(T1530-06-36) An End-to-End One-Pot Immunoaffinity-LC-MS/MS Assay for Simultaneous Quantitation of Total and Conjugated Antibody Levels of Antibody-Drug Conjugates with Protease-Cleavable Linkers

Purpose: To develop an end-to-end one-pot assay for quantitation of both total (TAb) and conjugated antibody (ADC) levels of antibody-drug conjugates (ADCs) from in vivo samples using only one digestion enzyme, proteinase K, to generate both surrogate analytes. The proposed assay workflow would streamline the current established workflow that utilizes two digestion enzymes, trypsin and papain/proteinase K, to generate the respective TAb and ADC surrogate analytes. This could potentially halve sample and consumable usage, and data acquisition time.

Methods: Six hIgG1 protease-cleavable ADCs were incubated in neat buffer with a proteinase K stock solution prepared in HBSS under standard quantitative assay conditions and analyzed via a peptide mapping workflow on a HPLC coupled to a high-resolution quadrupole time-of-flight (QTOF) mass spectrometer system. One heavy chain constant region peptide was chosen for further development. The digestion variables of pH (6–8) and temperature (40–60 °C) were optimized by testing various combinations thereof and comparing the abundance of the released surrogate peptide. Rat serum standards and samples of a hIgG1 isotype ADC conjugated via a Gly-Lys dipeptide linker to a small molecule payload, spiked with stable isotope-labeled full length IgG1 internal standard, were purified on an AssayMAP Bravo (Agilent Technologies) robotic liquid handler using immobilized protein A resin cartridges. After washing with PBS, samples were eluted under acidic conditions into alkaline neutralization buffer. The neutralized eluate was incubated with either proteinase K or trypsin to liberate both the payload and the signature peptide. After quenching with acid, the samples were centrifuged, and supernatant was analyzed on a LC-QTRAP MS system via selected reaction monitoring.

Results: All six ADCs subjected to peptide mapping analysis after proteinase K digestion generated the heavy chain constant region peptide ALPAPIEKT (subsequently referred to as ALPA in this abstract) at very high abundances. In fact, ALPA was the most abundant peptide by peak area detected for five of the six ADCs tested; it was the second most abundant peptide detected in the sixth ADC digest. ALPAPIEK, the slightly shorter tryptic peptide contained within ALPA, is a commonly utilized and well characterized surrogate peptide for preclinical IgG bioanalysis, validating the advancement of ALPA for further development. For optimization of pH and temperature, one test ADC was incubated in all combinations of pH 6, 7, or 8 and 40, 50, or 60 °C. Figure 1 shows the results of optimization, wherein lower temperatures and higher pHs generally produced stronger ALPA signal. The overall increase in signal from the worst to the best conditions tested was 40%, representing a useful but marginal increase. Altogether, these results indicate a relatively robust method that can be tailored to other constraints such as payload stability without losing much sensitivity. Next, the optimized workflow was used to assay in vivo samples and was compared to the established trypsin workflow. The concentration-time profiles are plotted in Figure 2, showing both datasets are nearly superimposable. In terms of assay performance, the linearity, accuracy, and reproducibility of the proteinase K assay were slightly superior but comparable to the trypsin assay (R² = 0.9936 vs. 0.9891). The accuracy of the low and high QCs was similar for both methods, within 85–110%. Both assays easily quantitated the highest and lowest prepared calibration standards of 400 and 1 ug/mL, respectively, and were capable of quantitating even lower concentrations based on the low co-eluting signal in the rat serum matrix blank. The trypsin assay, which utilizes a longer peptide, was more selective, with virtually no peak in the matrix blank (Figure 3), theoretically enabling LLOQs in the 0.1 ug/mL range. The proteinase K assay yielded a small interfering peak in the matrix blank that would likely restrict the LLOQ to 0.3–0.5 ug/mL in rat serum.

Conclusion: Proteinase K has found widespread use in the bioanalysis of protease-cleavable ADCs for quantitation of conjugated Ab levels. Our studies demonstrate that it is also a highly effective reagent for TAb assays, enabling it to be used in a true end-to-end one-pot bioanalytical assay for ADCs and other bioconjugates containing an Fc region. The one-pot workflow was comparable to the established two-enzyme workflow with respect to sensitivity, accuracy, linearity, and reproducibility. We also demonstrated the robustness of the method with respect to pH and temperature, allowing the assay to be customized easily.

Optimization of digestion pH and temperature.

Rat serum concentration vs. time profiles. Each point is the mean of three biological replicates, with error bars representing SD.

Chromatograms of matrix blank (left) and lowest standard (right) for proteinase K (top row) and trypsin (bottom row).