One blot: phospho-eIF2alpha and ATF4 protein in RS4;11 under VTP-50469, with a dose-response and a DOT1L comparator arm
also called the ATF4 protein blot · phospho-eIF2alpha under a menin inhibitor
Every claim this project makes about menin-blocking drugs switching off a stress programme rests on gene-activity measurements that cannot see the step where the switching actually happens. One western blot, on cells two laboratories already grow, would settle it.
The same thing, for a scientist
**No phospho-eIF2alpha or ATF4-protein measurement exists under a menin inhibitor in any cell, in any system, anywhere** - bounded by a PubMed query returning 6 fusion-scoped ATF4 records and 1 menin-drug-scoped record, all six titles read, none of them a protein measurement [E2]. Meanwhile this graph asserts "loss of the integrated stress response" from transcript data in which **ATF4's own transcript is flat at -0.86 log2 while its targets fall up to 5.11** [E1], and a 3-prime-end RNA-seq assay is blind to the 5-prime leader where ATF4's regulation lives. **The experiment is a single SDS-PAGE gel with five antibodies, run on RS4;11 cells under VTP-50469 with a matched EPZ5676 arm and a dose-response**, and it decides between four mechanisms the transcript data cannot separate: translational control via eIF2alpha phosphorylation, increased ATF4 degradation, loss of a co-activator or of chromatin access at target loci, and the targets falling for reasons unrelated to ATF4. **Cost tier: low.** RS4;11 is a standard line, VTP-50469 and EPZ5676 are commercially available, and the antibodies are all catalogue reagents. **Timeline: three weeks including the dose-response.**
Design
Cells. RS4;11 (infant t(4;11) KMT2A::AFF1 B-ALL) as the primary line, because it is the one line in both public datasets and the line every number in this beat comes from. SEM as a second t(4;11) line, because three of four array-measured cascade members disagree between RS4;11 and SEM and that discrepancy is unexplained. REH (KMT2A wild-type) as the negative control - and this arm is load-bearing, because the potent-compound dataset contains no wild-type line at all, so the fusion-dependence of this programme has only ever been tested with a weak compound.
Arms. VTP-50469 at 33, 100, 330 and 1000 nM; EPZ5676 at 100, 330, 1000 and 3000 nM; DMSO. The dose-response is not optional - it is the control for the single most serious alternative explanation of this beat's main finding, which is that the two compounds agree on the ISR arm because both are saturating it.
Timepoints. 6 h, 24 h, 48 h, day 4, day 7. Both public datasets sample only day 2 and day 7, which cannot separate a direct effect from a consequence of arrest, and the 6-24 h window is where a direct transcriptional component would appear.
Readouts, in priority order.
- Phospho-eIF2alpha (Ser51) and total eIF2alpha - the quantity the whole hypothesis turns on.
- ATF4 protein - the quantity the graph has been inferring. Run beside ATF4 mRNA by qPCR from the same lysate split, so the protein/mRNA dissociation is measured within-sample rather than across experiments.
- CHAC1 and TRIB3 protein, to confirm the transcript collapse transmits.
- Phospho-S6 or phospho-4E-BP1, because SESN2 (the mTORC1 brake), SLC7A5/SLC3A2 (the leucine importer) and RHEBL1 all fall together and the net mTORC1 direction is unpredictable and therapeutically actionable.
- MEIS1 and MEF2C protein or mRNA as the positive control for menin engagement, since those are the genes that discriminate the two compounds.
Controls. Tunicamycin as a positive ISR inducer; ISRIB or a GADD34 inhibitor as an eIF2alpha-axis manipulation; and total protein plus a housekeeping loading control, because the drug arrests growth and therefore changes cell size and total protein per cell.
How each outcome is read
| phospho-eIF2alpha | ATF4 protein | ATF4 mRNA | reading |
|---|---|---|---|
| down | down | flat | The hypothesis, confirmed. Translational control; the ISR is being switched off upstream of ATF4; "loss of the ISR" is a correct description of a cause |
| flat | down | flat | ATF4 protein is lost by degradation or reduced synthesis independent of eIF2alpha - a different mechanism with different drug implications (ATF4 is an SCF-betaTrCP substrate and p300-stabilised) |
| flat | flat | flat | The hypothesis is dead. ATF4 is intact and the lesion is downstream - at a co-activator, at chromatin access at the target loci, or the targets are falling for reasons unrelated to ATF4. This is the most informative negative result available |
| up | down | flat | Paradoxical, and the most interesting outcome: the ISR is being engaged while its output is blocked, which would point squarely at chromatin access at the ATF4 target loci - consistent with this beat's finding that a DOT1L inhibitor reproduces the whole arm |
| any | down | down >1 log2 at 6-24 h | The transcript null in this beat is a late-timepoint artefact, and Jimenez 2021's transcriptional component is real |
And the dose-response reads independently of all of that. If the ISR genes separate between VTP-50469 and EPZ5676 at sub-saturating doses, the attribution half of the hypothesis fails and menin specificity is restored. If they track together across four doses each while MEIS1 separates, it holds.
| design | Western blot panel plus matched qPCR on RS4;11 (primary), SEM (second t(4;11) line) and REH (KMT2A wild-type control), under VTP-50469 at 33/100/330/1000 nM and EPZ5676 at 100/330/1000/3000 nM versus DMSO, harvested at 6 h, 24 h, 48 h, day 4 and day 7, with protein and RNA taken from the same lysate split so the protein/mRNA dissociation is measured within-sample. |
|---|---|
| readout | Priority order: (1) phospho-eIF2alpha Ser51 and total eIF2alpha; (2) ATF4 protein beside ATF4 mRNA from the same lysate; (3) CHAC1 and TRIB3 protein; (4) phospho-S6 or phospho-4E-BP1 for the mTORC1 direction; (5) MEIS1 and MEF2C as the positive control for menin engagement. |
| controls | Tunicamycin as a positive ISR inducer; ISRIB or a GADD34 inhibitor as an eIF2alpha-axis manipulation; total protein plus a housekeeping loading control, because the drug arrests growth and changes total protein per cell; REH as the fusion-negative arm - which is load-bearing because the potent-compound public dataset contains NO KMT2A-wild-type line, so this programme's fusion-dependence has only ever been tested with the weak compound MI-2-2. |
| cost tier | low |
| timeline | three weeks including the dose-response and the early timepoints |
| why the dose response is not optional | It is the control for the single most serious alternative explanation of this beat's main finding - that VTP-50469 and EPZ5676 agree on the ISR arm because both saturate it at the two arbitrary single doses used in the deposited experiments. If the ISR genes separate at sub-saturating doses, menin specificity is restored and the attribution half of hyp-the-isr-collapse-is-a-fusion-programme-readout-not-a-menin-interface-readout fails. |
| prerequisite | OBTAIN THE FULL TEXT OF PMID 33741715 FIRST. Its abstract attributes menin-dependent ATF4 regulation to post-transcriptional mechanisms in MLL-rearranged B-ALL, language hard to justify without a protein measurement - so an ATF4 western under a menin inhibitor may already exist in its figures, in unnamed MLL-rearranged B-ALL lines, and would narrow this experiment considerably. |
| localization | NOT APPLICABLE as designed - a western blot on whole-cell lysate has no positional resolution. IF a nuclear/cytoplasmic fractionation is added at no extra cost, it would also test whether ATF4 is redistributed rather than lost, and would be the first localisation measurement of ATF4 in any KMT2A-rearranged cell. That addition is recommended and is not required for the primary readout. |
| subnuclear localization | NOT ADDRESSED BY THIS DESIGN. The subnuclear question - whether ATF4 occupies nuclear speckles in a KMT2A-rearranged cell, and whether it shares that compartment with AFF1 or AFF4 - requires immunofluorescence and is a separate experiment. It is worth noting that ATF4's only annotated subnuclear pool is p300-dependent (PubMed 16219772) and has never been imaged in a leukaemia cell, so the prior question is also open. |
| cell type | RS4;11 (infant t(4;11) KMT2A::AFF1 B-ALL) as primary, SEM (t(4;11) KMT2A::AFF1) as a second fusion-positive line chosen because it DISAGREES with RS4;11 on three of four array-measured cascade members, and REH (KMT2A wild-type ETV6::RUNX1 B-ALL) as the fusion-negative control. All three are standard, widely distributed lines. A patient-derived xenograft arm would be required to address the leukaemia-stem-cell question raised by PMID 37496439 and is explicitly OUT of scope here. |
| cell state | The design exists partly to control cell state, which is the largest uncontrolled variable in the public data. Requirements: report medium composition and amino-acid supply; feed on a fixed schedule; record cell density at every harvest; normalise to total protein AND a housekeeping control because the drug arrests growth and changes protein per cell; and include the 6 h and 24 h timepoints so a direct effect can be separated from a consequence of arrest. The public datasets sample only day 2 and day 7 and report none of this metadata. |
| epigenetic readout | NOT MEASURED BY THIS DESIGN, and that is a deliberate scope limit. The attribution half of the hypothesis - that the ISR arm sits downstream of H3K79-methylation-dependent elongation output - is tested here only pharmacologically, by whether the EPZ5676 arm tracks VTP-50469 across a dose-response. The measurement that would settle it is ChIP of KMT2A, menin and H3K79me2 at the CHAC1, TRIB3, ATF5 and STC2 loci in RS4;11 under both drugs, which does not exist for any of the 72 panel loci in any cell and is the natural follow-on experiment. |
| shown for fusion | DESIGNED FOR KMT2A::AFF1 specifically - two independent t(4;11) lines plus a KMT2A-wild-type control - which would make it the first protein-level ISR measurement in the index fusion. Nothing comparable exists: the only functional ISR-axis result in any KMT2A fusion is murine MLL-AF9 (PMID 37496439, PERK arm, LSC self-renewal), and the only menin-ATF4 result is in Ewing sarcoma with a one-sentence extension to unnamed MLL-rearranged B-ALL lines (PMID 33741715). |
| absence class | literature-absence-bounded |
| absence class note | RECLASSIFIED 2026-09-23 (cycle 76). This field held free text, which `scripts/check-absence-class.py` rejects; the prose is preserved here VERBATIM and unsummarised because it is more informative than the token that replaced it. Token chosen: literature-absence-bounded - the two-part prose says MEASURED-AND-<something> for the data arm and BOUNDED-UNPUBLISHED for the prose arm. The data arm is a RESULT, not an absence; the absence is the bounded literature, and the bound's own limits (titles and abstracts only, per rule 80) are stated in the preserved prose. ORIGINAL absence_class, verbatim: «BOUNDED-UNPUBLISHED, and the bound is over titles and abstracts only. No phospho-eIF2alpha, ATF4-protein, ribosome-profiling or polysome measurement under a menin inhibitor appears in the 6 fusion-scoped or 1 menin-drug-scoped ATF4 records, all titles read - but rule 80 applies with unusual force because a phospho-western is exactly the panel that never reaches an abstract, and PMID 33741715's full text was not obtained.» |
outcome table
{
"pS51 down + ATF4 protein down + ATF4 mRNA flat": "the hypothesis confirmed: translational control, ISR switched off upstream of ATF4",
"pS51 flat + ATF4 protein down + mRNA flat": "ATF4 lost by degradation or eIF2alpha-independent reduced synthesis - a different mechanism (ATF4 is an SCF-betaTrCP substrate and p300-stabilised)",
"pS51 flat + ATF4 protein FLAT + mRNA flat": "THE HYPOTHESIS IS DEAD and this is the most informative negative available: the lesion is downstream of ATF4 - co-activator, chromatin access at the target loci, or the targets falling independently of ATF4",
"pS51 UP + ATF4 protein down": "paradoxical and the most interesting: ISR engaged while output blocked, pointing at chromatin access at ATF4 target loci - consistent with a DOT1L inhibitor reproducing the whole arm",
"ATF4 mRNA down >1 log2 at 6-24 h": "this beat's transcript null is a late-timepoint artefact and Jimenez 2021's transcriptional component is real"
}tests hypotheses
[ "hyp-the-isr-collapse-is-a-fusion-programme-readout-not-a-menin-interface-readout" ]
- A negative result here (ATF4 protein flat) would invalidate a mechanistic story this graph has repeated across several entries, and it is a genuinely likely outcome - roughly as likely as the positive one. The experiment is worth running precisely because it can go either way, and the entries that would need rewriting should be identified before it runs, not after.
- The blot cannot distinguish 'ATF4 protein fell because the ISR switched off' from 'ATF4 protein fell because the cells stopped growing'. A drug that arrests the cell cycle lowers global translation, so the loading control and the total-protein normalisation are doing more work than usual, and a non-ISR translational-load control (any other uORF-regulated protein, or global puromycin incorporation) should be on the same gel.
- If the answer is that the ISR is engaged while its output is blocked - the paradoxical outcome - then the whole framing of this beat is wrong in an interesting way, and the follow-on is a ChIP experiment rather than any further protein work.
- Does PMID 33741715's full text already contain an ATF4 western under a menin inhibitor in MLL-rearranged B-ALL? Obtaining it is a prerequisite and could shrink this experiment to a dose-response plus the phospho-eIF2alpha lane.
- Which ATF4 antibody performs in RS4;11? ATF4 westerns are notoriously antibody-dependent and a validated reagent plus an ATF4-knockout or knockdown lysate control is a real practical prerequisite, not a formality.
- Should a nuclear/cytoplasmic fractionation be added? It costs almost nothing, tests redistribution rather than loss, and would be the first ATF4 localisation measurement in any KMT2A-rearranged cell.
- Is a patient-derived xenograft arm justified? The leukaemia-stem-cell prediction from PMID 37496439 - that lowering this axis might protect the LSC compartment - cannot be addressed in bulk cell lines, and it is the prediction with the most clinical weight.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset
In RS4;11, ATF4's own transcript moves 0 of 1 array probes and 0 of 4 RNA-seq line/day contrasts past 1 log2 (max -0.86) on a transcript expressed at 13.72 log2 RMA, while CHAC1 falls 5.11, STC2 4.40 and TRIB3 3.10 in the same samples, and all four eIF2alpha kinases plus eIF2alpha itself are flat within 0.38 log2 - so the transcript data cannot distinguish translational control, increased ATF4 degradation, loss of a co-activator, or the targets falling independently of ATF4.
What this evidence cannot show
Established as a statement about mRNA in these samples. The four candidate mechanisms are named as a complete-enough set for designing the blot, not as an exhaustive partition of possibility space. The assay used is 3-prime-end RNA-seq, which quantifies one window per gene and is blind to the 5-prime leader where ATF4's regulation lives - which is precisely why the protein experiment is required rather than optional. RS4;11 array DMSO arm n=2.
Evidence that argues against this (1)
- If -0.86 log2 is a real 1.8-fold transcriptional fall rather than noise, then part of the answer is already transcriptional and the blot may show a smaller protein effect than the framing implies. The 6-24 h arm is included specifically to catch that, and the experiment is designed to be informative whichever way that lane reads. Jimenez JA et al. 2021
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
PubMed returns 6 records for ATF4 jointly with an eight-operand KMT2A-fusion list and exactly 1 for ATF4 jointly with a thirteen-name menin-inhibitor list (PMID 33741715), all six titles read; none of the six reports a phospho-eIF2alpha or ATF4-protein measurement under a menin inhibitor, and the two zeros checked in the same session carry non-empty querytranslations and warninglist 'No items found.'
NCBI PubMed esearch, run 2026-09-22, with querytranslation and warninglist read back from the response for every query quoted · NCBI PubMed 2026What this evidence cannot show
RULE 80 IS THE BINDING LIMIT HERE AND IT MATTERS MORE THAN USUAL: this bound is over titles and abstracts only, and a phospho-eIF2alpha western is exactly the kind of supporting panel that never reaches an abstract. The claim that no such measurement exists is therefore a claim about the PROMINENTLY PUBLISHED record. The one menin-drug-scoped record, PMID 33741715, was read at abstract level only and its full text was NOT obtained - if its MLL-rearranged B-ALL arm contains an ATF4 western, part of this experiment is already done and this entry overstates the gap. Rule 94 form; rule 95 observed; rule 67 controls non-empty in the same session.
Evidence that argues against this (1)
- PMID 33741715's abstract says menin inhibition causes 'loss of ATF4' in MLL-rearranged B-ALL and attributes menin-dependent ATF4 regulation to post-transcriptional mechanisms - language that is hard to justify without a protein measurement. So the most likely state of the world is that an ATF4 western under a menin inhibitor DOES exist, in that paper's figures, in unnamed MLL-rearranged B-ALL lines. Obtaining that full text is a prerequisite to running this experiment and would narrow it considerably. Jimenez JA et al. 2021
one blot plus a dose-response that separates translational control from degradation, co-activator loss and ATF4-independent target loss, and that controls the dose-artefact alternative
- related to →KMT2A::AFF1 (MLL-AF4), t(4;11)(q21;q23)
designed in two independent t(4;11) KMT2A::AFF1 lines plus a KMT2A-wild-type control, which would make it the first protein-level ISR measurement in the index fusion
- related to →Pinometostat (EPZ-5676) — the DOT1L inhibitor, and the most instructive failure in this field
carries a matched EPZ5676 dose-response arm, which is the control for the dose-artefact alternative to the beat's main finding
includes a phospho-S6 or phospho-4E-BP1 lane, because SESN2 (the mTORC1 brake), SLC7A5/SLC3A2 and RHEBL1 all fall together and the net mTORC1 direction is unpredictable and therapeutically actionable
the ATF4-protein and phospho-eIF2alpha western should be folded into this design rather than run separately