American Association for Cancer Research 2018

Here, we describe novel and enabling cell-based Ab internalization assays that are turnkey, medium throughput and geared toward industrial biologics discovery. Internalization measurements are made over time on 96-well microplates using live-cell analysis (IncuCyte® S3) and a fluorescent pH-sensitive dye coupled antibody-binding fragment (FabFluor) that binds the test mAb Fc region using a single step, no wash labeling protocol. An increase in fluorescence signal is observed as the mAb complex is internalized into the acidic lysosome. To validate this approach trastuzumab (Herceptin, Her-2) or rituximab (Rituxan, CD20) were mixed with the hFabFluor reagent (1: 3 molar ratio, 15min), serially diluted in complete media (1:2) and added to pre-plated BT-474 or Raji cells (no wash). Cell images (10-20x) were taken and automatically analysed for fluorescence area every 30min for up to 48h. Both trastuzumab (BT-474) and rituximab (Raji) caused clear time- and concentration-dependent internalization (EC50 values 2.1 and 2.6nM, respectively). The fluorescence signal was punctate, outside of the nucleus and strongly co-localized with a lysosomal marker (LysoSensor). In line with known marker expression profiles, specific internalization of mAbs to CD45, CD71 and CD3, but not CD20, was observed in Jurkat T-lymphocytes and CD45, CD71 and CD20, but not CD3 in Raji B cells. As a proof of concept for screening and direct comparison of test mAbs, 6 commercially available CD71 (transferrin receptor) Abs were labeled with mFabFluor reagent, serially diluted (1:2, 4.6-10000ng mL-1) and added to HT1080 osteosarcoma cells. 3 of the Abs produced a large internalization signal with detection <50ng mL-1, whilst the other 3 were internalized weakly with signal only visible at higher concentrations. A mean Z’ value of 0.82 was calculated from control wells indicating a microplate assay with high precision and robustness. 

Taken together these data support the validation of a simple, integrated and quantitative solution for directly studying internalization of mAbs into cells which can easily be scaled to compare multiple Abs in parallel.  This method enables mAb and ADC internalization measurements to be implemented at early stages of the biologics discovery process and will prove valuable in efficacy, safety and pharmacokinetic optimization. 

Matrigel (Corning) was dispensed across a range of volumes (20 – 50 mL) and concentrations (1 – 5 mg/mL^-1) into flat-bottomed 96-well TC micro-plates to form a solidified base layer. Tumor cells (A549, MCF-7, SKOV-3, MDA-MB-231) were seeded on top (1 – 2K cells per well), and in some experiments a full ECM sandwich was created by addition of a further volume of Matrigel (2 – 25%, 0.2 – 5 mg/mL^-1). Using a custom autofocusing method, phase contrast, brightfield and fluorescence images (10x) were captured every 6h for 7 days from within the cell incubator (IncuCyte S3 live-cell analysis system). Typically, 20 – 80 spheroids were analyzed in each well. All four cell types formed multiple cell aggregates within the first 3 days, ranging in diameter from 30 – 80 microns. A549, SKOV-3 and MCF-7 multi-spheroids grew as round aggregates while MDA-MB-231 spheroids displayed stellate branching characteristic of an invasive morphology. At Matrigel volumes less than 40 microlitres  or concentrations less than 3 mg/mL^-1, cells penetrated to the base of the plate and grew as ‘flat 2D’ cultures. Using a novel brightfield image analysis algorithm, the number, area and average size of the spheroids could be computed over time non-invasively and without the use of fluorescent labels. Once formed, A549, SKOV-3, MCF-7 and MDA-MB-231 multi-spheroids increased 3.0-, 1.6-, 3.8- and 3.3-fold in size over 4 days, respectively. Treatment of A549 multi-spheroids with the DNA enzyme topoisomerase inhibitor camptothecin (1000 nM) inhibited growth with comparable spheroid size at day 0 and day 7 post treatment (average brightfield area 1.4 x 104 micron2). Using fluorescent protein reporters for apoptosis (Annexin V) and cell viability (IncuCyte CytoTox Green) we could verify camptothecin-induced cell death (fluorescence values 149±16% of control (Annexin V) and 243±51% of control (CytoTox). A concomitant decrease of stably expressed RFP (to 3±1% of control) was observed. 

The combination of protocol developments, novel image acquisition/analysis algorithms and cell health reporters creates an integrated solution for measuring growth and vitality of multiple small spheroids in a relevant and 3D bio-matrix over time. This approach should be applicable to primary- and patient–derived organoid tumor samples as well as cancer cell lines.

Pickup, MW, Muow, JK, Weaver, MW (2014), EMBO Rep. 15(12): 1243–1253

Traditional approaches to screen for antibody Fc effector functions focus on ADCC using a homogenous live/dead readout, which greatly limits the contextual and correlative value of the screening data. To overcome this limitation, a multiplex screening assay profiling ADCC using multiple cell death endpoints, and quantitating secreted proteins/cytokines was developed and analyzed using Intellicyt’s iQue Screener PLUS and integrated ForeCyt® software. The iQue Screener Plus is a high throughput flow cytometry platform featuring 3 lasers with 13 fluorescent channels and can sample a 384-well plate in 20 minutes. 

As a proof of concept, a small set of therapeutic antibodies directed against the same tumor antigen were used to induce immune cell-mediated killing of tumor cells using different effector to target cell ratios and a range of antibody concentrations. To demonstrate antibody specificity, antigen positive target cells, negative control target cells and effector immune cells were barcoded with different encoding dyes and included in the same well.  At specific times, a small aliquot from each well was transferred to a new assay plate to run a multiplexed cell/bead mixture assay by the addition of a cocktail of fluorescent immunophenotyping antibodies, fluorescent dyes measuring unique apoptosis parameters, and a panel of QBeads for secreted protein detection. The samples were assayed on the IntelliCyt iQue Screener PLUS with simultaneous data analysis including multi-plate analysis for cell killing kinetics. Positive/negative target cells and immune cells were digitally segregated by differently encoded fluorescence. Time-dependent cell killing of antigen positive target cells were observed using the apoptosis markers mitochondria depolarization staining and cell membrane integrity, whereas little killing was measured in antigen negative cells. The antibody set showed a range of ADCC mediated killing and granule exocytosis/cytokine release suggesting differences in Fc effector functions. These results highlight the streamlined workflow on IntelliCyt iQue Screener PLUS platform to profile therapeutic antibodies in a functional ADCC/cytokine release assay predicting their clinical efficacy.

We developed a genetically encoded ATP sensor to enable direct, automated analysis of cellular ATP levels using IncuCyte® S3. In this study, our live cell imaging approach was utilized to evaluate the effect of cancer therapeutics on ATP levels in tumor cell lines, focusing primarily on compounds that target metabolic vulnerabilities. Cell lines stably expressing a genetically encoded, fluorescent ATP sensor or a control (non-ATP binding) sensor were generated. ATP levels were monitored and analyzed using an IncuCyte® S3 equipped with a specialized filter set and data acquisition module. Cellular ATP levels were measured over the course of hours to days following compound treatment. Reductions in ATP levels could be observed within an hour of administration, highlighting the sensitivity and temporal resolution of the IncuCyte® S3 ATP sensor assay. Transient reductions in ATP followed by recovery, which would have been missed by typical endpoint assays, were noted using our live cell analysis approach. Sustained decreases in ATP were associated with enhanced antiproliferative efficacy compared to conditions under which recovery of ATP levels were observed. For example, triple-negative breast cancer (TNBC) cell lines have been shown to be more dependent on activity of glutaminase 1 (GLS1), which catalyzes the first step in utilization of glutamine to fuel mitochondrial metabolism, than their receptor-positive counterparts. Using our live cell analysis approach, we observed a rapid drop in ATP upon glutamine deprivation or inhibition of GLS1 by CB-839 in the TNBC cell line MDA-MB-231. ATP levels remained below that of vehicle-treated cells for the duration of the three-day time course. In contrast, the estrogen receptor-positive cell line MCF-7 displayed a more modest decrease in ATP and full recovery within 48 hours. Quantification of phase confluence confirmed that CB-839 had a stronger antiproliferative effect in MDA-MB-231 cells compared to MCF-7 cells. Visualization of morphology in tandem with automated ATP sensor analysis provide additional insight into the heterogeneity of cellular responses to compound addition. Overall, these data highlight the ability of IncuCyte® S3 to provide direct, kinetic measurement of ATP by live-cell analysis.