The incorporation of electrospray ionization (ESI)-based mass spectrometry (MS) into the portfolio of high-throughput capable readout technologies has historically been hampered by the lack of suitable interfaces delivering the sample to the ESI ion source. Here, classical approaches face different bottlenecks such as limited sampling speed, analyte carry over, sample consumption, as well as the need for laborious sample preparation. The introduction of acoustic ejection mass spectrometry (AEMS) has provided an elegant solution by combining fast and contactless acoustic sampling with sensitive and accurate MS-based analyte detection. Our investigations on this novel technology resulted in two cutting-edge setups. First, a breadboard system combining a generic acoustic droplet ejection (ADE) system with an open port interface (OPI) was used to capture nanoliter droplets directly ejected from microtiter plate wells and to transport these to the ESI-source for MS analysis. This system was used to assess and optimize critical performance parameters related to analytical throughput and system stability. Maximum achievable sampling rates of up to 6 Hz are demonstrated for various ADME assays together with stress tests regarding ion suppression and system robustness. Complementary, the first commercial AEMS system (ECHO MS, Sciex) was integrated into Boehringer Ingelheim screening automation and modified to meet high-throughput screening (HTS) robustness requirements. This was achieved by combining custom software interfaces, in-house hardware modifications, and learnings from a validation project aiming at the identification of inhibitors of the cyclic GMP-AMP synthase (cGAS). We describe the method optimization to enable sensitive and accurate determination of enzyme activity and inhibition in a miniaturized 1536-well microtiter plate format. Furthermore, we present results from both a validation single-concentration screen using a test set of 5,500 compounds, and the subsequent concentration-response testing of selected hits in direct comparison with a previously established MALDI-TOF readout. In agreement with our observations on the breadboard system, passive and active cleaning options are demonstrated to effectively remain optimum system performance during HTS campaigns or to recover the ready state of the system after a blocked OPI from samples with high matrix loads. Finally, the results from the first HTS campaign using the fully automated AEMS readout at Boehringer Ingelheim are presented as the final validation of the method, demonstrating the system’s efficient and robust performance in a 1.2 million compound library screen.
