May 15, 2015 (Vol. 35, No. 10)
Mei Cong, Ph.D. director Promega
Zhi-Jie (Jey) Cheng Ph.D. Promega
Natasha Karassina Promega
Jamison Grailer Promega
Jim Hartnett senior research scientist Promega
Neal Cosby Promega
Frank Fan Ph.D.
Bioluminescent Reporter-Based PD-1/PD-L1 Blockade Bioassay
Immunotherapy enables a patient’s own immune system to battle cancerous target cells instead of either attacking the target cells directly or using whole-body approaches. Programmed death receptor-1 (PD-1), expressed on cell types including T and B cells, serves as a negative costimulatory receptor. Its ligand PD-L1 (B7-H1) is not expressed by normal epithelial tissues but is aberrantly expressed on a wide array of human cancers. Engagement of PD-1 by PD-L1 expressed on tumor cells disables the host antitumor response, and the effect is associated with poorer prognosis of the patient. PD-1 and PD-L1 are important targets for cancer immunotherapy. Investigational anti-PD-1 immunotherapy drugs, such as nivolumab from BMS and lambrolizumab (MK-3475) from Merck, demonstrate significant overall survival rates in patients with advanced melanoma.
Currently methods to measure the potency of drugs targeting PD-1 or PD-L1 include in vitro binding assays, primary T cell-based cytokine release assays, and in vivo model systems. Binding assays do not demonstrate the drug’s functional impact; primary cell-based assays are tedious and have high variation; and although in vivo animal models can assess the drug’s antitumor effects, survival time and rate, they also have high variation and are cost and time prohibitive. There is urgent need for a simple, easy-to-use, robust, cell-based functional bioassay to measure potency of a candidate drug during the development of therapeutic mAbs that block the interaction of PD-1:PD-L1 for drug discovery and development. Such an assay requires high sensitivity with appropriate specificity, precision, and accuracy for drug screening and characterization in early drug discovery, and for lot release and stability studies in drug manufacture.
Improved Bioluminescent Reporter-Based PD-1/PD-L1 Blockade Bioassay
We developed a bioluminescent reporter assay that can be used for rapidly quantifying function of PD-1 or PD-L1 therapeutic antibodies in development as measured by activation of NFAT signaling pathway. We engineered a Jurkat T-cell line stably expressing NFAT-luciferase reporter and human PD-1 (PD-1 effector cells). Cells stably expressing human PD-L1 and a TCR activator act as antigen presenting cells (PD-L1+ cells). Three steps of signaling pathway alterations representing the antibody’s mechanism of action are achieved in this single-step, homogeneous cell-based assay (Figure 1). Co-cultivating the two cell lines induces activation of the Jurkat NFAT pathway via crosslinking of T-cell receptor (TCR) activator/TCR complex. PD-1 signaling in PD-1 effector cells following engagement of PD-L1+ cells inhibits T-cell function, and results in NFAT pathway inhibition. Blockade of PD-1:PD-L1 interaction with either anti-PD-L1 or anti PD-1 mAbs can reactivate the NFAT pathway in a dose-dependent manner.
Assay Specificity and Suitability for Monitoring Antibody Stability
We have developed both cells in the assay into a thaw-and-use format that removes the need for standard cell culture, and the cell batches are functionally tested. The PD-1/PD-L1 bioassay incorporating thaw-and-use cells is convenient, demonstrates consistent results (data not shown), and exhibits assay specificity (Figure 2A). No premixture of antibody drug and cells is required for efficient blockade. Our data demonstrate that the PD-1/PD-L1 blockade bioassay can be used to measure the potency of either anti-PD-1 or anti-PD-L1, and the assay is specific as demonstrated by a non-related antibody, anti-CTLA-4 drug Ipilimumab, which failed to activate the NFAT pathway.
Many human monoclonal antibodies display poor biophysical properties including sensitivity to pH and UV light, low stability, and a propensity to aggregate. These unfavorable properties are even more pronounced for human antibody fragments, which often require a considerable degree of optimization. Here we demonstrate that the bioluminescent reporter-based PD-1 Blockade Bioassay is suitable for detecting potency changes of stressed antibody samples (Figure 2B and C). Two PD-1/PD-L1 blocking test antibodies were heat stressed at either 42°C or 65°C for one or two days before being applied to cell-based PD-1/PD-L1 Blockade Bioassay with thaw-and-use cells. Compared to antibody stored at 4°C as positive control, both test antibodies retained functional activity after two days of heat stress at 42°C. Test antibody #1 completely lost its activity after one day of heat-stress treatment at 65°C, whereas potency of test antibody #2 was right shifted, indicating that test antibody #2 is more resistant to heat treatment at 65°C.
Figure 2A. Assay specificity of PD-1/PD-L1 Blockade Bioassay: Serial dilutions of anti-human CD274 (B7-H1, PD-L1) antibody (BioLegend, Cat.# 329709), anti-PD-1 antibody (BioLegend, Cat.# 329911), or Ipilimumab (Yervoy, BMS) were added into co-culture system of PD-1 effector cells and PD-L1+ cells, incubated for 6 hours at 37°C before luciferase activity was quantified using Bio-Glo™ Reagent. Data was fitted using 4PLC curve fit of GraphPad Prism® software. Figure 2B and C. PD-1/PD-L1 Blockade Bioassay is suitable for antibody stability studies. Anti-PD-1/PD-L1 blocking antibody #1 and #2 were heat stressed at either 42°C or 65°C for indicated times before being applied to cell-based PD-1/PD-L1 Blockade Bioassay. Loss of reporter response to heat stressed PD-1/PD-L1 blocking antibody after heat-treatment was detected.
Bioassay Qualification and Tolerance to Human Sera
Bioassay qualification was performed to demonstrate that the PD-1/PD-L1 Blockade Bioassay has great intra-plate accuracy, is able to measure relative potency for antibody biologics with good linearity, accuracy, precision, and reproducibility, from qualification assays run at four different test antibody potencies (range 50–200%) relative to 100% reference sample across three different assay days (data not shown). The assay was also challenged with various amounts of human sera from 1% to 10% in assay buffer. Resulting EC50s and fold inductions of test antibody were minimally impacted by human sera, suggesting that the bioassay is a good candidate for further development into a NAb assay.