The sesame varieties (Supplementary Figure S1) used in this study were provided by the National Sesame Medium-term Genebank (Wuhan, China). They were cultivated from June to October 2020 under identical growth conditions at the experimental station of the Oil Crops Research Institute of the Chinese Academy of Agricultural Sciences (OCRI-CAAS) located in Wuhan, China (N 30.57°, E 114.30°, 27 m altitude). Seed samples were prepared after harvest by mixing seeds from 15 individual plants. Samples used for the melanin extraction were kept at the seeds room of the Institute, and those for the widely targeted metabolomic profiling analysis were stored at −80°C.

The melanin was extracted from the black and brown seeds following previously reported methods with some little modifications (23, 24, 27). Briefly, 1 g of ground seeds was solubilized in 10 ml of 0.5 mol/L NaOH for 4 h, which generated a dark brown solution. After centrifugation at 5,000 rpm for 5 min, the supernatant was collected and filtered. The extraction process was repeated twice, and the supernatants were combined. The pooled supernatant was centrifuged again before the seed melanin was precipitated from the aqueous solution by acidification with 6 mol/L HCl at pH 2. The mixture was incubated for at least 12 h at room temperature. The crude melanin was obtained after centrifugation at 10,000 rpm for 10 min. The precipitate was redissolved in 0.5 mol/L NaOH, and 6 mol/L HCl was added again to ensure a higher purity of the melanin. Then, the precipitate of the seed melanin was washed thoroughly with distilled water until the supernatant pH became neutral. Thereafter the precipitate was sub sequentially washed with chloroform and ethanol to discard the residues of oil and other seed components. Finally, the crude seed melanin was dried overnight at 55°C, weighted, and kept for further analysis.

The methods used to characterize the black and brown sesame seed's melanin included the morphology and size distribution, solubility, precipitation, UV–Vis absorption spectroscopy, Fourier transform infrared spectroscopy (FT-IR), and thermogravimetric-differential scanning calorimetry (TG-DSC).

The morphology and size distribution of the black and brown sesame seeds melanin pigments were examined using scanning electron microscopy (SEM, S-4800, Hitachi Co., Ltd., Matsuda, Japan) as per Wang et al. (26). Prior to the observation, the melanin powders were coated with a platinum (Pt) layer to increase the conductivity of the samples.

The seed melanin solubility, precipitation, and absorbance were analyzed following the method reported by Qi et al. (24) with little modification. The dried melanins were suspended in aqueous (distilled water and 0.5 M NaOH) and organic solvents (methanol, ethanol, chloroform, isopropyl alcohol, and hexane). The mixture of the melanin and solvent (10 mg/ml) in 10 mL tubes was kept in a water bath at 25°C for 4 h. The solubility was recorded after the mixture was centrifuged at 5,000 rpm for 5 min. For the precipitation, 125 μL of 10% FeCl₃ and 80 μL of 6 mol/L HCl were added to 500 μL of the melanin solutions. The mixtures were kept at room temperature for 30 min followed by centrifugation at 10,000 rpm for 5 min. For the UV-Vis absorption spectroscopy, 10 mg of the seed and synthetic melanin was dissolved in 10 mL of 0.5 M NaOH. The UV-Vis absorption spectra of the solution were recorded on a UV- 2,450 spectrophotometer (Shimadzu, Japan) in the 220–900 nm range. The synthetic melanin (MelSYNTH, M8631) was purchased from Sigma-Aldrich (St. Louis, MO, USA). It is an L-Dopa melanin obtained from L-tyrosine in the presence of H₂O₂.

The FT-IR spectra of the sample were acquired using a Nicolet Is10 FT-IR spectrometer (Thermo Scientific, USA) equipped with a diamond ATR accessory as per Qi et al. (24). The measurements were conducted in the 4,000–600 cm⁻¹ range with a 4 cm⁻¹ optical resolution.

The TG-DSC curves were registered on NETZSCH STA 449 C (Netzsch-Gerätebau GmbH, Selb, Germany) thermal analysis instruments (29, 30). Each color seed melanin sample (10 mg) was heated from room temperature to 1,000°C at a heating rate of 10°C/min under a nitrogen flow.

Freeze-dried white and black sesame seed samples (Supplementary Table S1 and Supplementary Figure S1) were grounded using a mixer mill (MM 400, Retsch, Haan, Germany) with a zirconia bead for 1.5 min at 30 Hz. Further, 100 mg of lyophilized powder of each sample was extracted overnight at 4°C in 1.2 ml of 0.5 M NaOH and 70 % methanol (Supplementary Table S1). Next, they were centrifugated at 12,000 rpm for 10 min, and the supernatants were collected separately and filtrated through a 0.22 μm micropore membrane (SCAA-104, ANPEL, Shanghai, China). The extracts were conserved at −20°C up to the ultra-high-performance liquid chromatography coupled with electrospray ionization tandem triple quadrupole mass spectrometry (UHPLC-ESI-QqQLIT-MS/MS) analysis (9, 31). The QC (quality control) samples were prepared by mixing the different seed extracts. The metabolomics analysis and the qualitative and quantitative analyses of metabolites were performed similarly as we reported recently (9, 10).

The metabolomics data quality was prior assessed, and substances with large deviations (CV value > 0.5) were eliminated. Unsupervised PCA (principal component analysis) and HCA (hierarchical cluster analysis) were performed in R (version 3.5.0) using the statistics function prcomp and the R package pheatmap (www.r-project.org). Normalized data by log2-transformation was used for Orthogonal Partial Least Squares-Discriminant Analysis (OPLS-DA) in R. The p and fold change (FC) values were set to 0.05 and 2.0, respectively. The differential metabolites were filtered out based on the FC and VIP ≥ 1. Venn diagrams were constructed to illustrate differences in metabolite numbers. Differential metabolites were annotated using the KEGG Compound database (http://www.kegg.jp/kegg/compound/) and further mapped to the KEGG Pathway database (http://www.kegg.jp/kegg/pathway.html). Pathways with significantly regulated metabolites were considered metabolite sets enrichment analysis (MSEA). The significance of the enrichment was determined by the hypergeometric test's p-values. Besides, GraphPad Prism v9.0.0121 (GraphPad Software Inc., La Jolla, CA, USA) was used for some graphing.

To determine whether the dark sesame seed pigment is melanin, we followed the previously described melanin isolation method and extracted melanin from the black and brown sesame seeds. Each colored seed was represented by three varieties (Supplementary Figure S1), and the experiment was conducted in triplicate. The morphology of melanins is an important parameter to compare melanin samples from various sources. The SEM images showed that the black and brown sesame melanin samples exhibit amorphous form without self-organization (Figures 1A,B). The melanin content in the black and brown sesame seeds ranged from 95.8 to 105.5 mg/g and 51.3 to 76.5 mg/g, respectively (Figure 2A).

We tested the extracted pigments' solubility, precipitation, and absorbance to confirm their melanin nature. The black sesame pigment showed a perfect solubility in an alkaline solution (0.5 M NaOH) and could not be dissolved in water, isopropyl alcohol, chloroform, and hexane (Figure 2B). However, it was slightly soluble in methanol (80%) and ethanol. The extracted pigments exhibited precipitation in 10% FeCl3 and 6 mol/L HCl solutions (Figure 2C). The extracted pigments and synthetic melanin were dissolved in 0.5 M NaOH (1 mg/ml) to obtain their absorption in different spectra from 220 to 900 nm. The absorbance spectrum of the black seed extract included a linear increase of the absorbance from 900 to 400 nm, an exponential increase from 400 to 300 nm, then a decrease from 300 to 220 nm (Figure 2D). The brown seed melanins' absorbance spectra comprised a linear increase of absorbance from 900 to 400 nm, followed by an exponential increase of absorbance from 400 to 220 nm. In contrast to the black seed extract, the synthetic melanin absorbance spectrum showed an exponential increase from 900 to 600 nm and was almost stable between 600 and 300 nm (Figure 2D).

We performed the Fourier transform infrared spectroscopy (FT-IR) test to confirm that the isolated pigment from black and brown sesame seeds is melanin. The FT-IR spectra of the isolated black and brown sesame seed's pigment were similar and are shown in Figures 3A,B. The peak at 3444.6 and 3462.97 cm⁻¹ for the black and brown seeds melanin, respectively, corresponds to the OH group, and the peaks at ~2,855 cm⁻¹ can be attributed to the stretching vibration of the aliphatic CH groups (14, 24, 26). The characteristic strong bands at ~1,711–1,745 cm⁻¹ can be assigned to C = O stretching (26). The absorption at 1,650–1,500 cm⁻¹ corresponds to the vibration of aromatic ring C = C bonds or stretching of the COOH and C-N groups, or the N-H deformation (24, 26, 32). The peaks at ~1,465 cm⁻¹ are attributed to aliphatic C-H groups in the melanin pigment, and those at ~1,090 cm⁻¹ are assigned to the COH stretching of phenolic compounds or aromatic rings in substituted macromolecular systems (24, 26, 33). The peak centered at 1,049 cm⁻¹ for the pigment extracted from the black sesame seeds indicates CH in-plane of aliphatic structure, which is also a characteristic of melanin pigment (24). The observed peaks at ~722 cm⁻¹ can be attributed to out-of-plane bending of the aromatic C-H group (24).

The TG-DSC analysis helps investigate processes that occur in a material under different temperature conditions. To check the thermal stability of the black and brown sesame seed's melanin, we carried out the TG-DSC analysis in an open system under nitrogen flow in the temperature ranges from room temperature to 1,000°C at 10°C/min heating rate. The TG and DSC curves of the black and brown seed melanin are presented in Figures 4A,B, respectively. The TG and DSC curves of the two isolated melanins showed similar trends with three sharps weight loss, one endothermal peak at ~110°C, and one broad exothermal peak. The combustion (weight loss) of the two melanins started at ~50°C. The weight decay was slight between ~50–210°C; logarithmic between ~210–410°C; and slight from ~410°C. The total weight loss of the black and brown seed's melanin was 83.12 and 85.09%, respectively, indicating the sesame melanin is heat-stable. The broad exothermal peak ranged from ~450–1,000°C.

Based on the results of our recent study (10) and the fact that melanin components are primarily soluble in alkaline solvents, we performed a comparative metabolomics profile analysis of methanol and NaOH extracts of black and white sesame seeds. The aim was to investigate the effect of extraction solvents on sesame seed's metabolome profile and reveal the potential precursors of the sesame melanin. The UPLC-MS/MS-based widely targeted metabolomics analysis approach (9, 31) was applied, and three varieties of black and white sesame seeds were tested (Supplementary Table S1 and Supplementary Figure S1). In general, the total number of identified metabolites varied from 351 to 403 and 462 to 468 in NaOH and 70 % methanol extracts, respectively (Supplementary Table S1). Different classes of primary and secondary metabolites were identified, among which organic and phenolic acids, lipids, amino acids and derivatives, nucleotides and derivatives, and saccharides and alcohols were predominant (Supplementary Table S2).

To compare the seed's metabolites patterns in NaOH and methanol extracts, we performed the hierarchical cluster analysis (HCA) (Figure 5A). The samples extracted with 70% methanol clustered separately from those extracted with the NaOH solvent, indicating that the metabolite profiles of sesame seed in organic and aqueous solvents are different. Some metabolites exhibited high relative content in NaOH solvent. Principal component analysis (PCA) was further conducted to better understand the general metabolite variability between groups. As per the HCA result, samples extracted with NaOH and 70 % methanol were obviously separated, confirming that the metabolome profile of sesame seeds in methanol and NaOH solvents differed (Figure 5B). The HCA and PCA results also showed that metabolites accumulation in sesame seed is influenced by the genotypes.

The differential metabolite screening was performed in the pairwise comparison between the two solvent extracts of the sesame varieties Zhongzhi28 (WS01, white seed) and Zhongzhi33 (BS01, black seed) using the criteria of fold-change (FC ≥ 2 or ≤ 0.5) and VIP > 1. The results indicated that there were 337 (94 up-regulated and 243 down-regulated) and 371 (105 up-regulated and 266 down-regulated) significantly differential metabolites between WS01-Met and WS01, and between BS01-Met and BS01, respectively. In order to identify key metabolites that were more soluble in the NaOH solvent, we constructed a Venn diagram among the up-regulated differential metabolites between WS01-Met and WS01 and between BS01-Met and BS01 (Figure 6A). There were seventy-two (72) overlapping significantly differential metabolites among the two pairwise comparisons. The overlapping metabolites are listed in Supplementary Table S3. Phenolic acids, amino acids and derivatives, and organic acids accounted for 27.78, 19.44, and 20.83%, respectively, of the key sesame seed metabolites that exhibited high relative content in the NaOH solvent extract (Figure 6B).

As melanin is exclusively soluble in alkaline solutions (24), we firstly identified the differential metabolites between the NaOH solvent extract of Zhongzhi28 (WS01, white seed) and Zhongzhi33 (BS01, black seed). Differential metabolite analysis revealed that there were 204 significant differential metabolites, including 101 up-regulated and 103 down-regulated between the two sesame (Supplementary Table S4). The KEGG pathway enrichment analysis of the differential metabolites indicated that they were mostly involved in the biosynthesis of secondary metabolites; tyrosine, tryptophan, pyruvate, and phenylalanine metabolisms; phenylpropanoid biosynthesis; and isoquinoline alkaloid biosynthesis (Figure 7A), supporting the melanin nature of the dark pigment. Considering the polyphenolic structure of melanin, we then compared the relative content of the differential phenolic acids in white (WS) and black (BS) sesame seeds (Figure 7B). The result indicated that caffeic acid, indole-3-carboxylic acid, indole-5-carboxylic acid, protocatechuic acid, homogentisic acid, benzamide, 2,5-dihydroxybenzoic acid, 3,4-dihydroxybenzeneacetic acid, and ferulic and vanillic acid derivatives accumulated highly in black sesame seeds. Therefore, we speculated that they might be the precursors of melanin in sesame.