2.2.1. Physicochemical characterization
Adlay oil plays a significant role in various physiological and nutritional attributes [
8,
13], in addition to the general factors associated with oil in cereals [
16]. The average oil content estimated through nuclear magnetic resonance were similar with the values obtained from hexane extraction (4.5% ± 0.3) among different accessions of indigenous Adlay as illustrated in
Table 2.
This similarity in oil values by two different methods supports the accuracy of the measurement methodology employed. Differences in fat content among indigenous Adlay grains are readily apparent, driven by measurement, physical, geographical, and other inherent variations among the grains [
17,
18,
19,
20]. Black and white cultivars exhibit distinct oil contents, whereas variations in fat content between whole grain flour and degermed flour are minimal [
17]. Ding et al. [
18] also documented significant differences in fat content among cultivated Adlay varieties. Furthermore, defatted flour revealed an average protein content of 14.66% (± 1.14), with the highest protein content (15.82%) observed in BRN grains, as illustrated in
Table 2. In indigenous Adlay, the average protein content is higher than the maximum protein content of 13.78% found in the common wheat commercial variety, ‘Laasani 2008’ (personal communication). The differences in protein content among Adlay accessions seem closely linked to colour variations within the Adlay grains, specifically transitioning from green (GRN) to brown (BRN). It is notable that GRN Adlay grains tend to be relatively softer, which implies immature grains may have lower protein content [
18,
20]. Prior research has also highlighted substantial differences in protein content within various cultivated whole grain Adlay (ma-yun Stapf) varieties, which are widely consumed in China and Taiwan [
12]. However, no significant differences were detected among different compartments of grain [
18,
21]. Variations in protein content observed in Adlay can be attributed to several factors, including Adlay type, agro-ecological influences, genetic effects, biotransformation by fermentation, and the specific techniques employed for measurement [
14,
19,
20,
22,
23]. These factors collectively contribute to the differences in protein contents. Moisture and ash are considered crucial as they directly affect the stability and storage of food. The average ash value determined on a dry weight basis in indigenous Adlay collections was 2.24% ±0.36 (1.88% -2.63%), equally highest in OWT and GRN grains, followed by PRP (2.6%), as shown in
Table 2. Relatively higher ash contents are probably due to accessions hailing to Margalla hills areas. For example, husk colors or even grain compartments showed significant differences among ash contents [
18]. The digestible fiber average contents (2.02±0.2) are comparable to the previously reported 2.3% in cultivated Adlay and are found to be highest in BRN.
Among the indigenous adlay varieties, GRN adlay exhibited` highest P (0.3%), followed by GRY (0.27%) and OWT (0.25%). These differences, however, were not statistically significant. Notably, phosphorus levels have not been previously reported in hard-shelled adlay genotypes. Additionally, the levels of K in OWT (0.68%), and Ca in YLW (2.2%), were notably higher than the highest Na content (0.21%). These findings suggest the potential advantages of promoting indigenous adlay for their mineral contents. Boron is recognized as a crucial mineral influencing shell hardening, which, in turn, impacts milling quality [
24]. It was observed that BLK adlay contained the highest B (4 mg/kg), followed by BRN, GRN, and OWT varieties, respectively. The Boron levels 2.0-4.0mg/kg were found to be comparable to Indian cultivars. In contrast, P, K, Ca, and Na contents (0.3% to 2.2%), were notably higher than those typically found in cultivated adlay [
25]. These variations can be attributed to differences in soil chemistry, Adlay type, habitat, extraction method, and measurement techniques [
19,
22,
23,
25,
26,
27]. It’s worth noting that this is the first study to report Fe, Mn, Cu, and Zn contents in Adlay hard-shelled type. Prior studies had only assessed Fe, Cu, and Mn in soft-shelled Adlay, and Zn remained undetected in genotypes from the National Bureau of Plant Genetic Resources of India [
25]. Considering the strong relationship among the physicochemical characteristics of indigenous Adlay accessions as indicated in
Table 2, multivariate analysis was conducted using the statistical tool ‘Principal Components Analysis’ (PCA). PCA involves the generation of linear combinations of the variables, resulting in principal directions equal to the original variables. These principal directions are represented by vectors in the PCA-Biplot, as depicted in
Figure 3.
The analysis identified variables PRP and BLK Adlay accessions as PC1 and PC2, respectively. Notably, YLW clustered with BLK, while GRY clustered with PRP distantly apart from others, as shown in
Figure 3. These clusters accounted for a substantial proportion of the data variation, contributing nearly equally at 52.7% and 43.7%, respectively. This observation arises because these variables are more closely aligned with the axes and exhibit similar vector lengths. In contrast, parameters Zn and Cu appeared distant apart in the analysis. The PCA-Biplot presented in
Figure 3 represents the distribution of characteristics. It delineates these features into scattered positive loadings, as indicated by eigenvalues and eigenvectors, primarily driven by protein content. Conversely, negative loadings are associated with extracted oil content, Fe, and Na. Additionally, parameters such as ash, B, and Mn contents shape the data distribution within the biplot. Furthermore, the variables GRN, OWT, and BRN exhibit comparable vector lengths, and the angles between them in the biplot convey their inherent correlations in the multivariate space, as depicted above in PCA-Biplot
Figure 3.
2.2.2. Wild adlay triglyceride composition
Short and medium-chain fatty acids (SMCFAs) with carbon chain lengths fewer than fourteen have implications for human health [
28,
29]. In indigenous Adlay accessions, such fatty acids were not detected, as indicated by the chromatograms shown in the
Supplementary Figures S1–S7. Among the nine identified fatty acids, saturated fatty acids (SFAs) collectively constituted a significant portion (33.81% ±13.58%) of the total, with palmitic acid (C16:0) being the predominant SFA (26.3% ±4), followed by lignoceric (C24:0), behenic (C22:0), and arachidic (C20:0) acids, as illustrated in
Table 3.
As anticipated, unsaturated fatty acids (USFAs) expressed the dominant portion (53.1% ± 8.74) of the wild adlay triglycerides composition. The USFAs fraction mainly consists of PUFAs, with linoleic acid (C18:2n6C) being the major contributor (38.73% ±7.2) followed by MUFAs representing 7.42% ± 4.4 of the total USFAs. Oleic acid (C18:1n9C) was present as the prominent MUFAs (4.2% ±0.63) among the Adlay accessions, followed by eicosenoic acid (C20:1n9C) 3.13% ±0.4. Notably, other unsaturated fatty acids including palmitoleic (C16:1n9C), and eicosenoic acid (C20:1n9C) were present only in the YLW Adlay oils. Our results differ slightly from previously reported regarding the quantitative fractions of unsaturated fatty acids (85.1%) and their respective individual constituents [
30]. It is due to the measurement technique, adlay type and habitat etc. [
13,
22,
23,
27,
30,
31].
A striking observation was the higher distribution of PUFA (∑PUFAs) 38.73% ±7.2, making up a significant proportion of the wild adlay oil triglyceride composition in indigenous Adlay most accessions. The percentage contents of ∑PUFAs was 58.3% in YLW Adlay followed by GRY (55.1%), BRN (44.04%), GRN (38.04%), and PRP (33.0%). These values are comparable with earlier reports in cultivated Adlay [
30]. A similar study reported Triolein (1.04%) as the most abundant constituent in cultivated Adlay oil from Zhejiang province (China), and has isolated lipid markers consisting of various triglycerides, diglycerides, monoglycerides, sterols, glycerol trioleate, and fatty acids in Adlay seeds [
13,
22]. Lipid profile further revealed 32 peaks, consisting of 20 triglycerides and 12 diglycerides, 9 of them are useful in distinguishing the geographical origin of Adlay [
23,
32].
To comprehensively analyze the characteristics of nine identified fatty acids and their derivatives ΣUSFAs/ΣSFAs and ΣPUFA/ΣMUFA, PCA was carried out. PCA notably identified YLW, GRY, and BLK Adlay accessions as PC1, PC2, and PC3, respectively, with BRN, GRN, and OWT Adlay clustering together with GRY. YLW and GRY vectors are parallel to the axes and equal in length within the multivariate space, thereby making maximal and nearly equal contributions to the overall variation, accounting for more than 80% of the total (42% and 39% respectively), as indicated in the PCA-Biplot (
Figure 4).
The angle of PC3 (BLK Adlay) signifies its influence. Although it is not parallel to the axes and does not contribute to the same extent as PC1 or PC2, its vector length is substantial enough to warrant consideration in the variation summation. Moreover, PC1, PC2, and PC3 distribute characteristics into compact positive loadings, as revealed by linolenic acid, oil contents, long-chain saturated fatty acids (lignoceric, behenic, and arachidic acids), and the ratio of ΣUSFAs/ΣSFAs.
In contrast, other factors exhibit scattered loadings, with ΣMUFA in quadrant Q1, linoleic acid in Q3, and ΣPUFA in Q4. ΣUSFAs and ΣSFAs are distant apart from the rest of the data points. The angles between vectors YLW, GRY, BRN, GRN, and OWT represent their respective correlation coefficients in the multivariate space, with increased angles indicating proportional decreases in correlation as revealed by the reduced correlation associated with BLK Adlay and the complete lack of correlation, denoted by a wider angle for PRP Adlay.
2.2.3. Identification of functional groups or bonds in Adlay flour using FTIR
Fourier Transform Infrared (FTIR) is used for qualitative and quantitative assays with minimal sample preparation to outlook vibration of bonds within functional group leading to the metabolic fingerprint out of the molecule [
33]. Comparative IR spectra of seven indigenous Adlay flours have been produced in
Figure 5.
For simplicity, each spectrum in
Figure 5 has been viewed by two distinct frequency ranges according to its placement and application. Frequencies (υ expressed in cm
-1) possessing well-defined origins and are shared among all samples are enlisted in
Table 4.
The frequencies distributed unevenly among the samples and are elaborative only in combination with other frequencies, origin (s), or when multiple functional groups are considered simultaneously, due to their extensive data size, are illustrated in
Supplementary Table S1.
Frequencies consisting of 3350-3653 cm
-1, ~3000 cm
-1, 2348/2398 cm
-1, 1719/1721 cm
-1, 1702/1709 cm
-1, 1677/1678 cm
-1, 1134/1140 cm
-1, and 1088/1069 cm
-1 represent functional groups or bonds observable among all samples. A distinctive peak characterized by a single, broader/wider, or bold, or semi-global/rounded shape at 3350-3653 cm
-1 is always indicative of carboxylic acid, aldehyde/ketone, carbohydrates, aliphatic/acrylic alcohols, and or other related compounds [
34,
35]. Likewise, the well-known frequency at ~3000 cm
-1 is always associated with -CH3 or >CH2 symmetrical stretch (C-H bond), is an inherent feature of the fundamental molecular structure, universally present in organic molecules [
36]. Nevertheless, such frequencies are not usable as they are evenly distributed across all samples.
The frequencies which limit indigenous Adlay accession flours, enlisted in
Supplementary Table S1, can be categorized further into specific regions consisting of FPR (fingerprint region of protein) from 1500 cm
-1 to 1678 cm
-1 and NBR (narrow band region) 1188/1199cm
-1 to 1577cm
-1 in association with combination bands (“Comb band”) and or overtones (1600-2000 cm
-1). Additionally, there is the “OOP” (out of the plane region) 631/632 cm-1 to 943/945 cm-1, which intermingles with the FPR region to pose ring substitutions and ring isomerism. The comparative results of this data information are summarised in the ‘remarks column’ of
Supplementary Table S2. Further, when a frequency does not distinctly belong to a particular group, it is categorised as such [
27]. Briefly, vibrations in the range 410 cm
-1 to 2279/2281 cm
-1 indicate single or multiple functional groups present in a sample alone or multiple samples. The stretching frequencies 1134/44 and 1069/66 (vibration of ester group C-O bond) can be interpreted in conjunction with the FPR or OOP region [
34]. The frequencies at 738 and 687cm
-1 indicate ortho substitution, whereas 720 cm
-1 and 878 cm
-1 are significant in the context of 1,3,5-trisubstituted compounds. Similarly, frequency 1677/1678 cm
-1 suggests esters, dienes, trienes, or saturated alcohol groups present in six flours with the exception of PRP wild Adlay accession. The frequency 2158/2159cm
-1 is standard across all flours and indicates the isothiocyanate (-SCN) group [
33]. The frequency 2149cm
-1 is present only in BLK Adlay and marks R-N+≡C- or iso-nitrile group, whereas 1686cm
-1 is present only in GRY Adlay flour, indicating the R-HC=N- group [
33]. In both PRP and GRY Adlay flours, frequency 2863/2865 cm
-1 serves as a distinguishing feature attributed to H-C=O group, as elaborated in
Supplementary Table S2 in the remark’s column. With this approach, functional groups/ bonds are proportional to the number of vibration (s) frequency present in a spectrum provided factors such as resonance and or bonded hydrogen or others are considered the same and equally among all samples. BLK Adlay expressed a higher number of frequencies (53) corresponding to the equivalent number of functional groups followed by PRP and YLW (46 each), OWT, GRN and BRN (43 each) and GRY (42), respectively. Frequencies differences 1492/1493cm
-1 (~1500 cm
-1) - 656cm
-1 corresponding to serial number 26-35 (
Supplementary Table S2) profoundly discriminated flours in the context of FPR or NBR in association with Comb band and or overtones attributed of resonance and ring ‘OOP’ bending/scissoring, stretching frequencies.
In the regions 680-710 cm⁻¹ and 720-760 cm⁻¹, the absence of frequency implicates a lack of aromatic characteristics in lieu of the typical association of ring C=C bonds in this region. But vibrations at 745 cm⁻¹/747 cm⁻¹, 876 cm⁻¹/878 cm⁻¹, 877 cm⁻¹/905 cm⁻¹, 687 cm⁻¹, 738 cm⁻¹, and 410 cm⁻¹/419 cm⁻¹, coupled with the absence of well-defined peaks beyond 3000 cm⁻¹ due to various factors suppressing arenes, strongly suggest the presence of ring C=C bonds (arene). This interpretation is reinforced by the absence of frequency 895 cm⁻¹ - 915 cm⁻¹ and 985 cm⁻¹ - 995 cm⁻¹ attributed to alkene C=C bond vibrations (as illustrated in
Supplementary Table S2). From this insight, PRP and YLW Adlay flours contain ortho/para substitutions or a combination of both in lieu of frequency 745 cm⁻¹/747 cm⁻¹ or, more specifically, 1,2,3,4 or 1,2,4,5 or 1,2,3,5 or 1,3,4,5-tetra substitutions attributed to 876 cm⁻¹/ 878 cm⁻¹ frequency. OWT and BLK accessions exhibit a slight difference due to frequency 845 cm⁻¹/849 cm⁻¹ (indicative of para substitution) and, in some cases, both ortho/para substitutions. Moreover, OWT and BLK Adlay flours are distinguishable by 877 cm⁻¹/905 cm⁻¹ signifying penta- substitutions and frequency 410 cm⁻¹/419 cm⁻¹ pointing hexa- substitutions [
33]. In BRN Adlay, the combination of frequencies 687 cm⁻¹ and 738 cm⁻¹ suggests mono-substitution, 738 cm⁻¹ corresponds to ortho and/or 1,3,5-tri substitutions or vibration at 687 cm⁻¹ also indicates alkyl iodide (R-I stretch). GRY flour stands out with its unique absorbance of 878 cm⁻¹, signifying 1,2,3,4 or 1,2,3,5 or 1,3,4,5-tetra substitutions and/or penta- substitution. In the GRN flour spectrum, the vibration at 744 cm⁻¹ suggests ortho isomerism, whereas 827 cm⁻¹/828 cm⁻¹ indicates para isomerism and/or 1,2,3,4 or 1,2,3,5 or 1,3,4,5-tetra substitutions. The background frequency of 880 cm⁻¹ further supports the presence of penta-substitutions.