Targeting NF-κB with First in Class N-Quinoline Benzenesulfonamides (NQBS) Reveals Potent Activity in Models of Diffuse Large B-Cell Lymphoma (DLBCL)

The first two authors contributed equally to this work

Identifying pharmacologic strategies to inhibit the activation of NF-κB and its target genes has been a major research pursuit. To date, no direct inhibitors of the NF-κB subunits have been explored in the clinic. Based on the constitutive activation of NF-κB in diffuse large B-cell lymphoma (DLBCL), we used this disease model to develop drugs targeting NF-κB. Using a fluorescence-based high throughput screening (HTC) approach, a unique N-quinoline-benzenesulfonamide (NQBS) scaffold was identified as potential small molecule inhibitor of the NF-κB pathway.

A confocal microscopy based HTC assay performed in human umbilical vein endothelial cells (HUVEC) identified hit compounds that contained a unique NQBS core structure. The assay screened for compounds that inhibited nuclear translocation of NF-κB subunits in TNFα-induced HUVEC cells. To date over 100 NQBS analogs have been synthesized with varying potency and cytotoxicity in inhibiting growth of DLBCL lines (OCI-Ly10, RIVA, HBL-1 and OCI-Ly3).

Cytotoxicity assays demonstrated that the most potent compounds exhibit IC50s in the 0.5 to 1.5 µM range. These most potent NQBS analogs identified as CU-O42 CU-O47 and CU-O75 were also able to induce apoptosis and caspase activation. Apoptosis was preceded by exclusion of the NF-κB proteins from the nucleus.

To analyze the localization of NF-κB proteins within the cell compartments before and after the treatment with CU-O42, CU-O47 and CU-O75, we used confocal microscopy, electromobility shift (EMSA) and ELISA assays. Control cells tested positive for p50/p65 both within the cytoplasm and the nucleus. Following treatment with CU-O42 NF-κB was sequestered within the cytoplasm of the cell which occurred as early as 3h after exposure. In addition, all three analogs reduced the nuclear levels of NF-κB in a concentration-dependent manner when measured by EMSA and ELISA.

Furthermore, CU-O47 and CU-O75 were able to inhibit TNFα induced luciferase expression in a HEK293T cell model where luciferase is controlled by an NF-κB promoter. A KINOMEscan platform (examining the activity of over 450 different kinases) showed that no NQBS analog screened (CU-O42 and CU-O75) inhibited any of the kinases in the assay. In addition, a proteasome inhibition assay tested negative for trypsin-like and chromotrypsin-like protease activity (CU-O42, CU-O47 and CU-O75).

Stabilization of the inactive trimer of p50, p65 and IκBα was hypothesized as a potential mechanism of action of CU-O42 and CU-O75 through Internal Coordinate Mechanics (ICM) software. This binding hypothesis was further corroborated by cellular thermal shift assays (CETSA) with an increase of the IκBα melting temperatures (2.5-3°C) in whole cell lysates following rapid (30min) exposure to CU-O42 and CU-O75. Using a genome-wide regulatory network perturbation analysis (DeMAND) based on the RNA-Seq data collected from OCI-Ly10 cells treated with CU-O75, we identified IκBα as one of the potential targets of the compounds. Gene set enrichment analysis demonstrated NF-κB target gene downregulation using IC₂₀ of CU-O75 at 24h (p=0.045).

In vivo experiments were conducted in two models: (1) xenografts with human DLBCL cell lines of both ABC and GC subtype; and (2) myc cherry luciferase mouse model where mice spontaneously develop aggressive lymphomas. In both models, CU-O42 was able to inhibit tumor growth. Interestingly, in the xenograft model, malignant cell growth was inhibited in both ABC (HBL-1) and GC (OCI-Ly1) cells when compared to controls (p=0.01 and p=0.02). However, overall survival of mice with ABC xenografts treated with CU-042 significantly exceeded the survival of mice with GC xenografts (p<0.01) suggesting a more sustainable response in this subtype of disease, consistent with its dependency on NF-κB.

Identification of a unique NQBS scaffold has led to the chemical synthesis of over 100 structural analogs with a potent inhibition on NF-κB nuclear translocation. They display potent activity across a panel of lymphoma cell lines, producing a survival benefit in mice implanted with an ABC-subtype of lymphoma. ICM, CETSA and DeMAND suggest that this is a direct effect mediated on the proteins within the p65/p50/IκBα complex. These findings point to a novel mechanism of action and warrant further research into potential clinical translation of this class of small molecules.