are fertile, but Psip1/Hdgfrp2 double-knockout embryos die around E13.5 with a ventricular septal defect at E14.5, and TGF-beta signalling was the only pathway significantly deregulated by the double knockout [E2]. PSIP1/LEDGF is the KMT2A-fusion tethering factor this graph is built around, so its only documented lethal requirement - in combination with its paralogue - is cardiac. The third strand is pharmacological rather than genetic: selective MCL-1 inhibition raises plasma cardiac troponin in patients (4 of 8 in a first-in-human study) with no other corresponding cardiac findings, and the developer describes this as likely a **class effect** of MCL-1 inhibition [E3] - and MCL1 is a target this graph carries for infant KMT2A-rearranged disease.
The DOT1L result, and why it is the most consequential thing in this node
This graph holds drug-pinometostat, a DOT1L inhibitor, as one of the two main pharmacological routes into KMT2A-rearranged leukaemia. The cardiac genetics is unambiguous and mechanistically resolved: cardiac-specific Dot1L knockout mice show increased mortality, chamber dilation, increased cardiomyocyte death, systolic dysfunction and conduction abnormalities; the mechanism is DOT1L-dependent transcription of Dystrophin and therefore stability of the dystrophin-glycoprotein complex; and expressing a miniDmd construct largely rescues the phenotype, which is about as clean an epistasis result as this kind of study offers. The same paper notes DOT1L is downregulated in idiopathic dilated cardiomyopathy patient samples relative to controls 1.
What this is and is not. It is a genetic knockout, not catalytic inhibition, and the two are not the same for DOT1L any more than they are for other chromatin enzymes - a scaffolding function survives an inhibitor and does not survive a knockout. It is also mouse, and it is a complete, permanent, cardiac-wide deletion rather than weeks of partial pharmacological occupancy in a child. Those are real mitigations and they are why this is written as a liability to measure rather than a prediction of failure. But H3K79 methylation turns over slowly, pinometostat was dosed by continuous infusion, and the relevant question - what happens to Dystrophin transcription in a cardiomyocyte under a DOT1L inhibitor - has not been asked in any published work this beat could find.
The HDGFL2 result the gap was reported for, and its double negative
gene-hdgfl2 records this carefully and the double negative is the finding: Hdgfrp2-knockout mice develop normally and are fertile, and it is only the Psip1/Hdgfrp2 double knockout that dies, at approximately E13.5, with a ventricular septal defect associated with E14.5 embryos. RNA sequencing of ventricular tissue found that among the dozen genes encoding LEDGF/p75 binding factors only Nova1 was significantly deregulated, global splicing deregulation was not implicated, and TGF-beta signalling - which is central to cardiac morphogenesis - was the only pathway significantly deregulated by the double knockout relative to control and Psip1-knockout samples 2.
The interesting structure here is paralogue buffering: neither protein is individually required for cardiac development, and together they are. That is the same architecture this graph worries about on the leukaemia side, where HRP-2 is dispensable for MLL-rearranged leukaemogenesis while LEDGF/p75 is not. A therapy that hit both PSIP1 and HDGFL2 - which a shared-interface strategy would - would be removing a buffered requirement, and the tissue where that buffering is documented is the heart.
A pharmacological cardiac signal already measured in patients
Unlike the two genetic results, this one is human and clinical. In a first-in-human study of the selective MCL-1 inhibitor ABBV-467 in relapsed/refractory multiple myeloma (NCT04178902), treatment resulted in disease control in one patient and increases in plasma cardiac troponin in 4 of 8 patients, without other corresponding cardiac findings; the authors argue that the compound's selectivity implies the troponin release is a consequence of MCL-1 inhibition itself and therefore may represent a class effect of MCL-1 inhibitors in humans 3.
Troponin in plasma is, by definition, cardiomyocyte content that has left cardiomyocytes. This graph carries MCL1 as a target in infant KMT2A-rearranged disease, and risk-mcl1-inhibitor-cardiac-troponin-class-effect holds the risk argument in full; this node is the cell the risk is about.
What a cardiomyocyte is here, and the limits of the node
A terminally differentiated, post-mitotic striated muscle cell of the heart, with essentially no regenerative reserve after the neonatal period - which is the property that makes every claim above serious. Loss is not replaced.
This node makes no claim that cardiomyocytes have any role in leukaemia. Every result in it is either normal developmental biology or drug toxicity. That is the correct scope: a bystander-organ node exists so that on-target toxicity has somewhere to live in the graph, and so that a reader scanning a drug's page can reach the tissue that pays for it.
| ontology | CL:0000746 |
|---|---|
| ontology note | CL:0000746 is 'cardiac muscle cell'. CL:0002131 (regular cardiac myocyte) and CL:0002098 (regular ventricular cardiac myocyte) are the finer terms; the HDGFL2 phenotype is specifically ventricular. |
| key property | terminally differentiated and essentially post-mitotic after the neonatal period, with negligible regenerative reserve - which is why every requirement recorded here is a serious liability rather than a recoverable one |
| why this node exists | Q69 recorded a vocabulary gap: HDGFL2's only cell type with a loss-of-function phenotype is the embryonic cardiac ventricle (Psip1/Hdgfrp2 double knockout), and the graph had no cardiac node, so the answer was written into a field with nowhere to link. Writing the node surfaced the DOT1L cardiac knockout, which is the more consequential case because DOT1L is a drugged target here. |
| shown for fusion | NOT APPLICABLE in the usual sense, and the entry says so rather than writing a bare NO: cardiomyocytes have no role in leukaemia and this node makes no claim that they do. Every result here is normal developmental biology or drug toxicity in a bystander organ. What HAS never been measured is cardiac function or Dystrophin transcription under a DOT1L INHIBITOR, which is the fusion-relevant version of the question. |
| scope limit | This is a bystander-organ node. It exists so that on-target toxicity has a cell to attach to, not because the heart participates in this disease. |
markers
[ "TNNT2 (cardiac troponin T)", "TNNI3 (cardiac troponin I) - the analyte that rises under MCL-1 inhibition", "MYH6 / MYH7", "ACTN2", "NKX2-5 as the cardiac transcription factor", "dystrophin-glycoprotein complex at the sarcolemma - the structure DOT1L loss destabilises" ]
genes and targets in this graph with a cardiac requirement or signal
[
{
"entity": "gene-dot1l",
"finding": "cardiac-specific Dot1L knockout causes dilated cardiomyopathy with cardiomyocyte death and conduction abnormalities, via Dystrophin transcription; miniDmd largely rescues",
"pmid": "21289070",
"perturbation": "genetic knockout, NOT catalytic inhibition",
"drug_in_this_graph": "drug-pinometostat"
},
{
"entity": "gene-hdgfl2 with gene-psip1",
"finding": "double knockout is embryonic lethal at ~E13.5 with a ventricular septal defect and deregulated TGF-beta signalling; neither single knockout has the cardiac phenotype",
"pmid": "26367869",
"perturbation": "germline double knockout; paralogue-buffered requirement",
"drug_in_this_graph": "none - PSIP1 is a tethering factor, not a drugged target here"
},
{
"entity": "gene-mcl1",
"finding": "selective MCL-1 inhibition raised plasma cardiac troponin in 4 of 8 patients with no other cardiac findings; described by the developer as a likely class effect",
"pmid": "37880389",
"perturbation": "pharmacological, in humans - the only clinical cardiac signal in this node",
"drug_in_this_graph": "held in risk-mcl1-inhibitor-cardiac-troponin-class-effect"
}
]- Cardiomyocytes do not regenerate after infancy. Any on-target cardiac injury from a drug given to a child is a lifetime injury, and it compounds with anthracyclines, which are already in the backbone of every KMT2A-rearranged leukaemia protocol this graph carries.
- DOT1L's cardiac requirement runs through Dystrophin, the same protein whose loss causes Duchenne muscular dystrophy and its cardiomyopathy. That is a mechanism with a known, severe human phenotype behind it, not a theoretical concern - and it sits under a drug this project holds as one of two main routes into the disease.
- The MCL-1 troponin signal is already measured in humans and already called a probable class effect by the developer. Combining an MCL-1 inhibitor with anthracycline-containing induction in an infant stacks two cardiac liabilities, and the graph's own infant BCL-2/MCL-1 combination work does not measure cardiac endpoints.
- PSIP1 and HDGFL2 buffer each other in the heart. The leukaemia rationale for targeting the LEDGF interface is that it is a shared, conserved binding surface - which is precisely the property that would make a drug hit both paralogues and remove the buffering.
- None of these liabilities would be detected by the endpoints a leukaemia trial normally collects. Troponin is not routine, echocardiography is intermittent, and dilated cardiomyopathy in the cited model develops over time rather than acutely.
- Does a DOT1L INHIBITOR reduce Dystrophin expression in cardiomyocytes? The cardiac phenotype is established for gene knockout and the drug blocks catalysis while leaving the protein in place. Human iPSC-derived cardiomyocytes plus pinometostat, scoring DMD transcript and protein and contractile function, is a straightforward experiment that would convert an inference into a measurement - and it has not been published.
- Did the pinometostat trials measure cardiac endpoints? This beat did not establish what echocardiography, ECG or troponin monitoring those protocols carried. Given a published cardiac-knockout phenocopy of dilated cardiomyopathy for the target, the answer belongs in drug-pinometostat and is currently absent.
- Is the PSIP1/HDGFL2 cardiac requirement relevant to any realistic intervention? The phenotype requires loss of BOTH paralogues, and a shared-interface strategy is the only plausible way a therapy would do that. Whether a LEDGF-interface inhibitor engages HRP-2 at all is a chemistry question this graph has not answered.
- Is the MCL-1 troponin signal reversible permeabilisation or cumulative cardiomyocyte loss? The distinction matters enormously for a paediatric population that will live for decades after therapy, and 8 patients with no other cardiac findings cannot settle it.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
IN MICE, cardiac-specific knockout of Dot1L results in increased mortality with chamber dilation, increased cardiomyocyte cell death, systolic dysfunction and conduction abnormalities, phenocopying human dilated cardiomyopathy; mechanistically DOT1L acts in cardiomyocytes by regulating Dystrophin (Dmd) transcription and consequently the stability of the dystrophin-glycoprotein complex important for cardiomyocyte viability, expression of a miniDmd construct largely rescues the DCM phenotype, and DOT1L is downregulated in idiopathic DCM patient samples relative to normal controls.
DOT1L regulates dystrophin expression and is critical for cardiac function. · Nguyen AT, Xiao B, Neppl RL, Kallin EM, Li J, Chen T, et al., Zhang Y · Genes Dev 2011What this evidence cannot show
Murine conditional GENE KNOCKOUT, not catalytic inhibition - DOT1L has non-catalytic scaffolding roles that survive an inhibitor and do not survive a knockout, so this does not establish that a DOT1L inhibitor is cardiotoxic. It is complete cardiac-wide deletion from development rather than weeks of partial occupancy in an established heart. The human arm is an expression correlation in idiopathic DCM samples, not a causal result. Read: abstract only. Not previously cited in this graph.
Evidence that argues against this (1)
- The clinical experience with the DOT1L inhibitor pinometostat does not report this toxicity, and the distinction between genetic ablation and enzymatic inhibition is exactly where that difference would live: an inhibitor blocks H3K79 methyltransferase activity while leaving the protein in place, so a scaffolding contribution to Dmd transcription would be preserved. This is an argument from the mechanism of the two perturbations, not from a measurement - no published study has measured Dystrophin transcription or cardiac function in cardiomyocytes under a DOT1L inhibitor, so the reassurance is an absence of testing rather than a negative result. The graph's own drug-pinometostat entry is the place that evidence would sit. Wang H et al. 2015
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
IN MICE, Hdgfrp2-knockout animals develop normally and are fertile, whereas Psip1/Hdgfrp2 double-deficient mice die at approximately embryonic day 13.5, with histology revealing a ventricular septal defect in E14.5 double-knockout embryos; RNA sequencing of ventricular tissue showed that of the dozen genes encoding LEDGF/p75 binding factors only Nova1 was significantly deregulated, that global RNA-splicing deregulation was not a driving factor, and that TGF-beta signalling - a pathway central to cardiac morphogenesis - was the only pathway significantly deregulated by the double knockout relative to control and Psip1-knockout samples.
Embryonic Lethality Due to Arrested Cardiac Development in Psip1/Hdgfrp2 Double-Deficient Mice. · Wang H, Shun MC, Dickson AK, Engelman AN · PLoS One 2015What this evidence cannot show
Murine germline double knockout in embryonic development, which is the furthest thing from partial pharmacological inhibition in a child's established heart. The TGF-beta attribution is explicitly speculative in the authors' own wording. Note the symbol trap logged in data/symbol-collisions.json: the mouse gene is Hdgfrp2 and the human symbol HDGFL2 now returns only TDP-43 cryptic-exon papers, so this literature is unreachable from the current human symbol. Already cited in gene-hdgfl2 and in a PSIP1 isoform concept entry. Read: abstract only.
Evidence that argues against this (1)
- The single-knockout result is the more relevant one for any realistic intervention and points the other way: Hdgfrp2-knockout mice develop normally and are fertile, and Psip1 single knockout causes perinatal lethality without the cardiac arrest seen in the double. A therapy that reduced but did not abolish one of the two paralogues would be nowhere near the double-null condition, so the cardiac phenotype is a statement about the buffered pair rather than a prediction about inhibiting either protein. Read: abstract only. Nguyen AT et al. 2011
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
IN A FIRST-IN-HUMAN TRIAL (NCT04178902) of the highly potent and selective MCL-1 inhibitor ABBV-467 in patients with relapsed/refractory multiple myeloma, treatment resulted in disease control in one patient but also increases in plasma cardiac troponin in 4 of 8 patients without other corresponding cardiac findings, and the authors conclude that the compound's selectivity implies treatment-induced troponin release is a consequence of MCL-1 inhibition and may therefore represent a class effect of MCL-1 inhibitors in human patients.
Selective MCL-1 inhibitor ABBV-467 is efficacious in tumor models but is associated with cardiac troponin increases in patients. · Yuda J, Will C, Phillips DC, Abraham L, Alvey C, Avigdor A, et al., Souers AJ · Commun Med (Lond) 2023What this evidence cannot show
Eight patients, adults with multiple myeloma, not children and not leukaemia; troponin elevation without other cardiac findings is a biomarker signal whose clinical meaning at this n is unresolved. The class-effect conclusion is the developers' inference from selectivity, not a cross-compound comparison performed in this study, although risk-mcl1-inhibitor-cardiac-troponin-class-effect records a second compound with the same signal. Read: abstract only. Already cited in that risk entry, which holds the full argument.
Evidence that argues against this (1)
- A troponin rise with no other cardiac findings in 8 patients is a long way from demonstrated cardiomyocyte loss, and the alternative reading - that MCL-1 inhibition transiently permeabilises cardiomyocytes without killing them - predicts a reversible signal rather than cumulative damage. The genetic cardiac results in this node are permanent complete ablations from development and cannot arbitrate between those two readings, because neither measures troponin under partial pharmacological inhibition of an adult heart. The clinical question is unanswered in both directions. Read: abstract only. Nguyen AT et al. 2011
RESIDENCY: DOT1L performs its function in cardiomyocytes by regulating Dystrophin transcription, established by cardiac-specific knockout with miniDmd rescue; DOT1L is also downregulated in idiopathic dilated cardiomyopathy patient samples
REQUIREMENT: cardiac-specific Dot1L knockout causes chamber dilation, cardiomyocyte death, systolic dysfunction and conduction abnormalities phenocopying dilated cardiomyopathy. GENETIC KNOCKOUT, not catalytic inhibition - the distinction is load-bearing
- ← related toPinometostat (EPZ-5676) — the DOT1L inhibitor, and the most instructive failure in this field
UNMEASURED LIABILITY: the target's cardiac-specific knockout phenocopies dilated cardiomyopathy through Dystrophin, and no published work measures Dystrophin transcription or cardiac function in cardiomyocytes under a DOT1L inhibitor
the only cell type in which HDGFL2 loss produces a phenotype, and only with PSIP1 also absent: Psip1/Hdgfrp2 double knockout dies at ~E13.5 with a ventricular septal defect. The cell type gene-hdgfl2 named as missing vocabulary
paralogue-buffered cardiac requirement: neither Psip1 nor Hdgfrp2 single knockout gives the cardiac phenotype, the double knockout is embryonic lethal with a ventricular septal defect and deregulated TGF-beta signalling
the cell the risk is about: plasma cardiac troponin is cardiomyocyte content that has left cardiomyocytes, raised in 4 of 8 patients on a selective MCL-1 inhibitor and called a probable class effect by the developer
pharmacological cardiac signal in humans rather than a demonstrated cardiomyocyte requirement: selective MCL-1 inhibition raised plasma troponin with no other cardiac findings, and whether that is reversible permeabilisation or cell loss is unresolved