Visualization and identification of benzylisoquinoline alkaloids in various nelumbo nucifera tissues

Benzylisoquinoline alkaloids in lotus (Nelumbo nucifera) seed plumules and leaves exhibit significant tissue specificity for their pharmacological effects and potential nutritional properties. Herein, 46 benzylisoquinoline alkaloids were identified via UPLC-QTOF-HRMS, of which 9 were annotated as glycosylated monobenzylisoquinoline alkaloids concentrated in the seed plumules. The spatial distribution of targeted benzylisoquinoline alkaloids in leaves, seed plumules, and milky sap was determined via MALDI–MSI. Furthermore, 37 Nelumbo cultivars were investigated using targeted metabolomics to provide insights into functional tea development. While aporphine alkaloids comprised the main compounds present in lotus leaves, bisbenzylisoquinoline alkaloids were the main compounds in lotus plumules, where glycosylation primarily occurs. These findings can help understand the distribution of benzylisoquinoline alkaloids in lotus tissue and the directional breeding of varieties enriched with specific chemical functional groups for nutritional and pharmacological applications.


Introduction
Sacred lotus (Nelumbo nucifera Gaertn.) is a basal, economic crop that is used for both medicinal and food application. Lotus seeds and underground stems are frequently consumed in East and Southeast Asian countries, owing to their high nutritional properties. Lotus leaf extracts possess antiviral, diuretic, and astringent activities and are also used as an antipyretic [1]. Lotus seeds reportedly have anticancer, antidepressant, antioxidant, and anti-inflammatory properties [2,3]. Furthermore, various secondary metabolites in lotus leaves and seeds exhibit tissue specificity and are used in different functional and nutritional [4]. Therefore, understanding the biosynthesis, transportation, and distribution of these metabolites can facilitate to their development and utilization as functional foods [5].
Reportedly, benzylisoquinoline alkaloids (BIAs) constitute the chief components in sacred lotus and possess specialized nutritional value; however, they probably play a key role in defense against herbivores and pathogens in the plant [6]. BIAs constitute a large class of plant specialized metabolites derived from tyrosine [7]. Studying the synthesis, transportation, and distribution of such metabolites can greatly facilitate the directed breeding of varieties enriched with specific functional groups. The genome of the China Antique variety of the sacred lotus has been sequenced and assembled [8]. Over the previous decades, there have been increasing efforts and research interest in studying the biosynthesis and bioavailability of these specialized alkaloids, such as nuciferine, N-nornuciferine, armepavine, and neferine [9]. BIA metabolism in lotus has been presumed based on the biosynthetic genes and enzymes involved in the generation of benzylisoquinoline in the plant [10]. Further BIA studies can provide new findings and contribute to the development and application of nutritional food resources.
Mass spectrometry imaging (MSI) has emerged as a powerful analytical technique that can assess the spatial distributions of numerous compounds in biological tissues [11,12]. This technique has been extensively used to determine the spatial distribution of metabolites involved in plant biosynthesis [13,14]. Ultrahigh-performance liquid chromatography-ion-trap high-resolution time-of-flight mass spectrometry (UPLC-QTOF-HRMS) is suitable for analyzing structurally complex natural products. Furthermore, the combined application of MSI and UPLC-QTOF-HRMS has been employed to analyze the synthesis, biosynthesis, and distribution of specialized metabolites in licorice and hypericum [15,16].
This study aimed to investigate the biosynthesis and distribution of different BIAs in various lotus tissues. This analysis comprised four steps: 1) a UPLC-QTOF-HRMS analysis of the chemical profiles of BIAs in the leaves, seed plumules, and milky sap of Nelumbo; 2) matrix-assisted laser desorption and ionization MSI (MALDI-MSI) was used to characterize the spatial distribution of BIAs in these three tissues; 3) quantitative UPLC-QTOF-MS analysis of BIAs in 37 Nelumbo cultivars; and 4) elucidation of the putative biosynthetic pathways of BIAs. In addition, this study aimed to optimize the application of medicinal plants and promote the development of food resources based on tissue-specific germplasm resources.

Plant materials and sample preparation
The seeds of 37 N. nucifera (sacred lotus) cultivars were collected from China (Yunnan, Jiangsu, and HuBei Provinces), Japan, and Thailand and identified using previously described morphological and histological methods [17]. Subsequently, these cultivars were grown in the United Lotus Germplasm Resource of the Amway Botanical Research Center (Wuxi, China) and collected in August (Supplementary Table 1). For each germplasm, three replicates of three different lotus seeds were used. Subsequently, 0.5 mL of fresh milky sap (obtained from Nelumbo stem) was extracted using 3 mL of extract solution (0.3 M hydrochloric acid:methanol, 1:1, v/v) via sonication for 20 min followed by centrifugation at 12,000 rpm for 3 min. Each sample extraction was repeated twice, and the supernatant was collected and filtered using a 0.22-μm Millipore filter (Alltech Scientific Corporation, Beijing, China) before UPLC-mass spectrometry (UPLC-MS/MS). Next, 0.5 g of fresh seed plumules and leaves was accurately weighed and then powdered in liquid nitrogen using an analytical mill (IKA A11 basic machine, Germany). The samples were extracted in a 6-mL mixed solution (0.3 M HCl: methanol, 1:1; v/v) for 20 min via ultrasonication followed by centrifugation at 5000×g for 10 min, and the supernatant was filtered using a 0.22-μm filter (Alltech Scientific Corporation, Beijing, China) before UPLC-MS/MS analysis [18,19]. Then, 60 μL of a methanol:water, (1:1, v/v) solution was added to the milky sap and centrifuged at 12, 000×g for 5 min, followed by the removal of 0.5 μL of supernatant. An equal volume of 2,5-dihydroxybenzoic acid solution was added to the mixed well, smeared onto a slide, dried, and analyzed using MSI.
The fresh leaf and seed plumules were mixed with 5% carboxymethyl cellulose solution and stored in Tissue-Tek® molds. Cryosections of 16-μm thickness were obtained at − 20 • C and mounted on glass slides for MSI measurements.

MALDI-MSI
All measurements were performed using AP-SMALDI 10 high-resolution MALDI imaging ion source (TransMIT GmbH, Giessen, Germany), which was operated at atmospheric pressure and coupled to a Q-Exactive Orbitrap mass spectrometer (Thermo Fisher Scientific, Bremen, Germany). The ion source is a solid laser (λ = 343 nm) operating at a repetition rate of 2000 Hz. For each mass spectrum, ions from 50 laser pulses were accumulated in the C-trap before being sent to the Orbitrap mass analyzer. All the experiments were performed in the positive-ion mode with the target voltage at +4.3 kV.

Liquid chromatography-mass spectrometry analysis
Agilent UPLC 1290II system combined with a 6540 QTOF (Agilent Technologies, Santa Clara, CA, USA) was used to determine the accurate mass of the metabolites. UPLC equipped with a binary solvent delivery system, autosampler, and column compartment was used in this study. Chromatographic separation was performed on a Waters BEH C 18 column (2.1 × 100 mm, 1.7 μm), and the elution conditions were as follows: 0-15 min, 5%-95% B. A and B indicate 0.1% formic acid water (formic acid:water, 0.1:100, v/v) and acetonitrile, respectively.
Sample ionization was acquired in both positive and negative modes within the mass/charge (m/z) range of 50-1000. The electrospray ionization (ESI) source operating parameters in both the positive and negative modes and the ESI-MS conditions were as follows: gas temperature, 325 • C; gas flow, 5 L/min; nebulizer, 35 psig; and sheath gas temperature, 350 • C. Internal references (purine and HP-0921) were adopted to modify the measured masses in real time, and the reference masses were m/z 121.0509 and 922.0098 in the positive-ion mode and 119.0363 and 1033.9881 in the negative-ion mode. The accurate mass of each metabolite was used for quantification.

Statistical analysis
Origin 2021b for Windows® (Origin Lab Corp., USA) was used to calculate the correlation coefficients of the alkaloids in the various tissues. Principal component analysis (PCA) was performed using SIMCA software version 14.1 (Umetrics, Sweden) to elucidate the differences between the four types of alkaloids detected in the three different tissue types.

Metabolic profiling of BIAs in lotus
Overall, 46 alkaloids were identified in the leaves, seed plumules, and milky sap of lotus. The accurate masses, fragmentation ions of MS/MS, and retention times of 19, 25, and 30 alkaloids were identified in lotus leaves, lotus plumules, and milky sap, respectively, were determined. (Table 1).
The accurate masses of the metabolites were determined by performing a full scan using QTOF-HRMS. The available fragmentation profiles of the compounds were obtained via product ion or a neutral loss scan. Compounds coclaurine 4, armepavine 8, N-nornuciferine 26, nuciferine 27, liensinine 31, isoliensinine 32, and neferine 33 were identified by matching the MS and MS/MS spectra to the authentic reference standards. The most common modification reaction of the metabolites was glycosylation, which increases the solubility and stability of compounds [20]. Glucoside (Glc)-substituted compounds were common in lotus. Compounds 11-19 were deduced to have undergone glycosylation with monobenzylisoquinoline alkaloids. Five alkaloid compounds (11)(12)(13)(14)(15)  O] + m/z 175.0497 were observed; therefore, these compounds were identified as isococlaurine and coclaurine, respectively, and this has been confirmed using the authentic standards. Furthermore, the elution times of the substituent C-7 alkaloids were longer than those of C-6 alkaloids on C 18 columns; thus confirming the identities of 3 and 4, which has been previously reported in lotus plumules [21]. In the MS/MS spectra of 2, the ion [M + H-CH 5 N] + m/z 255.1027 was detected, hence, 2 was determined to be N-methylhigenamine.

Aporphine alkaloids
In total, 10 aporphine alkaloids were identified in N. nucifera; among them, 21, 23, 24, 26, 27, and 29 were identified and quantified in lotus leaves, lotus milky sap, and seed plumules. For typical aporphines, the substitution of methoxy (OCH 3 ) or hydroxyl occurs mainly in the A ring, and major methyl substitution occurs in the N linkage [22]. If it is an adjacent methoxy and hydroxyl substitution, there are abundant fragment ions due to the loss of CH 3 OH and CO. Because of α cracking that usually occurs at the N groups, the loss of NH 3 and CH 3 NH 2 fragments could be present in the MS of aporphine alkaloids.
For typical aporphines, the Retro Diels-Alder reaction usually occurs in the B ring via the loss of a CH 2 --N-R group. Compound 29 was deduced to have a molecular formula of C 19 3 ] + m/z 223.077, and hence, it was tentatively identified as glaziovine, which has been previously reported in lotus leaf stalks [24].

Other alkaloids
Thirteen other alkaloids, including amine, pyridine, isoquinoline, and indole alkaloids, were identified in lotus leaves, milky sap, and seed plumules. Overall, 21 alkaloids were identified in the sacred lotus, including 9 glycosylated monobenzylisoquinoline alkaloids. Seven glycosylated monobenzylisoquinoline alkaloids were detected in lotus seed plumules, indicating it as the primary site of glycosylation.

In situ detection of BIAs in lotus
The distribution of alkaloids in lotus leaf, lotus milky sap, and lotus plumules was studied using MALDI-MSI. The nutritional applications of lotus were associated with the visualization results of the BIAs. The distribution of specific BIAs in different tissues as shown via MSI can benefit the commercial production alongside clinical applications of lotus. BIAs exhibited obvious specific tissue distributions, which supported the speculation that different BIAs were synthesized in the different tissues of lotus (Fig. 2). Combining these findings with those of fluorescence intensity in QTOF MS, we concluded that the typical bisbenzylisoquinoline compounds, such as liensinine/isoliensinine (31/32; [M+H] + m/z 611.3108) and neferine (33; [M+H] + m/z 625.3266) were concentrated and distributed in lotus seed plumules and milky sap ( Fig. 2A). Seven aporphine alkaloids were detected in the three tissues; however, their contents were the lowest in the seed plumules ( Fig. 2A). The results also suggested that O-nornuciferine/N-nornuciferine/lirinidine (24/26/25; [M+H] + m/z 282.1489) and dehydroanonaine (20; [M+H] + m/z 264.1062) could be generated in leaves, which has been previously reported in lotus [26]. Moreover, pronuciferine (29; [M+H] + m/z 312.1599) may be produced by laticifers and transferred to other tissues. It was apparent that very small amounts of bisbenzylisoquinoline were distributed in the center of lotus leaves and widely distributed in the milky sap, primarily around the seed plumules, suggesting that latex affected its transportation, similar to that in opium poppy. Nuciferine (27; [M+H] + m/z 296.1645) has been previously reported to be the most important alkaloid in all the growth and development stages of lotus leaves [27], and to be only widely distributed in lotus leaves. Similarly, high levels of roemerine (23) were detected in lotus leaves and milky sap. Hence, it can be concluded that the metabolism of aporphine and bisbenzylisoquinoline alkaloids are different in various lotus tissues, which has been previously reported [28]. However, some of the compounds were produced during the growth and development of the tissues. Laticifers and milky sap have an indispensable effect on the generation and transportation of these compounds for stress resistance [5].
The specificity distribution of methylation and glycosylation in the seed plumule is shown in Fig. 2B. Notably, glycosylated monobenzylisoquinoline alkaloids were detected not only around the seed plumules but also in the vascular bundle. O-Glc, and isococlaurine-Glc/Coclaurine-Glc (11/12/13/14/15), were detected in the same position, suggesting that these four compounds might be involved in monobenzylisoquinoline glycosylation. The distribution of bisbenzylisoquinoline and aporphine alkaloids differed between leaves and plumules. The areas of glycosylation with benzylisoquinolines were determined using MSI. Additionally, the discovery of trigonelline as a hormone associated with seeding supports the glycosylation process of alkaloids [29,30].

UPLC-QTOF-MS-based metabolomics analysis of BIAs in lotus
The accurate masses of 46 BIAs identified in lotus were quantitatively analyzed in lotus leaves, milky sap, and seed plumules. Lotus leaves, milky sap, and seed plumules demonstrated clear separation in the PCA plot by principal component (PC)2 (Fig. 3). There were obvious considerable differences among the different tissues (Fig. 3B), with lotus leaves near the top of the PC2 axis, characterized by high levels of aporphine alkaloids (partitioned as 19.1% of the variance). Lotus milky sap was characterized by lower levels of  Table 1, Monobenzyl isoquinoline (blue circle), aporphine (green box), bisbenzylisoquinline (yellow diamond), other (pink triangle) alkaloids. Numbers in (B) represent tissue-sample. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) monobenzylisoquinoline alkaloids and higher levels of bisbenzylisoquinoline alkaloids and other content located in the middle of the PC2 axis. Lotus seed plumules were located in the fourth quadrant and characterized by bisbenzylisoquinoline alkaloids and some monbenzylisoquinoline alkaloids. These alkaloids, including glycosylated and methylated alkaloids, were distinguished by tissuespecific variations (Fig. 3A), probably due to variations in genetic and epigenetic regulation. Furthermore, the plot revealed strong positive correlations among three bisbenzylisoquinoline alkaloids (liensinine [31], nelumboferine [30], neferine [33]), one aporphine alkaloid (nuciferine [27]), and one monobenzylisoquinoline alkaloid (isococlaurine [3]). Additionally, two glycosylated bisbenzylisoquinoline alkaloids (N-methylhigenamine-6-O-Glc [13] and N-methylisococlaurine-6-O-Glc [17]) were clustered away from the other monobenzylisoquinoline alkaloids and were significantly and positively associated with N-methylhigenamine (2) and armepavine (8). These results suggest that these compounds may be associated with seed development because of the epigenetic modification imparting stress resistant [31].
The BIA content in the three lotus tissues of the 37 lotus cultivars are shown in Fig. 4; nuciferine (27), N-nornuciferine (26), Onornuciferine (24), dehydroanonaine (20), and caaverine (22) were the main compounds in lotus leaves, accounting for >60% of the total BIAs in most of the analyzed cultivars. Aporphine and monobenzylisoquinoline chief were the main alkaloids in lotus leaves (Fig. 4B). Bisbenzylisoquinoline alkaloids (nelumboferine, neferine, liensinine, and isoliensinine) and monobenzylisoquinoline (armepavine, coclaurine, and isococlaurine) alkaloids were the predominant compounds in lotus seed plumules, accounting for >80% of the total BIAs in most of the analyzed cultivars. Liensinine and neferine were the two predominant compounds in lotus milky sap,   Table 1.

C. Hao et al.
accounting for nearly 90% or more of the total BIAs in some cultivars for "N1" (Fig. 4A). The contents of the lotus milky sap was more comprehensive determined compared with those of the other tissues and included monobenzylisoquinoline, aporphine, bisbenzylisoquinoline, and other alkaloids. As different alkaloids exhibit different pharmacological and physiological activities, quantitative studies can guide tissue selection in the analyzed cultivars to improve the bioavailability of these alkaloids.

Conclusion
Herein, 46 alkaloids, including 9 glycosylated monobenzylisoquinoline alkaloids, were identified using UPLC-QTOF-HRMS and shown to be localized in various tissues (leaves, milky sap, and seed plumules) of sacred lotus using MALDI-MSI. These results directly demonstrate the differences in distribution of these alkaloids in various lotus tissues and further evince the role of secondary metabolites in plant stress resistance. In addition, the glycosylated metabolites were differentially localized, being mostly distributed near the vascular bundles of the lotus seed plumules. These findings lay a foundation for the development and application of natural products for glycosylation. While more information is required to identify the sugar moiety of glycosylated monobenzylisoquinoline alkaloid, the discovery of the glycosylation of monobenzylisoquinoline alkaloids explains the enrichment mechanisms of the different types of naturally glycosylated alkaloids in the different tissues and provides useful insights into functional food and tea development. Furthermore, the content correlations of BIAs among the 37 representative lotus cultivars also provided insights into BIA biosynthesis. To the best of our knowledge, herein, the total alkaloid synthetic pathway in lotus was mapped and the known and possible metabolic pathways were labeled based on quantitative data and reported gene functions for the first time, thus providing new ideas for the further analysis of alkaloid synthetic mechanisms in lotus and BIA biosynthesis. Furthermore, the quantitative data serves as a scientific reference for planting selection, and the development and application of different varieties of lotus as well as provides guidance for lotus breeding.

Author contribution statement
Chenyang Hao, Wei Yang, Yuetong Yu, Yan Liu, Xiaolu Wei and Sha Chen: Conceived and designed the experiments; Analyzed and interpreted the data.
Gangqiang Dong, Yongping Zhu and Xiaolu Wei: Contributed reagents, materials, analysis tools or data. Chenyang Hao, Jun Zhang and Sha Chen: Performed the experiments; Wrote the Paper.

Data availability statement
Data included in article/supplementary material/referenced in article.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.