Metabolite Profiling of Growth Regulatory Hormones from Maize Tissue

Standards Limitation—Chemical Synthesis Gap

LC/MS experiments detect “mass features,” which are the result of ionized molecules hitting a detector within a time window of the chromatography. As such, there may be many mass features with the same chemical formula as a compound of interest, and a feature with the correct mass of a hormone is insufficient evidence that this feature is due to the observation of that hormone. Comparison of both the MS spectra and chromatographic retention time of a feature to a known standard is typically required for compound identification. The difficulty and expense of producing chemically known standards has limited the number of laboratories and teams that can study plant regulatory molecules and hormones. A “chemical synthesis gap,” where proposed chemicals have not yet been or cannot readily be synthesized, can occur when there is purification and detection of a molecule that does not have a synthesized standard to which it can be compared for chemical identification during MS analyses. For some hormone classes, particularly the BRs and GAs, multiple compounds have the same mass, because they consist of the same chemical formula arranged in a different structure. Thus, differences in retention time caused by differences in structure (e.g., different orientation of an –OH group resulting in a change in interaction with the column and mobile phase) is key to determining molecule identity (see Fig. 1) and, unlike m/z, which can be arrived at theoretically, retention times on a column must be measured using bona fide standards. In addition to the traditional use of known chemical standards, metabolites purified from mutant plants with disruptions in hormone biosynthetic pathways can be used to identify hormones. Tracking the accumulated and missing metabolites from mutant and wild-type tissues can further help in the identification of peaks at novel retention times.

View larger version:

• In this window
• In a new window

Figure 1.

Chemical structure of pairs of brassinosteroid and gibberellin metabolites with the same mass and chemical formula but different chemical structures. Chemical structure of 22α-hydroxy-campesterol (A), 6-oxocampestanol (B), gibberellic acid₁₅ (C), and gibberellic acid₅₃ (D).

The addition of deuterium-labeled standards can further improve LC/MS analysis and allows for absolute quantification of the hormones. Deuterated compounds are added to ground tissue at known concentrations and are nearly chemically identical to nondeuterated hormones during extraction and separation. They can be distinguished by mass during analysis and used to create a standard curve. In the absence of deuterium-labeled standards for every individual metabolite in a hormone class (e.g., in the presence of a chemical synthesis gap), indirect quantification of hormones can be carried out using a standard curve from a chemically similar compound. For example, our associated protocol (Dilkes and Best 2025b) uses a deuterated GA₅₃ (GAs are numbered in order of discovery) standard to produce a standard curve for the entire GA class of metabolites. Using a representative metabolite as a standard is an effective approach, especially when the cost and availability of deuterated bona fide standards are prohibitive. This approach allows quantification of one chemical hormone to act as the proxy for others in the same class, and thus the accumulation of a whole class of molecules can be assessed. A potential pitfall in the approach that can confound interpretations is that even though compounds are in the same hormone class and are structurally similar, they may have altered ionization or differential extraction from the compound used for the standard curve.

Purification and Signal Improvement

The total amount of hormone that can be loaded onto a column and sprayed into a mass spectrometer limits the amount of signal that can be observed. Prior methods required purification of each hormone away from other metabolites before detection. Improvements in the sensitivity and mass accuracy of instruments to detect these hormones now allow for simultaneous observance of multiple hormones in the same sample. However, this less targeted approach can result in unwanted metabolites that can interfere with the detection and quantification of the wanted hormones remaining in the analyzed sample. The addition of a solid phase extraction step improves the signal to noise ratio by removing some unwanted metabolites from the samples. Our method uses graphitized carbon black to achieve this purification, but other sorbents could be used, such as activated carbons. Carbon black can react with hormones and, if left in contact with the sample for too long, will reduce signal levels.

Derivatization is another form of signal improvement. We propose that the combination of derivatization, targeted (tandem MS/MS) and untargeted (quadrupole time of flight [QTOF] MS) profiling, and metabolite measurements in mutants can ultimately identify all hormone pathway intermediates. The QTOF MS allows detection of all metabolites in a sample without previously knowing their exact masses prior to analysis. In targeted MS/MS analyses, extraction and separation of specific metabolites are carried out and when metabolites of the expected mass are detected, these metabolites are further fragmented, and their fragments are then detected to provide further structural detail. All forms of extraction and derivatization are designed to bias the signal for a specific hormone class. However, extraction and derivatization can also cause some distortion of that signal, which can be problematic when looking at multiple metabolites within the same class. For example, the method we present in the associated BR protocol (Dilkes and Best 2025a) uses derivatization at the 3′ hydroxyl of sterols with picolinic acid. Some BRs have a 3′ ketone group at this location and therefore will not derivatize and will not be detected in this protocol. There are alternate methods to derivatize these 3′ keto sterols, such as the Amplifex Keto Reagent, which derivatizes 3′ keto steroids with a permanently changed quaternary amine that stabilizes the structure and improves ionization in positive mode (Zhu et al. 2015). This derivatization allows the efficient ionization of the few BR intermediates with a 3′ ketone, including the pregnane derivatives formed by side-chain cleavage of the steroid by enzymes in the plant (Yang et al. 2018; Carroll et al. 2023; Kunert et al. 2023). If desired, a second round of analysis incorporating this method will allow for a global analysis inclusive of the 3′ keto steroids (Dilkes and Best 2025b, in this collection).

These methods are not exclusively useful for previously identified compounds. For example, performing an untargeted LC/MS profiling of the picolinate derivatives of all steroids, and not just the known BRs, can allow a search of the data for the picolinate fragment as an informatic anchor, and the potential identification of novel sterols. This technique was previously successful on tissue from the Setaria BR-deficient cytochrome P450 724 (CYP724) mutant bristleless1, which accumulated a C21 pregnane (Yang et al. 2018). In underivatized samples, we also detected a mass consistent with the 3′ ketone derivative of this same compound and hypothesized the existence of a plant desmolase (Yang et al. 2018). Recently, the pathway to produce these compounds was elucidated (Carroll et al. 2023; Kunert et al. 2023), confirming the hypothesized pathway uncovered by untargeted metabolite profiling of a hormone biosynthetic mutant.

Many more such opportunities for characterizing and identifying hormones exist across the plant kingdom. For example, the ultimate bioactive products of the BR biosynthetic pathway in maize and other monocots are still unclear. The diversification of the cytochrome P450 90s (CYP90) genes that encode the BR biosynthetic pathway enzymes that we are familiar with in flowering plants occurred after the split between vascular and nonvascular plants. As a result, the CYP90s in plants other than the angiosperms may synthesize different polyhydroxylated sterols that act as BRs in these plants. GAs may also be different across plant taxa. For example, the sex-determining antheridogens in some homosporous ferns are GAs. To date, all chemically identified antheridogens are GAs, but many antheridogens are not yet chemically identified and may be novel GAs or other signaling molecules (Takeno 1991; Tanaka et al. 2014). Additional opportunities for chemical identification include the metabolic targets of the CYP78A enzymes, the KLUH pathway, members of which are also as yet unidentified (Anastasiou et al. 2007; Stransfeld et al. 2010). Studying mutants in these pathways in maize, where greater volumes of tissue are readily obtained for metabolite purification than in other plants, may aid in the chemical identification of this and other proposed hormones.