(1096-A) Streamlined and semi-automated MS-based proteomics pipeline from protein extraction to real-time data analysis by coupling BeatBox, PreON and PaSER platforms

Abstract: Introduction
Sample preparation is an often overlooked yet crucial component of successful LC-MS-based proteomics studies. Conventional manual workflows often have various limitations, such as low throughput, insufficient reproducibility, and time-consuming protocols. Here, we present a streamlined workflow from sample homogenization to data analysis by combining highly efficient protein extraction on the BeatBox platform with the robust iST sample preparation, fully automated on the PreON platform (PreOmics). Coupling this workflow with the newly developed PaSER technology (Bruker) provides improved and accelerated data processing by making the results available immediately after the run and eliminating the need for separate processing on high-performance computers.
Methods
The described sample preparation workflow was evaluated with human cell lines and various mammalian tissue types. Protein extraction was performed on the BeatBox using sample-specific settings, followed by automated digestion and peptide clean-up on the PreON applying the iST protocol. This combined workflow was compared to traditional sonication and manual iST sample preparation. Data were acquired by a timsTOF mass spectrometer (Bruker) coupled to an Easy nLC 1200 (Thermo) and processed in real-time using PaSER (Bruker), effectively generating a streamlined workflow from sample preparation to real-time results. To validate the real-time processing, the data were also calculated in MaxQuant.
Preliminary data
From intact sample to finished data analysis (DDA on timsTOF Pro with PaSER data processing, Bruker) in less than one working day and with less than 30 minutes of hands-on time, we processed and analyzed 12 samples in parallel using the described proteomics pipeline. We identified ~4900 protein groups from HeLa cells and ~3600 protein groups from mouse liver tissue using a short 30-minute LC gradient. Compared to sonication, BeatBox homogenization provides not only an excellent extraction rate but also higher flexibility of input material, starting amount, and the number of samples. At the same time, fully automated sample preparation on the PreON platform minimizes technical variability and sample loss and dramatically reduces hands-on time compared to manual processing. The CVs of the technical replicates prepared on the PreON were 1~2% lower for intra-day and 2~4% lower for inter-day repeatability than manual processing. PaSER processes the data in real-time during the MS acquisition for accelerated data analysis. Comparing the PaSER output with a traditional MaxQuant offline search revealed similar protein IDs but significantly higher peptide IDs, indicating higher sequence coverages.
Novel aspect
The presented workflow sets a new standard in speed and robustness in LC-MS-based proteomic sample preparation and data processing.