Design and functional characterization of a first-in-class irreversible inhibitor of HOIL-1-interacting protein (HOIP) with selective antitumor activity against B-cell non-Hodgkin lymphoma

Activating single-nucleotide polymorphisms of HOIL-1-interacting protein (HOIP), the catalytic subunit of the linear ubiquitin chain assembly complex (LUBAC), have recently been shown to promote myeloid differentiation primary response 88 (MYD88)-mediated B-cell lymphomagenesis. To assess the relevance of targeting HOIP in the activated B-cell-like (ABC) subtype of diffuse large B-cell lymphoma (DLBCL) with MYD88 mutation, we used systems biology to build a mathematical model aimed at evaluating the impact of HOIP depletion on DLBCL. From this model, an AI-mediated query demonstrated that HOIP blockade was associated with the suppression of three main pathophysiological motifs in malignant B cells, i.e., cell growth and proliferation, apoptosis evasion and deregulated metabolism. We then undertook a computational study that consisted of a combinatorial substitution of several α-β unsaturated moieties to be used as covalent binding warheads to the catalytic cysteine residue of HOIP. The resulting chemical library was used on subsequent molecular docking to assess the best HOIP binding candidates. Out of the four candidates synthesized from this library, we isolated compound A (Cpd A), a covalent irreversible inhibitor of HOIP with a pyrido[2,3-d]pyrimidine core, which exerted selective antitumor activity in a panel of ABCL-DLBCL cell lines (mean IC50 at 48hours: 90.7± 13.09 µM) while sparing normal B cells. Immunoprecipitation studies demonstrated that Cpd A was able to modulate the interaction between HOIP and the LUBAC component, SHARPIN, leading to the blockade of NF-κB signaling and to the downregulation of several downstream effectors, including CCL3, IL6 and IRF4. Specificity of Cpd A towards HOIP was confirmed by a drug affinity responsive target stability (DARTS) assay based on the immuno-detection of persistent HOIP peptides after Cpd A-mediated blockade of enzymatic proteolysis, and thereafter in a CRISPR-engineered HOIP-knockout ABC-DLBCL model. Finally, efficacy of the compound was confirmed in vivo in a chicken embryo chorioallantoic membrane (CAM)-derived model of ABCL-DLBCL subjected to a twice weekly dosing with Cpd A, in which the compound achieved a 25% tumor growth inhibition, associated with a 60.5% and 89% reduction in brain and bone marrow infiltration by lymphoma cells, respectively. Altogether, our results confirm that HOIP represents a promising therapeutic target for ABC-DLBCL with activating MYD88 mutation, and that a pyrido[2,3-d]pyrimidine derivative can successfully and specifically block HOIP, resulting in NF-κB disruption and selective antitumor activity in this aggressive subtype of B-cell lymphoma.