Blood cancers endure relentless organellar stress—DNA damage, ER overload, and mitochondrial dysfunction—that pushes cells to the brink of apoptosis. Survival depends on pro-survival BCL-2 proteins, whose inhibition with venetoclax (VEN) delivers potent clinical responses. Yet resistance arises as malignant cells exploit alternative anti-apoptotic circuits.
Here we uncover a previously unrecognized safeguard: a tripartite complex of MARCH5, UBE2J2, and MFN2 that assembles at mitochondria–ER contact sites (MERCS). This stress-sensitive module integrates organellar damage signals and fortifies local BCL-2/BCL-XL function. When dismantled—by genetic deletion, disrupted protein–protein interaction, or stress itself—blood cancer cells become acutely vulnerable to BH3-mimetics and diverse cytotoxic insults.
Using an augmented CRISPR/Cas9 dropout screen in BH3-mimetic–resistant mantle cell lymphoma, we identified MARCH5 and UBE2J2 as synthetic lethal partners of BCL-2 family proteins. Loss of MARCH5 sensitized multiple hematologic malignancies (MCL, AML, DLBCL, MM), regardless of TP53 status, but spared solid tumors—except for selective vulnerability to BCL-XL inhibition. Importantly, MCL-1 was unaffected. MARCH5-deficient blood cancer cells exhibited striking intolerance to DNA damage, ER stress, and mitochondrial poisons.
UBE2J2 proved essential for this apoptotic restraint, but only in blood cancers. MARCH5 mutants unable to bind UBE2J2 failed to suppress apoptosis, underscoring the therapeutic potential and tissue selectivity of targeting this interaction. Proteomic and genetic reversal screens suggested diverse relevant substrates, yet mitofusin-2 (MFN2) emerged as indispensable. MFN2 bound both MARCH5 and UBE2J2, anchoring the complex at MERCS. Subcellular fractionation confirmed co-localization of this tripartite axis with BCL-2, BCL-XL, BAX, and BAK—but not MCL-1—providing a spatial explanation for selective sensitization. Loss of the MARCH5 C-terminal tail abolished MFN2 binding, enforcing dependence on BCL-2/BCL-XL and driving stress intolerance.
Physiologic stress rapidly disassembled MFN2 and UBE2J2 from MARCH5, lowering the apoptotic threshold. In vivo, disruption of MARCH5:MFN2 markedly prolonged survival with VEN in an otherwise refractory MCL xenograft model.
In summary, MARCH5:UBE2J2:MFN2 constitute a MERCS-based stress sensor that suppresses apoptosis under basal conditions but collapses under organellar damage, forcing addiction to neighbouring BCL-2/XL. This axis defines a hidden layer of apoptotic control in blood cancers and presents a tractable target for therapeutic development.