Can KT-253, an MDM2 degrader, eliminate malignant myelofibrosis stem cells?

Why destroying MDM2 differs from blocking it

About 96% of myelofibrosis patients carry wild-type p53, the tumor suppressor that can force damaged or aberrant cells into apoptosis. In myelofibrosis, hematopoietic stem and progenitor cells overexpress MDM2, an E3 ubiquitin ligase that tags p53 for proteasomal destruction, so p53 activity is suppressed even though the gene is intact. That makes MDM2 an attractive target: remove the brake and p53 should return.

Small-molecule MDM2 inhibitors such as AMG-232 bind MDM2 and stabilize p53, and they have shown clinical activity in myelofibrosis. But the paper explains a built-in limit: rising p53 transcriptionally drives more MDM2 production, a p53-dependent feedback loop that restores MDM2 levels and blunts the inhibitor's effect. KT-253 takes a different route. It is a targeted protein degrader that uses the ubiquitin-proteasome system to eliminate MDM2 protein itself, which the authors argue prevents that feedback restoration and sustains p53 activation.

p53 wild-type cells die, p53-mutant cells survive

The clearest test of mechanism used three JAK2V617F-positive cell lines: UKE-1, which has wild-type p53, and SET-2 and HEL, which carry p53 mutations. In UKE-1 cells, KT-253 was more than tenfold more potent than AMG-232 at reducing cell numbers, with a 2-day IC50 of 74 nM versus 740 nM, and about twentyfold more effective after 5 days (31 nM versus 730 nM). KT-253 increased the sub-G0/G1 fraction and reduced S-phase cells, and it induced apoptosis at tenfold lower concentrations than AMG-232.

Neither drug inhibited growth or increased apoptosis in the p53-mutant SET-2 and HEL lines. Protein analysis showed KT-253 depleted MDM2 within 0.5 hours, whereas a tenfold higher dose of AMG-232 progressively raised MDM2 levels over two hours. KT-253 produced larger early increases in phospho-p53, p53, p21 and PUMA, lowered the anti-apoptotic protein MCL-1, and raised cleaved caspase-3. In SET-2 cells KT-253 still lowered MDM2 protein, but p53 pathway readouts did not change — consistent with the authors' interpretation that the killing effect requires functional wild-type p53.

Patient CD34+ cells respond at tenfold lower dose

The clinically relevant test used primary CD34+ cells from nine myelofibrosis patients. Doses as low as 10 nM KT-253 induced apoptosis, while AMG-232 up to 500 nM did not produce comparable apoptosis. At 20 nM, KT-253 raised CDKN1A (p21) transcripts more than fourfold at 24 hours; a tenfold higher AMG-232 dose (200 nM) produced less than a twofold increase. Both agents raised BBC3 (PUMA), PMAIP1 (NOXA) and BAX transcripts and lowered MCL-1 and BCL2L1 transcripts, but p21 protein induction was markedly stronger with KT-253 even at tenfold lower doses.

These are short ex vivo exposures of patient cells, so they measure direct drug sensitivity of the progenitor compartment rather than what a patient would experience under systemic dosing. Within that boundary, the comparison is consistent: the degrader achieves stronger p53 target activation and apoptosis at concentrations an order of magnitude below the inhibitor.

Colony assays show selective loss of mutant progenitors

Colony formation assays ask whether progenitor cells can still grow into hematopoietic colonies. In nine myelofibrosis samples, 6.25 nM and 12.5 nM KT-253 reduced total colony numbers by 40% and 60%, while the same doses did not reduce colony formation by normal donor CD34+ cells (four donors). Tenfold higher AMG-232 doses produced only 20% and 30% reductions. A higher KT-253 dose did show toxicity to normal donor cells, so the selectivity is dose-dependent rather than absolute.

The mutation-specific readout is sharper. At 12.5 nM, KT-253 reduced JAK2V617F-positive colonies by 71%, whereas 125 nM AMG-232 reduced them by only 10%. In four of five patient cases, JAK2 wild-type colonies persisted after KT-253 treatment even as mutant colonies fell. The authors read this as a potential therapeutic window in which malignant clones are eliminated while non-mutated CD34+ cells are spared — an interpretation of the colony data, not a demonstrated clinical margin.

Mouse xenografts show depletion of repopulating cells

To test whether KT-253 reaches the stem cell compartment, the team pre-treated JAK2V617F-positive patient CD34+ cells with vehicle, KT-253 (25 nM or 100 nM) or AMG-232 (500 nM) for three days, then transplanted the surviving cells into NSGS mice and followed the animals for up to 15 weeks. KT-253 reduced CD34+ cell numbers by 20% and 70% at 25 nM and 100 nM, while more than 95% of cells persisted after AMG-232.

In the transplanted mice, 40% of animals receiving vehicle-treated cells and 20% receiving AMG-232-treated cells died by week 15, whereas all mice given KT-253-treated cells remained alive. Human CD45+ chimerism in bone marrow and spleen, multiple human lineage markers, and spleen weights were lower in recipients of KT-253-treated grafts. The JAK2V617F variant allele frequency of recovered human CD45+ cells fell only in the KT-253 groups, and bone marrow fibrosis was reduced in those recipients. The authors note that this phenotype combines targeted JAK2V617F clearance with broad stem cell cytotoxic pressure, and that secondary transplantation — a stricter stem cell assay — was not performed.

What the preclinical design cannot settle

The xenograft work exposed cells to drug before transplantation rather than treating recipient mice directly. That design cannot capture systemic pharmacokinetics, bone marrow microenvironment effects, or hematologic toxicities that would shape a clinical therapeutic window. The authors state that direct in vivo dosing in patient-derived xenograft or JAK2-mutant syngeneic models is needed to establish true in vivo selectivity and toxicity.

The paper also points to early clinical signals outside this study's own experiments: in a phase 1 trial in relapsed or refractory solid tumors and high-grade myeloid malignancies, the most common adverse events were gastrointestinal, with no evidence of neutropenia or thrombocytopenia, and two of five treated leukemia patients — both with MPN-associated secondary leukemia — achieved a complete or partial remission. Those observations support further evaluation but do not establish efficacy in myelofibrosis.

The priority is sustained MDM2 removal

The decision-relevant comparison in this paper is not simply degrader versus inhibitor, but what happens to MDM2 protein over time. AMG-232 raises MDM2 levels as p53 activity rebounds; KT-253 removes MDM2 and keeps p53 target genes induced. That difference appears in the potency gap — 31 nM versus 730 nM IC50 in UKE-1 cells at 5 days, apoptosis of patient CD34+ cells at 10 nM versus failure at 500 nM AMG-232, and 71% versus 10% reduction in JAK2V617F-positive colonies at a tenfold dose difference.

The practical reading is that MDM2 degraders merit clinical testing specifically in wild-type p53 myelofibrosis, where the target biology applies, and that the key question for trials is whether sustained MDM2 removal translates into deeper malignant stem cell depletion than inhibitors achieved. The p53-mutant cell line results mark the boundary: this strategy is not expected to work where p53 is already lost or mutated.

Sources used in this answer