(W1030-06-38) Use of Human Plasma as a Surrogate Matrix for Quantitation of Modified Protein Therapy across Multiple Tumor Tissues

Purpose: These data support a way to overcome the challenges encountered with identifying proper surrogate matrices for the quantitation of novel drug modalities in various tumor tissue types. Traditional small molecule analysis in tissue can utilize harsh solvents for homogenization and extraction of analytes from various tissue types. Targeted immunotherapies such as monoclonal antibodies, cannot be extracted from tissues using harsh solvents if structural integrity needs to be maintained for subsequent extraction or sample cleanup. Due to variability in the composition of tissues, it is challenging to select a representative surrogate tissue as a calibration or QC sample for protein quantification. Therefore, we proposed to investigate the use of plasma as a surrogate matrix for the quantitation of proteins in tissue samples. Our model system is a modified protein in tumor tissue and we seek to understand if plasma has broad application as a surrogate matrix across multiple tumor tissue types.

Methods: The therapeutic agent for this application is a modified protein, a part of which is a blocking domain of the active moiety of the molecule (Intact molecule) of which prior to binding to the target the blocking domain is cleaved off. A generalized scheme for specificity towards the active form is shown in Figure 1. Using immunoprecipitation with an anti-idiotypic (ID) capture antibody, an LC-MS/MS method was developed to quantify the cleaved moiety in several distinct tumor types. Tissue homogenization was optimized for the consistent recovery of cleaved modified protein over the assay range. Briefly, 15-20 mg of tumor tissue were cut using a scalpel and individually weighed. Tissues were added to 2 mL tough tubes containing Ruptor beads. The proprietary modified protein was spiked on top of the tissue. Chilled IP-Lysis buffer containing Halt protease inhibitor was spiked into the tough tube. Tissues were homogenized using an Omni Bead Ruptor at a speed of 3.90 m/s for 30 sec (2 cycles, 10 sec pause) at -5°C. Tubes were then centrifuged and the tissue lysate was removed. Streptavidin beads were added to a KingFisher 96 deep-well plate and incubated with the anti-ID. Samples were diluted with PBST containing 1% BSA prior to incubation with the capture antibody for 1 hour at 19^oC and 950 rpm using a Thermomixer. Streptavidin beads were washed using a KingFisher. 20µL of calibrators prepared in human EDTA plasma and 20µL tissue lysate were added to the wells and incubated for 2 hours at 19^oC and 950 rpm using a Thermomixer. The KingFisher was used to elute the therapeutic agent. 1M NH₄HCO₃ was then added to all wells along with a working internal standard solution and mixed by vortexing the plate. 10% w/v deoxycholate solution prepared in 50 mM NH₄HCO₃ was added to each well and incubated for ~30 min at 80^oC and 750 rpm using a Thermomixer. Samples were subjected to an enzymatic digest using Trypsin (0.125 mg/mL in 50 mM NH₄HCO₃) and incubated overnight at 37^oC and 750 rpm using a Thermomixer. After incubation, the plate was centrifuged and extracts were quenched using 10% v/v HCOOH in water. The plate was then vortex mixed and centrifuged. Extracts were transferred to a clean 96-well plate for injection. The established range for the assay is 10.0 – 5,000 ng/mL (100 – 50,000 ng/g). The unique peptides shown in Figure 2 were identified using Skyline and screened against the human proteome. Separation of the unique peptides was achieved using an Avantor ACE Excel 2 C18, 2.1 x 100 mm (2.0 µm) column at 50^oC and using 0.1% HCOOH in H₂O as mobile phase A and 96:4 ACN:DMSO (v/v) as mobile phase B operating at a gradient with post-elution wash-off with an initial flow rate of 0.6 mL/min. A triple quadrupole Sciex API6500 mass spectrometer equipped with a turbo ion spray source is used for detection in positive ion mode.

Results: Initial development was conducted in malignant melanoma (MM) and renal cell carcinoma (RCC) tumors. The accuracy and precision of the tissue lysates are shown in Table 1. The accuracy and precision for both RCC and MM tumor tissues were within ±22% relative to the plasma standard curve.

Conclusion: Preliminary data suggest that plasma can be used as a surrogate matrix for preparing a calibration curve for at least two different tumor tissues, RCC and MM. This significantly reduces the costs of tissue assays and reduces the complexity of development if animal tissue matrices were used as surrogate matrices. Evaluation of additional tumor types (breast, lung, Merkel cell carcinoma, and pancreas) using plasma as a surrogate matrix is currently ongoing. A formal analysis of the statistical variability across these tissues will be presented during the poster session.