Back

CD59 organizes the plasma membrane to sustain oncogenic Ras-MAPK signaling and is a targetable vulnerability in acute myeloid leukemia

Maher, A.; Manikoth Ayyathan, D.; Cathelin, S.; Roehrig, P.; Liu, S. Z.; Yang, Y.; Liu, A. C. H.; Hosseini, M.; Quadri, E.; Villeneuve, T.; Kaur, S.; Schoof, E. M.; Wang, V.; Minden, M.; Marshall, C. B.; Schimmer, A. D.; Xie, S.; Dick, J. E.; Chan, S. M.

2026-08-12 cancer biology
10.64898/2026.08.11.744285 bioRxiv
Show abstract

Acute myeloid leukemia (AML) is a clinically heterogeneous disease. Although the genetic abnormalities associated with poor prognosis are well defined, how they drive unfavorable outcomes remains unclear. Using published gene-expression and dependency datasets, we searched for cell-surface protein-coding genes associated with poor survival and required for AML growth, prioritizing this class of proteins for its accessibility to biologics. This search identified CD59, a GPI-anchored protein with a canonical role in complement regulation, whose high mRNA expression correlates with adverse-risk genetics and stemness signatures. CD59 silencing impaired proliferation across genetically diverse AML cell lines, reduced leukemic burden, and extended survival in cell xenograft models. Moreover, CD59 expression was enriched on leukemic stem cells (LSCs), and its depletion impaired LSC self-renewal and primary AML engraftment in vivo while sparing normal hematopoiesis. Mechanistically, these effects reflected a non-canonical role for CD59 in sustaining Ras-MAPK signaling, whereby its loss depleted inner-leaflet phosphatidylserine and impaired Ras and c-Raf membrane recruitment and activation. rILYd4, a recombinant fragment of the bacterial toxin intermedilysin that binds and degrades CD59, recapitulated these effects and sensitized cells to venetoclax in vivo. These findings reveal CD59 as a critical regulator of Ras-MAPK signaling required for AML growth and nominate its rILYd4-mediated degradation as a therapeutic strategy.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.