Background

Structural modifications of metabolites in plants play a pivotal role in various biological processes, including growth and development [1,2]. Different types of structural modifications endow metabolites with distinct chemical properties. For example, immature tomato fruits contain the toxic substance α-tomatine, which provides protection against predation. During fruit ripening, α-tomatine undergoes further acylation and glycosylation to produce the non-toxic substance esculeoside A, thereby enhancing the nutritional quality of the fruit [3]. These diverse types of modified metabolites constitute the metabolite modificomics, offering valuable insights into the diversity of plant metabolites.

There are many unknown types of structural modifications in plant metabolites, and current metabolomics techniques for detecting modified metabolites each have their own limitations and advantages. Nontargeted approaches allow for large-scale detection of metabolite signals but often lack the sensitivity required to detect ultra-trace levels of unknown modified metabolites [4]. Conversely, targeted methods are highly sensitive but can only detect known modifications, making it difficult to systematically study the full range of chemical modifications in plants [5]. The strategy presented in this study integrates the strengths of high sensitivity, broad coverage, and high resolution to enhance the detection sensitivity and identification efficiency for modified metabolites.

Here, we describe a widely targeted metabolite modificomics (WTMM) strategy [6], which employs the stepwise neutral loss-enhanced product ion (NL-IDA-EPI) method using ultra-high-performance liquid chromatography–quadrupole linear ion trap (UHPLC-Q-Trap) to capture metabolic signals containing various modified groups. High-resolution mass spectrometry (HRMS) data collected via ultra-high-performance liquid chromatography-Q-Exactive Orbitrap (UHPLC-QE-Orbitrap) are then used, alongside metabolite databases, for structural annotation of these signals, enabling the construction of a WTMM database. Furthermore, we utilize the scheduled multiple reaction monitoring (sMRM) mode of Q-Trap to perform quantitative analysis of the signals in the WTMM database, optimizing detection parameters to improve data quality. Using tomato as a model, we applied the WTMM strategy to establish a tomato-specific WTMM database, identifying 165 novel modified metabolites. Additionally, we demonstrated the utility of this database by validating diagnostic biomarkers for bacterial wilt in tomatoes [6].