Abstract
Osteogenic differentiation is crucial in normal bone formation and pathological calcification, such as calcific aortic valve disease (CAVD). Understanding the proteomic and transcriptomic landscapes underlying this differentiation can unveil potential therapeutic targets for CAVD. In this study, we employed RNA sequencing transcriptomics and proteomics on a timsTOF Pro platform to explore the multiomics profiles of valve interstitial cells (VICs) and osteoblasts during osteogenic differentiation. For proteomics, we utilized 3 data acquisition/analysis techniques: data-dependent acquisition (DDA)-parallel accumulation serial fragmentation (PASEF) and data-independent acquisition (DIA)-PASEF with a classic library-based (DIA) and machine learning-based library-free search (DIA-ML). Using RNA sequencing data as a biological reference, we compared these 3 analytical techniques in the context of actual biological experiments. We use this comprehensive dataset to reveal distinct proteomic and transcriptomic profiles between VICs and osteoblasts, highlighting specific biological processes in their osteogenic differentiation pathways. The study identified potential therapeutic targets specific for VICs osteogenic differentiation in CAVD, including the MAOA and ERK1/2 pathway. From a technical perspective, we found that DIA-based methods demonstrate even higher superiority against DDA for more sophisticated human primary cell cultures than it was shown before on HeLa samples. While the classic library-based DIA approach has proved to be a gold standard for shotgun proteomics research, the DIA-ML offers significant advantages with a relatively minor compromise in data reliability, making it the method of choice for routine proteomics.