qHTS Assay to Find Inhibitors of Chronic Active B-Cell Receptor Signaling
Diffuse large B-cell lymphoma (DLBCL) is the most common category of non-Hodgkin lymphoma, with ~20,000 new cases per year in the United States. DLBCL should be regarded as a collection of different neoplasms. Microarray gene expression analysis has defined two principal types of DLBCL, named for their similarity to normal B cell stages or activation states: a germinal center B-cell (GCB) type, more ..
BioActive Compounds: 168
NIH Molecular Libraries Probe Production Network [MLPCN]
NIH Chemical Genomics Center [NCGC]
MLPCN Grant: 1 R03 DA031086-01A1
PI Name: Dr. R. Eric Davis
NCGC Assay Overview:
Diffuse large B-cell lymphoma (DLBCL) is the most common category of non-Hodgkin lymphoma, with ~20,000 new cases per year in the United States. DLBCL should be regarded as a collection of different neoplasms. Microarray gene expression analysis has defined two principal types of DLBCL, named for their similarity to normal B cell stages or activation states: a germinal center B-cell (GCB) type, and an activated B-cell (ABC) type. In ABC-DLBCL, a constitutive high-level activity of the NF-kappaB family of transcription factors has been observed (1) and NF-kappaB target genes are more highly expressed in ABC primary tumors than in GCB tumors. Functional studies have shown a constitutive high-level activity of the IkappaB (inhibitor of kappa-B) kinase (IKK) complex. Using a loss-of-function short hairpin RNA (shRNA) interference screen in ABC-DLBCL lines, it has been found that IKK activation in all ABC-DLBCL lines requires CARD11, MALT1, and Bcl-10 (2). Further analysis of DLBCL primary tumors for NF-kappaB target gene expression imply that although CARD11 is required for constitutive IKK activity in all ABC-DLBCL lines, mutations in CARD11 may be the root cause of IKK activation only in lines or tumors with such mutations. In other words, CARD11 mutations may explain NF-kappaB activation in only 10% of ABC-DLBCL cases. In subsequent work (3), a mechanism for NF-kappaB activation has been found that occurs in at least another 20% of ABC-DLBCL cases: the B-cell antigen receptor (BCR) is spontaneously active in all ABC-DLBCL lines with WT CARD11, leading to IKK activation through proteins known to be required for BCR surface expression and transmission of its signals: CD79a, CD79b, BTK, and (variably) SYK. This process has been termed 'chronic active BCR signaling' (CABS).
The goal of this screen is to find small-molecule inhibitors of the CABS pathway found in ABC-DLBCL lines using a cell-based assay that measures IkappaBalpa stability, which has already been extensively validated in the OCI-Ly3 line (PubChem AID:445; REF: 4). The BCR-dependent ABC-DLBCL cell line TMD8 (which has a Y196C mutation in CD79b) shows activation of WT CARD11, and ultimately NF-kappaB through IkappaBalpa degradation in a CABS dependent mechanism
A cell-sensor assay for I-kappa-B-alpha stabilization has been developed. The assay uses green (CBG68) and red (CBR) emitting beetle luciferases [Chroma-GloTM developed by Promega Corp.] where the green luciferase is fused to IkappaBalpha and the red luciferase is present in its native state. Both luciferase genes are stably present in the TMD8 human ABC DLBCL cell line, but their expression is under the control of tetracycline-inducible promoters. Following the addition of doxycycline to induce expression of the luciferases, the green luminescence in control cultures rises only slowly because the CBG fusion partner IkappaBalpha is targeted by IKK activity for rapid degradation. In contrast, green luminescence rises rapidly in the presence of proteasome inhibitors or IKK inhibitors. The rise in red luminescence of CBR is not affected by proteasome or IKK inhibitors, and serves as normalization for cell number and nonspecific effects. Compounds were screened as a concentration-titration series that ranged from 46 uM to 3 nM.
NCGC Assay Protocol Summary:
Cells were seeded in 1536-well plates at 5000 cells/4uL in IMDM medium containing 0.5% FBS, w/o phenol red, 1% penicillin/streptomycin. Then, 23 nL of compounds or DMSO were delivered to each well using a pin tool. One uL 20ng/ml doxycycline was then dispensed into white solid 1536-well plates and the plates were incubated at 37C/5% C02 in a cell incubator for 4 hrs. Then 4 uL luminescent substrate mix (in-house prepared: 150 mM Tris-Cl, pH 7.6, 3 mM MgCl2, 1% Triton X-100, 5 mM DTT, 0.2 mM CoA, 0.15 mM ATP, 5 mM D-luciferin, final concentrations) was added to each well. The plate was incubated at r.t. for 10-15 min. The plates were measured on a ViewLux plate reader for green luminescence (540/20 nm filter) and red luminescence (618/8 nm filter). The green and red luminescent signals were corrected for red and green luminescence filter overlap using previously determined filter calibration constants (determined using the procedure outlined in the Promega Technical Manual; REF 4). The %Activity was determined from the ratio of corrected green/red luminescent values. MG132 were used as control. Green luminescent and ratio %Activity was determined by normalizing to the difference in signal between basal cells (0% Activity) and cells incubated with 10 uM of the proteasome inhibitor MG-132 (100% Activity). Red luminescent %Activity was normalized to the difference between red luminesence in basal cells (0%Activity) and zero luminescence (-100% Activity).
Concentration-response curves were fitted to the signals arising from the green luminescence (I-kappa-B-alpha-luciferase fusion; IkB-luc) and the red luminescence (unfused luciferase; luc) as well as the calculated ratio. The concentration-effect curves were then classified based on curve quality (r2), response magnitude and degree of measured activity. Compounds were then categorized based on their concentration-response curves for all three datasets (IkB-luc, luc and ratio). Inhibitors showed concentration-dependent decrease in the IkB-luc and ratio data with little or no effect on the luc data. Activators showed concentration-dependent increases in both the IKb-luc and luc data with either no effect or a modest effect on the ratio data. Inactive compounds showed no effect in all three datasets.
Keywords: IKK, I-kappa-B-alpha, IkB, IkBa, NF-kappaB, cytotoxicity, luminescence, MLSMR, MLPCN, NIH Roadmap, qHTS, NCGC
1. Compounds are first classified as having full titration curves, partial modulation, partial curve (weaker actives), single point activity (at highest concentration only), or inactive. See data field "Curve Description". For this assay, apparent inhibitors are ranked higher than compounds that showed apparent activation.
2. For all inactive compounds, PUBCHEM_ACTIVITY_SCORE is 0. For all active compounds, a score range was given for each curve class type given above. Active compounds have PUBCHEM_ACTIVITY_SCORE between 40 and 100. Inconclusive compounds have PUBCHEM_ACTIVITY_SCORE between 1 and 39. Fit_LogAC50 was used for determining relative score and was scaled to each curve class' score range.
* Activity Concentration. ** Test Concentration.
Data Table (Concise)