Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

The Diversity of Chemoprotective Glucosinolates in Moringaceae (Moringa spp.)


Glucosinolates (GS) are metabolized to isothiocyanates that may enhance human healthspan by protecting against a variety of chronic diseases. Moringa oleifera, the drumstick tree, produces unique GS but little is known about GS variation within M. oleifera, and even less in the 12 other Moringa species, some of which are very rare. We assess leaf, seed, stem, and leaf gland exudate GS content of 12 of the 13 known Moringa species. We describe 2 previously unidentified GS as major components of 6 species, reporting on the presence of simple alkyl GS in 4 species, which are dominant in M. longituba. We document potent chemoprotective potential in 11 of 12 species, and measure the cytoprotective activity of 6 purified GS in several cell lines. Some of the unique GS rank with the most powerful known inducers of the phase 2 cytoprotective response. Although extracts of most species induced a robust phase 2 cytoprotective response in cultured cells, one was very low (M. longituba), and by far the highest was M. arborea, a very rare and poorly known species. Our results underscore the importance of Moringa as a chemoprotective resource and the need to survey and conserve its interspecific diversity.


Glucosinolates (GS) account in part for the remarkable medicinal potential of Moringa oleifera (“moringa,” Moringaceae, Brassicales; apparently native to the sub-Himalayan lowlands in NW India)1. Glucosinolates (β-thioglucoside N-hydroxysulfates), mostly restricted to the angiosperm order Brassicales2, are metabolized by the enzyme myrosinase to their biologically active, cognate isothiocyanates (ITC)1,3,4. Isothiocyanates have long been known for their herbivore deterrent, fungicidal, bacteriocidal, nematocidal, and allelopathic properties5,6,7,8,9,10,11. Isothiocyanates such as sulforaphane from broccoli have antibiotic activity against numerous human pathogens including Escherichia coli, Salmonella typhimurium, Candida sp., and Helicobacter pylori12,13,14,15,16,17,18. These medicinal properties have been ascribed both to temperate cruciferous plants that are well-known sources of glucosinolates, and to Moringa oleifera, the most widely cultivated and economically important species of the monogeneric tropical family Moringaceae19,20,21,22. Because moringa is highly drought resistant, it can provide benefits to the large and often underserved human populations in the tropics and sub-tropics worldwide.

Many of the medicinal properties such as cancer treatment, regulation of blood glucose levels, and antibiosis that have long been ascribed to M. oleifera in traditional medicine are likely attributable to its glucosinolates or isothiocyanates22. For example, one of the main uses of M. oleifera in Ayurvedic tradition is cancer treatment23. The biomedical research literature now contains numerous animal studies showing preventive effects against carcinogenesis22 that are plausibly accounted for by known mechanisms of action of non-moringa GS and ITC7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25. Additionally, we showed that 4-α-L-rhamnopyranosyloxy)benzyl isothiocyanate (4RBITC), the isothiocyanate created by hydrolysis of “glucomoringin” (4RBGS or 4-(α-L-rhamnopyranosyloxy)benzyl glucosinolate) from M. oleifera is a potent and selective antibiotic against H. pylori15. Other studies have shown that the antibiotic activity of 4RBITC from M. oleifera is selective and potent against other important human pathogens such as Staphylococcus aureus and Candida albicans26. It also appears to be effective in controlling certain manifestations of both ALS and multiple sclerosis in mouse models27,28. A growing number of epidemiologic, animal, and clinical studies link dietary glucosinolates and their cognate isothiocyanates to protection against chronic diseases including a variety of cancers, diabetes, and autism spectrum disorder via the Keap1-Nrf2-ARE-mediated induction of phase 2 cytoprotective enzymes29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44. The coordinated Nrf2-mediated upregulation of this large group of enzymes is responsible for the very important indirect antioxidant activity of these isothiocyanates31,34,45,46,47,48,49,50.

It is likely that the chemistry of the uniquely rhamnosylated glucosinolates from Moringa spp. (compared to all 120 or so other glucosinolates) might provide special advantages to mammals consuming them at moderate levels22,37. Outside Moringaceae, rhamnosylated glucosinolates have only been documented in the related Resdaceae (Reseda spp.)51,52 and Brassicaceae (Noccaea caerulescens)53. One reason for particular interest in the rhamnosylated glucosinolates is the strong possibility of biologically significantly different absorption, distribution, metabolism, and excretion compared to other glucosinolates2,22,43,54.

With these considerations in mind, we screened glucosinolate diversity and phase 2 enzyme induction potential across Moringa. We focus primarily on adult leaves, because in M. oleifera and M. stenopetala leaves are the most commonly used parts of the plant, providing nutritious vegetables that are consumed both fresh and cooked. We also examine young leaves and the exudates from leaf glands of some plants. Glands on young leaves secrete a clear, sticky exudate that often attracts ants55. Because glucosinolates, via their cognate isothiocyanates, serve anti-herbivory and other protective functions, these compounds are often present in highest quantities at early, more vulnerable ontogenetic stages56,57. Young leaves, mature leaves, seeds, flowers, and extrafloral nectaries might all be expected to have differing proportions of glucosinolates. Our study thus provides a survey of the diversity and relative amounts of glucosinolates across the leaves, seeds, and exudates across this small but chemically and morphologically diverse family.

To the extent that morphological diversity reflects potentially different ways of interacting with herbivores, it is reasonable to presume that surveying across Moringa species could identify species with compounds that are even more efficacious than those currently known. In addition to the commonly grown M. oleifera, there are 12 other species in this monogeneric family. All of the species are native to the dry tropics of Africa, Asia, and Madagascar, with the center of diversity being the Horn of Africa at the intersection of Kenya, Ethiopia, and Somalia. The most commonly cultivated species M. oleifera is apparently native to northwestern Indian lowlands, and seems likely to have been domesticated in India thousands of years ago, with the domesticate differing markedly from the wild plants in its much faster growth rate, shorter maturation time, and softer leaflets. Its close relative M. concanensis is also native to the Indian subcontinent, where it is relatively widespread in dry tropical woodlands. The closest relative to the Indian species is M. peregrina, which is found from the Dead Sea south to the northern Horn of Africa. Four species with massive, water-storing trunks are found in Madagascar (M. drouhardii and M. hildebrandtii), Namibia and Angola (M. ovalifolia), and Kenya and Ethiopia (M. stenopetala). A group of closely-related species is restricted to the Horn of Africa, and is made up of medium sized to small trees with tuberous roots (M. arborea, M. rivae, M. ruspoliana), to dwarf shrubs to tiny herbs with massive underground tubers (M. borziana, M. longituba, M. pygmaea). Sampling across the diversity from massive trees to tiny herbs, we explore the diversity of glucosinolates in leaves, seeds, bark, inflorescences, and glandular exudates from both field and cultivated specimens of 12 of the 13 species.

We assess the glucosinolate contents of 12 of the 13 known species of Moringa, mostly in leaves (dried and fresh), but also when available seeds, stems, and leaf gland exudates. We document the occurrence of 2 heretofore unidentified glucosinolates as major components of 6 Moringa species, though they are not abundant in the 2 most common species (domestic M. oleifera and M. stenopetala). We report on the occurrence of simple alkyl glucosinolates in four species (M. peregrina, M. ruspoliana, M. rivae, and M. longituba), in one of which (M. longituba) they are the dominant glucosinolates. Finally, we document potent chemoprotective potential in most of the species, show that the activity comes from their glucosinolates, and measure the cytoprotective activity of 6 of these glucosinolates in purified form in a variety of cell lines. We could find no obvious relationships between growth habit, geography, phylogenetic relatedness or other obvious variables, and glucosinolate content or spectrum, underscoring the importance of conserving and studying all the species in the genus.


We evaluated the major GS in leaf samples from 14 Moringa taxa, including 12 of the 13 known wild Moringa species, plus samples from domesticated M. oleifera as well as a hybrid between M. concanensis and M. oleifera. Given availability, our sampling also included samples from seeds of 9 taxa, leaf gland exudate from 6, bark from one species, and inflorescence axis samples from another. All GS were initially separated by HPLC. Peaks were only purified and/or collected for mass spectrometry (MS) and nuclear magnetic resonance spectrometry (NMR) if the matching peak from a second otherwise identical extract disappeared after enzyme treatment with myrosinase, which is highly specific for GS. While confirming the presence of some compounds, our sampling revealed a rich diversity of compounds and concentrations across species, including two apparently newly identified GS. Associated with this diversity was a very wide range of potency of induction of the phase 2 detoxication response in mammalian cells, including some responses that were very high, surpassing that of the well-characterized isothiocyanates such as sulforaphane.

We confirmed the presence of the previously identified rhamnose-containing glucosinolate (4RBGS 1m/z 570.0957) and its acetylated derivatives (e.g. 2m/z 612.1067) as the predominant GS in Moringa oleifera domestic and M. stenopetala (Fig. 1). Also in these two species, we identified two multiply glycosylated GS which to our knowledge have never before been described. They are shown as 3 & 4 in Fig. 1. They contain a glucopyranosyloxy-benzyl moiety as their R-group (not described previously), as opposed to the rhamnopyranosyloxy-benzyl moiety that has been previously reported in GS from Moringa spp. Upon further investigation, we found this glucopyranosyloxy-benzyl GS (3) in all Moringa species except M. hildebrandtii. We found the former, well-documented rhamnopyranosyloxy-benzyl GS (1) in all species except M. longituba.

Figure 1
figure 1

The predominant glucosinolates (GS) in the genus Moringa were isolated from plant organ extracts of the 12 Moringa spp. described herein, by complementary HPLC techniques, and identities were confirmed initially by direct mass spectroscopy (m/z), and by MS/MS. Accurate mass determination is given in parentheses under the nominal mass designation. NMR was utilized to confirm the identities of compounds 1 and 2 which are described in the literature, and of 3 and 4 which to our knowledge have heretofore not been described. GS 5 is not found in Moringa spp. but is an isomer of 1 that we isolated from Reseda odorata and provide by way of comparison. The biological activities of ITCs generated from these GS are presented in Table 1.

We obtained unequivocal structural identification of these compounds from nominal and accurate masses via electrospray MS, from LC/MS/MS (m/z 586.0900 and m/z 628.1014 for compounds 3 and 4 respectively), and by 400 and 600 MHz NMR according to methods described previously58 (Supplemental Data & Supplemental Fig. S1).

Given nomenclatural disagreements in the literature and configurational differences (e.g. isomers that are acetylated at the 2′ or 3′ position on the side-chain sugar (rhamnose or glucose), rather than at the 4′ position we have indicated, we refer in the ensuing presentation of data only to “monoacetyl-4-(α-L-rhamnopyranosyloxy)benzyl-GS” or “monoacetyl-4-(α-L-glucopyranosyloxy)benzyl-GS” for the sake of simplicity and accuracy. In Fig. 1, we show only 4′-O-acetyl-4-(α-L-rhamnopyranosyloxy)benzyl GS (2) and 4′-O-acetyl-4-(α-L-glucopyranosyloxy)benzyl GS (4). In all of our samples the molecular weights of compounds 2 and 4, the acetylated derivatives of compounds 1 and 3 respectively, were consistent with the identities provided.Other GS were identified in Moringa spp. that were consistent with those described previously52,59,60. Identities were matched with authentic standards where possible and accurate masses were obtained by mass spectroscopy of HPLC peaks. As noted above NMR confirmation was also obtained with compounds 1, 3, and 4. Thus, in addition to the GS noted above, we identified benzyl GS 6, and 1-methylethyl-GS 7. These GS are all shown in Fig. 1. The final GS found in the 12 Moringa species examined was not unequivocally identified and has been referred to in Fig. 1 as “alkyl-GS”. This compound or compounds has an accurate mass of 390.0534, which is consistent with 5 different isomeric GS: 3-hydroxybutyl GS, 4-hydroxybutyl GS, 1-(hydroxymethyl)propyl GS, 2-hydroxy-2-methylpropyl GS, and 1-ethyl-2-hydroxyethyl GS2. This GS was only found in M. longituba, a dwarf species with red, tubular flowers unique in Moringaceae, from Kenya, Ethiopia, and Somalia.

We also have included a non-Moringa GS in Fig. 1 for comparison. It is a benzyl GS that is rhamnosylated at the ortho- or 2- position on the benzyl ring (2-(α-L-rhamnopyranosyloxy)benzyl GS; 5). We have isolated this compound from Reseda odorata, which like Moringa is a member of Brassicales. The purified compound a structural isomer of compound 1 from Moringa in which the rhamnose group attaches at the 4- or para- position of the benzyl ring, is a much less potent inducer of cytoprotective enzyme than compound 1, as shown in Table 1.

Table 1 Phase 2 enzyme inducer potencies of selected isothiocyanates (ITC) produced by myrosinase hydrolysis of purified glucosinolates (GS).

Detailed comparisons of silica gel dried and fresh leaves in M. oleifera found that silica gel drying satisfactorily preserved GS levels and myrosinase activity61. We therefore treated silica gel and fresh leaf samples as equivalent and pooled them in subsequent analyses. We then asked whether young leaves, which had not finished expanding and were largely non-lignified, differed in their glucosinolate levels from mature leaves, which had ceased expansion and were more lignified. We also found no significant differences, though p-values were often marginal, suggesting that more detailed analysis might detect different profiles between mature and immature leaves (Table 2). Finally, we observed significant differences in at least one species with respect to the others for each of the GS (Table 2). These differences across species are most obvious in the leaves, of which we were able to secure the most extensive collection of plant material, allowing us to make some strong comparisons between species and across GS (Fig. 2). Though sampling was relatively limited, our results suggest major quantitative differences between plant organs in their GS content and spectrum. Leaves differ both quantitatively and qualitatively in their GS profiles from seeds (9 species compared), and from leaf gland exudates (7 species compared) (Fig. 3, Supplemental Table S1).

Table 2 Wilcoxon signed-rank tests for differences in glucosinolate levels between silica-gel dried and fresh, and mature and immature leaves, and Kruskal-Wallis tests for differences in glucosinolate levels between species.
Figure 2
figure 2

Differences in percentages of the glucosinolates (GS), by glucosinolate, across leaf samples of all species (species designations indicated in the lower left panel of this Figure). (A) Alkyl-GSs (1 or more of 5 different isomeric GSs not represented in Fig. 1 –3-hydroxybutyl GS, 4-hydroxybutyl GS, 1-(hydroxymethyl)propyl GS, 2-hydroxy-2-methylpropyl GS, and 1-ethyl-2-hydroxyethyl GS); (B) Methylethyl-GS (compound 7); (C) Benzyl-GS (compound 6); (D) 4RB (compounds 1 and 2 combined); (E) 4GB (compounds 3 and 4 combined). Different letters at the tops of 2D and 2E indicate statistically homogeneous groups identified by Kruskal-Wallis test followed by post-hoc comparisons.

Figure 3
figure 3

The 13 known species of Moringa (family Moringaceae) are shown on the left, with a cladogram showing their relationships. Bars are color coded to represent approximate relative amount of each glucosinolate (GS)(sample size “N” given at right), in the cases of leaves and seeds, and only very small fresh leaf exudate collections. Differences in shading intensity corresponds to percentages of glucosinolates. GSs are grouped along structural lines in some cases, thus: MeEt (compound 7); Alkyl (one or more of 5 different isomeric GSs not represented in Fig. 1: 3-hydroxybutyl GS, 4-hydroxybutyl GS, 1-(hydroxymethyl)propyl GS, 2-hydroxy-2-methylpropyl GS, and 1-ethyl-2-hydroxyethyl GS); Benzyl (compound 6); 4RBGS (compounds 1 and 2); 4GBGS (3 and 4). Note that source material for M. rivae included shoot apices.

Overall, cytoprotective enzyme inducer potency for 11 of 12 Moringa leaf extracts was comparable to that observed for broccoli seeds, which are the most potent plant source of this activity62. Potency was highly correlated with the plants’ content of 4RBGS 1 on both a dry and a fresh weight basis (Pearson’s r = 0.850 and 0.844 respectively) (Fig. 4). Moringa longituba was unique among the species examined in that it contained predominantly the very simple alkyl side-chain GS as already described in this section, none of which are very potent inducers of the Keap1-Nrf2-ARE mediated cytoprotective enzymes, typified by NAD(P)H:quinone oxidoreductase 1 (NQO1) when assayed separately (data not shown). Interestingly, it also contained no detectable 4RBGS.

Figure 4
figure 4

NQO1 inducer potential strongly correlated with 4RBGS content. Single extracts of representative leaves from each of the 12 Moringa species were assayed by the “Prochaska assay”47,85 with added myrosinase to convert glucosinolates (GS) to their biologically active cognate isothiocyanates, and to then determine relative potency as an inducer of the phase 2 cytoprotective response. Aliquots of the same samples were evaluated by HPLC to determine concentration of 4RBGS 1. Results are reported as NQO1 inducer potential (million Units). Both dry (; blue) and fresh (; red) leaf samples were assayed, however the latter were normalized to a dry weight basis by assuming 75% moisture content, in order to overlay data. (The mean moisture content of experimentally measured values ranged from 70 to 80%, with a mean of 75%). Potency was highly correlated with 4RBGS content (Pearson’s r = 0.825 for the overall regression).

Intrigued by the almost 10-fold stronger NQO1 induction in Hepa1c1c7 cells by 4RBITC compared to sulforaphane (4-methylsulfinylbutyl-ITC), which comes from broccoli and is not present in Moringa spp., uptake of the ITCs 4RBITC and sulforaphane was compared in a number of cell lines. A representative uptake curve in cultured Hepa 1c1c7 cells is presented (Fig. 5), in which kinetics were similar, and “area under the curve” (AUC; a commonly used pharmacokinetic metric) was essentially the same as that produced by equimolar quantities of 4RBITC, making it appear that factors beyond differences in uptake must account for the large difference in potency between 4RBITC and sulforaphane.

Figure 5
figure 5

Pharmacokinetics of uptake of 4RBITC (produced from compound 1) and sulforaphane (Inducer potency of SF is similar whether produced in situ by myrosinase, or purchased as either R- [natural] or RS- [synthetic] SF added to culture medium for Hepa1c1c7 murine hepatoma cells). Area-under-the-curve (AUC) is very similar for the two compounds, 57.77 for SF and 62.62 for 4RBITC, but 4RBITC does not reach its maximum level in cells until 4 hours, whereas SF has already peaked by 1 hour post-addition. 4RBITC is about 10-fold more potent an inducer in this cell line. The difference in uptake kinetics may be reflected in greater potency. Values for duplicate determinations at each time point are connected by solid vertical lines.


Our results document very wide variation in GS profiles and potency of induction of mammalian phase 2 cytoprotective enzymes across Moringa species. These results illustrate the importance of examining as wide an array as possible of germplasm, both domesticated and wild, because the variation across species greatly exceeded that within the cultivated domestic variants of M. oleifera, by far the most commonly cultivated and readily obtainable species. Underscoring this diversity, our survey highlights two heretofore unrecognized GS, one very abundant in the putative wild ancestor to M. oleifera. We discuss this diversity and show how variation in GS profiles is of great potential applied importance, with the potency of phase 2 enzyme induction varying far beyond that known of M. oleifera, the only species studied to date. We identified a species with much higher phase 2 enzyme induction potential than M. oleifera, as well as species with no or very low phase 2 response. We conclude by emphasizing the urgency of conserving all of the species in this genus, not just the relatives of the domesticated M. oleifera.

A New Glucosinolate within M. oleifera

Sampling widely across the family highlighted the presence of heretofore unrecognized glucosinolates. One of these compounds, 4-(-L-glucopyranosyloxy)benzyl GS (4GBGS), is very abundant in the wild type of M. oleifera, and because we first found it in this species, we name it after this taxon. The most abundant glucosinolate in M. oleifera is known as glucomoringin. Our name employs a parallel construction based on the common name of the wild type of M. oleifera in its area of natural occurrence. Though the putative wild type of M. oleifera is grown widely across eastern Pakistan and lowland northern India from Haryana, Punjab, and Jammu and Kashmir east to Bihar [Garima, Olson, and Nouman unpublished observation], it is native only to a small area. This area is found only on the Punjab-Himachal Pradesh border in northwestern India where the most frequent common name for the plant is “soonjna.” We assign the common name for this glucosinolate as “glucosoonjnain.” The common name for the isothiocyanate resulting from its hydrolysis would thus be “soonjnain”.

Previous surveys of glucosinolates in M. oleifera63 did not find evidence of glucosoonjnain because they all examined domestic, rather than wild type material. Our detailed survey of 4GBGS and 4RBGS across domestic and wild type accessions of M. oleifera61 showed marked differences in profiles. While both glucosinolates are present in both species, 4GBGS is present in very low amounts in domestic M. oleifera but is the most abundant one in wild type M. oleifera. The opposite is true of 4RBGS, which is present in very low amounts in wild type M. oleifera but is the most abundant one in domestic M. oleifera. Because the two variants differ markedly in taste61, it is possible that these dramatically different profiles between the putative wild ancestor and the domesticate are due to selection favoring more agreeable taste. That a novel glucosinolate was waiting to be discovered in the best-known species, M. oleifera, emphasizes the importance of examining as diverse an array as possible of samples62,64, both within and across species.

In addition to 4RBGS, first documented by Badgett65 and later reported by others60 the following other GS have been reported to occur in M. oleifera: 4″-acetyl-4-RBGS66, benzyl GS66; isopropyl- or methylethyl GS, 2-methylpropyl GS, isobutyl GS20,59,60,67; and more recently 4-(2′-O-acetyl-α-L-rhamnosyloxy)benzyl ITC68,69, 4-(4′-O-acetyl-α-L-rhamnosyloxy)benzyl ITC43; 2′-O-acetyl-α-L-rhamnosyloxybenzyl ITC70; 3′-O-acetyl-α-L- rhamnosyloxy)benzyl ITC70; 4′-O-acetyl-α-L-rhamnosyloxy)benzyl ITC70. Note that nomenclature of the acetylated derivatives is inconsistent but the most often reported of these compounds is 4′-O-acetyl-4-(α-L-rhamnopyranosyloxy)benzyl GS (2).

Glucosinolate Diversity Across Species

In addition to novel GSs, our results highlight remarkable variation across species in their glucosinolate profiles (the relative abundances of the different GS). The well-known domestic form of M. oleifera had as its predominant GS the relatively well-characterized 4RBGS (“glucomoringin”), as did species such as M. arborea, M. hildebrandtii, and M. stenopetala. Other species, including the wild type of M. oleifera, M. concanensis, M. longituba, and M. rivae, had dependably low levels of 4RBGS. Moringa longituba, a dwarf species from Kenya, Ethiopia, and Somalia with a slender aboveground stem springing from a large belowground tuber, appears to have no or trivial amounts of 4RBGS, unique in the family. The newly-documented 4GBGS (“glucosoonjnain”) varied conspicuously across species, and was especially prominent in accessions of M. ovalifolia and M. rivae, and in wild type M. oleifera. 4RBGS and 4GBGS were present in almost all species, and the remarkable variation in the levels of these GS across accessions and across species strongly suggests that the levels of these compounds is malleable under selection, giving promise to efforts to shape the levels of these GSs under breeding programs.

In contrast to 4RBGS and 4GBGS, which were widespread across the genus, the other main categories of GS were observed only in certain species. The most widespread of these was 1-methylethyl-GS (compound 7, Fig. 1). Absent from most species, this compound was observed in M. ruspoliana, a species with very large leaflets from the Horn of Africa, as well as M. peregrina, a species with minute, deciduous leaflets found in the Arabian Peninsula and adjacent areas. Although levels of methylethyl-GS were consistently very low in domestic M. oleifera, it was often present in appreciable proportions in the wild species from the Indian subcontinent, M. concanensis and M. oleifera wild type. More remarkably, compound 6, benzyl-GS, was only observed to any appreciable degree in young leaves of Moringa rivae, a poorly-studied shrubby species from remote drylands of Kenya, Ethiopia, and Somalia. In addition to benzyl GS, M. rivae also had appreciable levels of 4RBGS and 4GBGS. 4RBGS and 4GBGS were nearly absent from M. longituba, as were all other GS except for the alkyl GS. The alkyl GS, in turn, were absent in all other species. With both structural variation and variation in relative proportions across species, our results highlight the vast variation potentially available for breeding across the small number of taxa studied.

Glucosinolates Differ with Ontogeny

In addition to variation across species, our data suggest that GS profiles may vary between mature and immature leaves. We did not, however, detect statistically significant differences between mature and immature leaves in the proportions of any of the GS structural categories (Table 2). In the case of methylethyl-GS and 4RBGS, the p-values were nearly significant. Since GS are favored in herbivore defense in poorly lignified ontogenetic stages, it seems plausible that more detailed study will recover significant differences between immature and mature leaves. Inspection of the patterns in our data (Fig. 3; Supplemental Table S1) suggest that in most species (M. arborea, M. borziana, M. drouhardii, M. hildebrandtii, M. oleifera domestic, M. peregrina, M. ruspoliana, and M. stenopetala), 4RBGS was the predominant GS in young leaves. Only in M. longituba, M. ovalifolia, M. rivae were GS other than 4RBGS predominant in young leaves. Our sampling of immature leaves included very small leaves still undergoing some cell division, as well as much larger ones mostly undergoing cell expansion and cell wall lignification. These observations closely parallel our observations in Brassica oleracea var. italica (broccoli) in which GS profile changes dramatically in ontogeny10,71, with genetics72,73, and environment71,73. This heterogeneity across the ontogenetic stages sampled here means that there is opportunity for more detailed examination of the ways that GS profiles vary from emergence to maturity in leaves of mature trees, as well as how GS vary in Moringa from seedlings to saplings.

Glucosinolates Differ Across Organs

Just as different ontogenetic stages of the same organ could be expected to differ in glucosinolate profiles, organs with different roles can also differ, and our results are suggestive in this regard. Seeds (n = 9) were the non-leaf organ for which we had the most data. Inspection of Supplemental Table S2 suggests that the profile of GS in seeds generally parallels that of those in mature leaves. Leaf gland exudate was available from 6 species, and in contrast to seeds, exudates differed from leaves in their GS profiles. Glucosinolate diversity was always very low in the exudates, usually limited to a single GS with the exception of M. borziana, which bore two. The GS present in exudate were always the predominant GS in the leaves from a species. Moringa arborea, M. drouhardii, M. peregrina, and M. stenopetala exudates had as their only glucosinolate 4RBGS, which was the predominant GS in leaves (and seed, when available). Similarly, exudate in M. longituba contained only alkyl glucosinolate(s), which are also predominant in M. longituba leaf and seed profiles. Only in M. borziana leaf exudates were two GS (4RBGS and 4GBGS) observed. While 4RBGS predominated in the leaves of the plants of the same provenance as the leaf gland exudate, the proportions were nearly equal in the leaf gland exudate. Finally, we collected bark from a M. longituba population and from M. arborea because the plants were leafless at the time of our field visit. The bark sample of M. longituba is noteworthy because it was the only sample of that species that had appreciable levels of 4RBGS. The inflorescence axis of M. arborea had slightly higher levels of 4GBGS than the other samples but otherwise paralleled the leaf and exudate profiles. Though the sample sizes were small, our results comparing leaf glucosinolate profiles with other plant parts are sufficient to suggest strongly that exploring not only across ontogenetic stages but also across plant organs is very likely to recover biologically significant variation in glucosinolate profiles.

Glucosinolates and Phylogeny

There was no tendency for the GS profiles of closely related species, or species of similar habit or habitat to resemble one another more than GS profiles of species that were distantly related or dissimilar in habit or habitat. Perhaps most notably, the domestic M. oleifera differed strongly from the wild type in having high levels of 4RBGS and low levels of 4GBGS. Like wild type M. oleifera, the other wild Indian species, M. concanensis, also had much higher levels of 4GBGS than the domestic, an example of presumably close relationship not predicting similarity of GS profiles. Another striking example is provided by M. longituba and M. ruspoliana, which are sister species, i.e. sharing a most recent common ancestor and each other’s closest living relative. They share morphological features from wood anatomy to flower color and leaf texture. However, their GS profiles were maximally different, with alkyl GS, abundant in M. longituba, not detected in M. ruspoliana, and methylethyl GS, which are relatively abundant in M. ruspoliana, entirely absent from M. longituba. 4RBGS, which is abundant in M. ruspoliana leaves, was observed only in leaf gland exudate in M. longituba.

A major advantage often cited for phylogenetic classifications is that they are predictive. Predictive in this sense means that phylogenetic affinity should predict similarity in characteristics such as morphology or phytochemistry. While this is surely true at the ordinal level—Moringa species are more similar to one another in their GS profiles in general than they are to members of any other family in the order—it is not so within the family. It is essentially impossible to predict the GS profile of one species given its phylogenetic position and the profiles of the other species. This marked variation in important biologically active compounds strongly underscores the need both to study comprehensively the diversity in the family as well as the need to conserve the species. With no way of predicting the GSs present, there is no way of knowing what is lost if any taxa or populations become extinct before they are studied.

Glucosinolate Structure and Function

Comparing the rhamnopyranosyloxylbenzyl- bearing GSs in Moringa with those in Reseda, the only other plants known to produce such GSs, provides clues regarding the molecular structural features that are correlated with biological activity in Moringa. Moringa oleifera is of interest because its leaves have high levels of 4RBGS, a GS with a unique rhamosyloxy residue in the side chain. Myrosinase hydrolysis of 4RBGS leads to a highly chemoprotective ITC, such that in certain cell types it is a more potent inducer of the prototypical phase 2 enzyme response than sulforaphane from broccoli sprouts, the most potent naturally occuring phase 2 inducer known to date31,37. Moringa arborea, a poorly known species that has only been seen twice by botanists, induced an extremely high cytoprotective response, and was correspondingly rich in 4RBGS. It seems plausible that structural variation in the rhamnosyloxy moiety could affect functional aspects such as lipophilicity, solubility, and bioavailability47.

We examined the functional impact of variation in the position of the rhamnosyloxy moiety by measuring the phase 2 induction capacity of Compound 5, which is from Reseda odorata and is not known from Moringa. As compared to 4RBGS, Compound 5 showed 40-fold lower induction capacity in Hepa1c1c7 cells and 4-fold lower in RAW264.7 cells, did not appreciably change it in PE or ARPE-19 cells, and actually increased by 3-fold its ability to induce glutathione synthesis in Hepa1c1c7 cells (Table 1). 4RBGS and Compound 5 are isomers that differ only in the position of the rhamnosyl group on the side-chain benzyl ring. In 4RBGS, the rhamnosyl group is in the the para- (or C4- position; see Compound 1 in Fig. 1), whereas it is at the ortho- (or C2- position) in Compound 5; Fig. 1). This positional difference is associated in most cases with dramatically lower chemoprotective potency of Compound 5 as compared to 4RBGS. The meta- or 3-substituted isomer has to our knowledge never been identified in the plant kingdom or elsewhere, but examining its activity if found or synthesized is a priority. With regard to the activities of other Moringa compounds, neither the alkyl GS 7, nor the aryl GS 6 (Fig. 1) were very potent inducers of cytoprotective enzymes. Likewise, the new “glucosoonjnain” (compound 3), is 20-, 2-, 3-, and 3-fold less potent than the rhamnosylated congener (compound 1) as an inducer of the cytoprotective enzyme NQO1 in Hepa1c1c7, RAW 264.7, PE, and ARPE-19 cells, respectively. Though the mechanism remains to be elucidated, our findings implicate the rhamnosoxyl group and its position on the benzyl ring as strongly affecting biological activity.

How best to ingest Moringa to maximize the health benefits deriving from its GSs has been little studied, but it is likely to be similar to other GSs, for which abundant information is available. Myrosinase, as with most other enzymes, is irreversibly inactivated by the heat of cooking3,74,75,76, and so the co-administration of myrosinase can potentially offset some of the lack of myrosinase in cooked vegetables77. As a result, consuming Moringa leaves with minimal heat, either raw or only barely cooked, would seem to maximize chances of receiving biologically significant doses of ITC. Crushing of leaves before consumption is also expected to raise ITC levels significantly. Though most of these observations have not been replicated with the GS-myrosinase-ITC system in moringa, it is extraordinarily unlikely that these observations would not also apply to this genus.


We must now start to better understand not only why these compounds are produced by plants (for their defense, as both anti-feedants and attractants), but how and why different relative quantities are made, what cues changes in these ratios, and most importantly, how each of them works to alter human healthspan when ingested or applied topically. Much work has been done clinically with broccoli sprouts and sulforaphane, but 4RBITC is even more potent than sulforaphane in some assays (Table 1, and refs15,25,47). Further understanding of the biological properties of these compounds, and of their pharmacokinetics and pharmacodynamics in human beings, will permit more rational and targeted prescriptive and personalized nutrition – “green chemoprotection” or “frugal medicine”37,38. Identification of Moringa GS has with few exceptions been limited to the most commonly cultivated species M. oleifera and M. stenopetala. Knowledge of the unique GS and powerful chemoprotective activity from members of this genus should bolster efforts to prevent the impending extinction of these invaluable additions to our medical and nutritional armamentarium78.

Materials and Methods

Plant Collection

Samples of 12 of the 13 species of Moringa were collected in the field or obtained from cultivated specimens. Pressed, dried herbarium specimens were prepared as vouchers for most collections. Most of the species grow in remote and difficult to access localities, so this is the first time that material of so many species has been assembled for study of the glucosinolate diversity in the group. This rarity means that some species were available only in very limited quantities. For example, the collections here of M. arborea represent only the second time the plant has ever been seen by scientists. With regard to M. pygmaea, this species is native to a very remote area of the Somaliland-Puntland border79 and samples were not available. The plants sampled and locality information are listed in Supplemental Table S2. Voucher specimens are deposited in the Missouri Botanical Garden, Kew, East African, Mexican National, and other herbaria. Samples indicated in Supplemental Table S2 as leaflets dried in silica gel were collected following protocols routinely used for the collection of plant material for DNA extraction80. Approximately 10 dried leaflets were ground with a small amount of sterile silica sand. A subsample was used for DNA extraction for phylogeny reconstruction80. Some plants were leafless at the time of collection. Therefore, one sample of M. longituba (Olson 710) was prepared by separating the bark from the xylem cylinder and drying the bark components that remained after separating the phellem. A sample of M. arborea (Olson 714) was prepared from an inflorescence axis. Samples not collected in the field were from plants grown in the Biology Department greenhouse at Washington University, St. Louis, Missouri, or the International Moringa Germplasm Collection in Jalisco, Mexico. Glandular exudates were collected from actively-secreting glands on young leaves with sterile glass capillary tubes. Fresh leaves were also gathered from greenhouse grown plants, packed between moist paper towels, and overnight shipped on ice to Johns Hopkins University (Baltimore, MD, USA). To test ontogenetic differences in GS distribution, we gathered both young and mature leaves. Young leaves were the most apical available and had not completed expansion, whereas leaves considered mature were the lowermost healthy leaves available on the stems, had ceased expansion, and contained more lignified tissue.

Glucosinolate Extraction

GS extracts were made by homogenizing either dry plant tissues ground with an equal weight of silica sand, or fresh leaves, for 3 minutes in cold solvent (equal parts of acetonitrile, dimethyl sulfoxide, and dimethylformamide)62 or in 80% boiling methanol using a Polytron Homogenizer (Brinkman Instruments, Westbury, NY, USA) at ½ speed for 3 min. Homogenates were centrifuged (5600 × g for 3 min at 25 °C) and the supernatant was stored at −20 °C until analysis.


Extracts were analyzed by Hydrophilic Interaction Liquid Chromatography (HILIC) according to published methods81,82. Briefly, confirmation of GS identities was performed by comparison to analytical standards prepared as described83 and by using a complementary protocol for HPLC of intact GS followed by mass spectroscopy58. All HPLC was performed with the plant extracts, followed directly by the same extracts after incubation with exogenous myrosinase, which uniquely degrades GS, thus either reducing or eliminating only peaks attributable to GS. Thus, HPLC peaks were judged to be GS only if the spectrum of an isolated peak eluted by HPLC was characteristic of the specific GS, and if repeated chromatography of the extract after hydrolysis with purified myrosinase resulted in a substantial diminution and/or disappearance of the peak. Multiple pooled HPLC peaks were collected for direct injection mass spectroscopy. If, following mass spectroscopy in two separate systems, both absolute mass and MS/MS transitions, were not congruent with known GS, NMR was performed to anchor structural assignments.

Analysis: Mass Spectroscopy

Confirmatory analysis was performed on all GS, first using mass spectroscopy (MS) in both the positive mode with electrospray ionization tandem MS (ESI-MS/MS using a Thermo-Finnigan TSQ Advantage Triple Quadrupole MS coupled to a Thermo-Finnigan Accela UPLC and HTC Pal autoinjector)84 and the negative mode (capillary temperature 320 °C, sheath gas 20 a.u., spray voltage 4 kV)36 and with a PE Sciex Q-Star Hybrid Quadrupole/time-of-flight mass spectrometer. With these protocols, masses (m/z), and MS/MS transitions were matched with expectations and with known standards36,84. Absolute mass was ultimately obtained for all compounds. Accurate mass analyses were performed by pooling analytical HPLC peaks obtained from sequential injections, concentrating, and introducing directly on an Agilent 6520B quadrupole time-of-flight (QTOF) mass spectrometer in negative mode with drying gas at 350 °C, 11 L/min and 40 psi, in MS mode. This has a mass accuracy of 2 ppm (ca. ± 0.001 amu), and in practice these measurements are usually within ± 0.0003 amu.

Analysis: NMR Spectroscopy

Proton and carbon NMR data were acquired at 23 °C with a 400 MHz Mercury spectrometer equipped 5 mm z-axis gradient probe. Samples were dissolved in 600 μL of D2O at concentrations of 1–10 mM. The observed 1H chemical shifts are reported with respect to the residual H2O/HOD signals, which is at 4.80 ppm downfield from external sodium 3-(trimethylsilyl) propionate-2-2-3-3-d4 (TSP) in D2O. One dimensional 1H, 13C natural abundance, and two-dimensional 1H-1H COSY were acquired for each glucosinolate sample.

Cytoprotective Phase 2 Enzyme Induction Bioassay

Extracts were diluted 200-fold into microtiter plates for bioassay as described47,62,85. Briefly, induction of NQO1 catalytic activity was measured in Hepa 1c1c7 murine hepatoma cells grown in microtiter plate wells each containing 150 µL of medium. Eight or more replicates of 1:1 serial dilutions in culture medium were assayed. To measure total inducer activity in plant extracts [from GS and their cognate isothiocyanates], excess purified myrosinase3,4 and 500 µM ascorbate to activate the myrosinase were added directly to the microtiter plates. One unit of inducer activity is the amount that doubles NQO1 specific activity in 48 h in a well containing 150 µL of medium. Other cell lines were tested in the same manner. Cell lines were maintained as described previously and reviewed in47 and detailed as referenced: Hepa1c1c7 murine hepatoma cells62, RAW264.7 murine macrophage cells86; PE murine keratinocytes87, and APRE-19 human retinal pigment epithelial cells45.

Cellular Uptake Studies

Hepa 1c1c7 cells were seeded at 5 × 105 cells per well in 3 mL culture medium in 6-well (3.5 cm diameter) tissue culture plates for 24 h. Uptake was then monitored over the next 48 h following addition of 10 μM final concentration of each of the isothiocyanates delivered in DMSO (0.1% final concentration). Wells containing ITC but no cells served as blanks; DMSO was added to control wells containing cells only. Culture medium was aspirated and cells were harvested at time points up to 24 h as previously described, using a method designed to rapidly remove all traces of cell culture medium and the ITC contained therein, from cells88. Briefly, cells were immediately layered onto a 1:1 mixture of diisononyl phthalate and dibutyl phthalate, centrifuged gently in a swinging bucket centrifuge, chilled, re-suspended in water, stored at −70 °C until assayed, then thawed, sonicated and ITCs were measured by HPLC14.


All reagents used were analytical or HPLC grade and were purchased from Sigma/Aldrich Chemical Company, Inc., St. Louis, MO., USA) or from Fisher Scientific (Fairlawn, NJ, USA), USB/Affymetrix (Santa Clara, CA., USA; G6PD), Gibco Life Technologies Corp. (Grand Island, NY, USA; FBS & culture media).

Data Analysis: Glucosinolate Profiles

Given the uneven distribution of sample sizes across species, we used Wilcoxon signed-rank tests to compare the levels of GS between silica gel dried versus fresh leaf samples, with data paired by species. For these tests, we used mean values of each GS for silica gel dried and fresh leaves per species. Because they represented the bulk of the sample, we used only mature leaves for these tests. We then tested for differences between immature versus mature leaves, using a similar procedure. Finally, we tested for differences between species in each of the GS categories using Kruskal-Wallis tests followed by non-parametric multiple comparisons. Analyses were carried out in R v.3.3.189.

Ethics Approval

Not applicable since no human or animal studies were conducted herein.


  1. 1.

    Olson, M. E. Moringaceae: Drumstick Family. Pp. 167–169 in Flora of North America Editorial Committee, eds 1993+. Flora of North America North of Mexico. 15+ vols. New York and Oxford. Vol. 7 (2010).

  2. 2.

    Fahey, J. W., Zalcmann, A. T. & Talalay, P. The chemical diversity and distribution of glucosinolates and isothiocyanates among plants. Phytochemistry 56(1), 5–51 (2001).

    Article  PubMed  CAS  Google Scholar 

  3. 3.

    Shikita, M., Fahey, J. W., Golden, T. R., Holtzclaw, W. D. & Talalay, P. An unusual case of ‘uncompetitive activation’ by ascorbic acid: purification and kinetic properties of a myrosinase from Raphanus sativus seedlings. Biochem. J. 341(Pt 3), 725–732 (1999).

    PubMed  PubMed Central  CAS  Google Scholar 

  4. 4.

    Wade, K. L., Ito, Y., Ramarathnam, A., Holtzclaw, W. D. & Fahey, J. W. Purification of active myrosinase from plants by aqueous two-phase counter-current chromatography. Phytochem. Anal. 26(1), 47–53 (2015).

    Article  PubMed  CAS  Google Scholar 

  5. 5.

    Brown, P. D. & Morra, M. J. Control of Soil-Borne Plant Pests Using Glucosinolate-Containing Plants. Adv. Agronomy 61, 167–231 (1997).

    Article  CAS  Google Scholar 

  6. 6.

    Rosa, E. A. S., Heaney, R. K., Fenwick, G. R. & Portas, C. A. M. Glucosinolates in crop plants. Hortic. Rev. 19, 19–215 (1997).

    Google Scholar 

  7. 7.

    Halkier, B. A. & Gershenzon, J. Biology and biochemistry of glucosinolates. Ann. Rev. Plant Biol. 57, 303–333 (2006).

    Article  CAS  Google Scholar 

  8. 8.

    Zhang, Y., Talalay, P., Cho, C. G. & Posner, G. H. A major inducer of anticarcinogenic protective enzymes from broccoli: isolation and elucidation of structure. Proc. Nat. Acad. Sci. USA 89(6), 2399–2403 (1992).

    ADS  Article  PubMed  PubMed Central  CAS  Google Scholar 

  9. 9.

    Fahey, J. W. & Stephenson, K. K. Cancer chemoprotective effects of cruciferous vegetables. Hortscience 34(7), 1159–1163 (1999).

    CAS  Google Scholar 

  10. 10.

    Pereira, V. F. M. et al. Influence of temperature and ontogeny on the levels of glucosinolates in broccoli (Brassica oleracea var. italica) sprouts and their effect on the induction of mammalian phase 2 enzymes. J. Agric. Food Chem. 50(21), 6239–6244 (2002).

    Article  PubMed  CAS  Google Scholar 

  11. 11.

    Dinkova-Kostova, A. T. & Kostov, R. V. Glucosinolates and isothiocyanates in health and disease. Trends Mol. Med. 18(6), 337–347 (2012).

    Article  PubMed  CAS  Google Scholar 

  12. 12.

    Dornberger, K. et al. Investigations of the isothiocyanates erysolin and sulforaphan of Cardaria draba L. Die Pharmazie 30(12), 792–796 (1975).

    PubMed  CAS  Google Scholar 

  13. 13.

    Drobinca, L. et al. Antifungal activity of isothiocyanates and related compounds. I. Naturally occurring isothiocyanates and their analogues. Appl. Microbiol. 15, 701–709 (1967).

    Google Scholar 

  14. 14.

    Fahey, J. W. et al. Sulforaphane inhibits extracellular, intracellular, and antibiotic-resistant strains of Helicobacter pylori and prevents benzo[a]pyrene-induced stomach tumors. Proc. Nat. Acad. Sci. USA, 99(11), 7610-7615 (2002).

  15. 15.

    Haristoy, X., Fahey, J. W., Scholtus, I. & Lozniewski, A. Evaluation of the antimicrobial effects of several isothiocyanates on Helicobacter pylori. Planta Med. 71(4), 326–330 (2005).

    Article  PubMed  CAS  Google Scholar 

  16. 16.

    Yanaka, A. et al. Dietary sulforaphane-rich broccoli sprouts reduce colonization and attenuate gastritis in Helicobacter pylori-infected mice and humans. Cancer Prev. Res. 2(4), 353–360 (2009).

    Article  CAS  Google Scholar 

  17. 17.

    Nowicki, D., Rodzik, O., Herman-Antosiewicz, A. & Szalewska-Palasz, A. Isothiocyanates as effective agents against enterohemorrhagic Escherichia coli: Insight to the mode of action. Sci. Rep. 6, 22263, (2016).

    ADS  Article  PubMed  PubMed Central  CAS  Google Scholar 

  18. 18.

    Kaiser, S. J., Mutters, N. T., Blessing, B. & Günther, F. Natural isothiocyanates express antimicrobial activity against developing and mature biofilms of Pseudomonas aeruginosa. Fitoterapia 119, 57–63 (2017).

    Article  PubMed  CAS  Google Scholar 

  19. 19.

    Palada, M. C. Moringa (Moringa oleifera Lam.): A versatile tree crop with horticultural potential in the subtropical United States. Hortscience 31(5), 794–797 (1996).

    Google Scholar 

  20. 20.

    Fahey, J. W. Moringa oleifera: A review of the medical evidence for its nutritional, therapeutic, and prophylactic properties. Trees for. Life J. 1, 5 (2005).

    Google Scholar 

  21. 21.

    Olson, M. E. & Fahey, J. W. Moringa oleifera: A multipurpose tree for the dry tropics. Revista Mexicana De Biodiversidad 82(4), 1071–1082 (2011).

    Google Scholar 

  22. 22.

    Fahey, J. W. Moringa oleifera: A review of the medicinal potential. Acta Hortic. 1158, 209–224 (2017).

    Article  Google Scholar 

  23. 23.

    Meulenbeld, J. The trees called śigru (Moringa sp.), along with a study of the drugs used in errhines (Vol. 1). (Barkhuis, 2009).

  24. 24.

    Dinkova-Kostova, A. T., Fahey, J. W., Kostov, R. V. & Kensler, T. W. KEAP1 and done? Targeting the NRF2 pathway with sulforaphane. Trends Food Sci.Technol. 69(Part B), 257–269 (2017).

    Article  PubMed  CAS  PubMed Central  Google Scholar 

  25. 25.

    Fahey, J. W., Stephenson, K. K., Wade, K. L. & Talalay, P. Urease from Helicobacter pylori is inactivated by sulforaphane and other isothiocyanates. Biochem. Biophys. Res. Comm. 435(1), 1–7 (2013).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  26. 26.

    Galupo, M. et al. Antibacterial activity of glucomoringin bioactivated with myrosinase against two important pathogens affecting the health of long-term patients in hospitals. Molecules 18(11), 14340–14348 (2013).

    Article  CAS  Google Scholar 

  27. 27.

    Galuppo, M. et al. Administration of 4-(α-L-Rhamnosyloxy)-benzyl isothiocyanate delays disease phenotype in SOD1G93A rats: A transgenic model of amyotrophic lateral sclerosis BioMed Res. Internat., ID# 259417, 12 pp. (2015).

  28. 28.

    Galuppo, M. et al. Antiinflammatory activity of glucomoringin isothiocyanate in a mouse model of experimental autoimmune encephalomyelitis. Fitoterapia 95, 160–174 (2014).

    Article  PubMed  CAS  Google Scholar 

  29. 29.

    Zhang, Y., Kensler, T. W., Cho, C. G., Posner, G. H. & Talalay, P. Anticarcinogenic activities of sulforaphane and structurally related synthetic norbornyl isothiocyanates. Proc. Nat. Acad. Sci. USA 91(8), 3147–3150 (1994).

    ADS  Article  PubMed  PubMed Central  CAS  Google Scholar 

  30. 30.

    Stoner, G. D. & Morse, M. A. Isothiocyanates and plant polyphenols as inhibitors of lung and esophageal cancer. Cancer Lett. 114(1–2), 113–119 (1997).

    Article  PubMed  CAS  Google Scholar 

  31. 31.

    Talalay, P. & Fahey, J. W. Phytochemicals from cruciferous plants protect against cancer by modulating carcinogen metabolism. J. Nutr. 131(11 Suppl), 3027S–33S (2001).

    Article  PubMed  CAS  Google Scholar 

  32. 32.

    Hecht, S. S. Chemoprevention by isothiocyanates. J. Cell. Biochem. Supp. 22, 195–209 (1995).

    Article  CAS  Google Scholar 

  33. 33.

    Kensler, T. W. Chemoprevention by inducers of carcinogen detoxication enzymes. Environ. Health Perspect. 105(Suppl 4), 965–970 (1997).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  34. 34.

    Fahey, J. W. & Talalay, P. Antioxidant functions of sulforaphane: A potent inducer of Phase 2 detoxication enzymes. Food Chem. Toxicol. 37(9–10), 973–979 (1999).

    Article  PubMed  CAS  Google Scholar 

  35. 35.

    Singh, K. et al. Sulforaphane treatment of autism spectrum disorder (ASD). Proc. Nat. Acad. Sci. USA 111(43), 15550–15555 (2014).

    ADS  Article  PubMed  PubMed Central  CAS  Google Scholar 

  36. 36.

    Egner, P. A. et al. Rapid and sustainable detoxication of airborne pollutants by broccoli sprout beverage: results of a randomized clinical trial in China. Cancer Prev. Res. 7(8), 813–823 (2014).

    Article  CAS  Google Scholar 

  37. 37.

    Fahey, J. W., Talalay, P. & Kensler, T. W. Notes from the field: “Green” chemoprevention as frugal medicine. Cancer Prev. Res. 5(2), 179–188 (2012).

    Article  Google Scholar 

  38. 38.

    Fahey, J. W. & Kensler, T. W. Health span extension through green chemoprevention. Virtual Mentor 15(4), 311–318 (2013).

    Article  PubMed  Google Scholar 

  39. 39.

    Kensler, K. H. et al. Genetic or pharmacologic activation of Nrf2 signaling fails to protect against aflatoxin genotoxicity in hypersensitive GSTA3 knockout mice. Toxicol. Sci. 139(2), 293–300 (2014).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  40. 40.

    Riedl, M. A. & Nel, A. E. Importance of oxidative stress in the pathogenesis and treatment of asthma. Curr. Opin. Allergy Clin. Immunol. 8(1), 49–56 (2008).

    Article  PubMed  CAS  Google Scholar 

  41. 41.

    Heber, D. et al. Sulforaphane-rich broccoli sprout extract attenuates nasal allergic response to diesel exhaust particles. Food Funct. 5(1), 35–41 (2014).

    Article  PubMed  CAS  Google Scholar 

  42. 42.

    Axelsson, A. S. et al. Sulforaphane reduces hepatic glucose production and improves glucose control in patients with type 2 diabetes. Sci.Trans. Med. 9(394), (2017).

  43. 43.

    Waterman, C. et al. Stable, water extractable isothiocyanates from Moringa oleifera leaves attenuate inflammation in vitro. Phytochemistry 103, 114–122 (2014).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  44. 44.

    Waterman, C. et al. Isothiocyanate-rich Moringa oleifera extract reduces weight gain, insulin resistance, and hepatic gluconeogenesis in mice. Mol. Nutr. Food Res. 59(6), 1013–1024 (2015).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  45. 45.

    Gao, X., Dinkova-Kostova, A. T. & Talalay, P. Powerful and prolonged protection of human retinal pigment epithelial cells, keratinocytes, and mouse leukemia cells against oxidative damage: The indirect antioxidant effects of sulforaphane. Proc. Nat. Acad. Sci. USA 98(26), 15221–15226 (2001).

    ADS  Article  PubMed  PubMed Central  CAS  Google Scholar 

  46. 46.

    Gao, X. & Talalay, P. Induction of phase 2 genes by sulforaphane protects, retinal pigment epithelial cells against photooxidative damage. Proc. Nat. Acad. Sci. USA 101, 10446–10451 (2004).

    ADS  Article  PubMed  PubMed Central  CAS  Google Scholar 

  47. 47.

    Fahey, J. W., Dinkova-Kostova, A. T., Stephenson, K. K. & Talalay, P. The “Prochaska” microtiter plate bioassay for inducers of NQO1. Meth. Enzymol. 382, 243–258 (2004).

    Article  PubMed  CAS  Google Scholar 

  48. 48.

    Dinkova-Kostova, A. T. et al. Phenolic Michael reaction acceptors: Combined direct and indirect antioxidant defenses against electrophiles and oxidants. Med. Chem. 3, 261–268 (2007).

    Article  PubMed  CAS  Google Scholar 

  49. 49.

    Dinkova-Kostova, A. T. & Talalay, P. Direct and indirect antioxidant properties of inducers of cytoprotective proteins. Mol. Nutr. Food Res. 52, S128–S138 (2008).

    PubMed  Google Scholar 

  50. 50.

    Joko, S. et al. Comparison of chemical structures and cytoprotection abilities between direct and indirect antioxidants. J. Funct. Foods 35, 245–255 (2017).

    Article  CAS  Google Scholar 

  51. 51.

    Olsen, O. & Sorensen, H. Isolation of glucosinolates and the identification of o-(alpha-l-rhamnopyranosyloxy)benzylglucosinolate from Reseda odorata. Phytochemistry 18, 1547–1552 (1979).

    Article  CAS  Google Scholar 

  52. 52.

    Sorensen, H. o-(alpha-L-Rhamnopyranosyloxy)benzylamine and o-hydroxybenzylamine in Reseda odorata. Phytochemistry 9, 865–870 (1970).

    Article  CAS  Google Scholar 

  53. 53.

    De Graaf, R. M. et al. Isolation and identification of 4-a-rhamnosyloxy benzyl glucosinolate in Noccaea caerulescens showing intraspecific variation. Phytochemistry 110, 166–171 (2015).

    Article  PubMed  CAS  Google Scholar 

  54. 54.

    Maldini, M. et al. Moringa oleifera: study of phenolics and glucosinolates by mass spectrometry. J. Mass Spectrom. 49(9), 900–910 (2014).

    ADS  Article  PubMed  CAS  Google Scholar 

  55. 55.

    Olson, M. E. Intergeneric relationships within the Caricaceae-Moringaceae clade (Brassicales) and potential morphological synapomorphies of the clade and its families. Internat. J. Plant Sci. 163(1), 51–65 (2002).

    Article  CAS  Google Scholar 

  56. 56.

    Rangkadilok, N. et al. Developmental changes of sinigrin and glucoraphanin in three Brassica species (Brassica nigra, Brassica juncea and Brassica oleracea var. italica). Sci. Hortic. 96(1–4), 11–26 (2002).

    Article  CAS  Google Scholar 

  57. 57.

    Rangkadilok, N. et al. The effect of post-harvest and packaging treatments on glucoraphanin concentration in broccoli (Brassica oleracea var. italica). J. Agric. Food Chem. 50(25), 7386–7391 (2002).

    Article  PubMed  CAS  Google Scholar 

  58. 58.

    Prestera, T. et al. Comprehensive chromatographic and spectroscopic methods for the separation and identification of intact glucosinolates. Anal. Biochem. 239(2), 168–179 (1996).

    Article  PubMed  CAS  Google Scholar 

  59. 59.

    Daxenbichler, M. E. et al. Glucosinolate composition of seeds from 297 species of wild plants. Phytochemistry 30(8), 2623–2638 (1991).

    Article  CAS  Google Scholar 

  60. 60.

    Kjaer, A., Malver, O., El-Menshaw, B. & Reisch, J. Isothiocyanates in myrosinase-treated seed extracts of Moringa peregrina. Phytochemistry 18, 1485–1487 (1979).

    Article  CAS  Google Scholar 

  61. 61.

    Chodur, G. M., Olson, M. E., Wade, K. L., Stephenson, K. K. et al. Wild type and domesticated Moringa oleifera differ markedly in taste, glucosinolate composition, and antioxidant potential, but not myrosinase activity or protein content. Sci. Rep. 8,(2018).

  62. 62.

    Fahey, J. W., Zhang, Y. & Talalay, P. Broccoli sprouts: an exceptionally rich source of inducers of enzymes that protect against chemical carcinogens. Proc. Nat. Acad. Sci. USA 94(19), 10367–10372 (1997).

    ADS  Article  PubMed  PubMed Central  CAS  Google Scholar 

  63. 63.

    Agerbirk, N. & Olsen, C. E. Glucosinolate structures in evolution. Phytochemistry 77, 16–45 (2012).

    Article  PubMed  CAS  Google Scholar 

  64. 64.

    Doerr, B., Wade, K. L., Stephenson, K. K., Reed, S. B. & Fahey, J. W. Cultivar effect on Moringa oleifera glucosinolate content and taste: A pilot study. Ecol. Food Nutr. 48(3), 199–211 (2009).

    Article  PubMed  Google Scholar 

  65. 65.

    Badgett, B. L. Part I. The mustard oil glucoside from Moringa oleifera seed. PhD Thesis – (Rice University: Houston, TX., 1964).

  66. 66.

    Bennett, R. N. et al. Profiling glucosinolates and phenolics in vegetative and reproductive tissues of the multi-purpose trees Moringa oleifera L. (horseradish tree) and Moringa stenopetala L. J. Agric. Food Chem. 51(12), 3546–3553 (2003).

    Article  PubMed  CAS  Google Scholar 

  67. 67.

    Eilert, U., Ehmke, A. & Wolters, B. Elicitor-induced accumulation of acridone alkaloid epoxides in Ruta graveolens suspension cultures. Planta Med. 50(6), 508–512 (1984).

    Article  PubMed  CAS  Google Scholar 

  68. 68.

    Park, E. J., Cheenpracha, S., Chang, L. C., Kondratyuk, T. P. & Pezzuto, J. M. Inhibition of lipopolysaccharide-induced cyclooxygenase-2 and inducible nitric oxide synthase expression by’ 4-[(2′-O-acetyl-alpha-L-rhamnosyloxy) benzyl]isothiocyanate from Moringa oleifera. Nutr. Cancer 63(6), 971–982 (2011).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  69. 69.

    Cheenpracha, S. et al. Potential anti-inflammatory phenolic glycosides from the medicinal plant Moringa oleifera fruits. Bioorg. Med. Chem. 18(17), 6598–6602 (2010).

    Article  PubMed  CAS  Google Scholar 

  70. 70.

    Amaglo, N. et al. Profiling selected phytochemicals and nutrients in different tissues of the multipurpose tree Moringa oleifera L., grown in Ghana. Food Chem. 122, 1047–1054 (2010).

    Article  CAS  Google Scholar 

  71. 71.

    Farnham, M. W., Wilson, P. E., Stephenson, K. K. & Fahey, J. W. Genetic and environmental effects on glucosinolate content and chemoprotective potency of broccoli. Plant Breeding 123, 60–65 (2004).

    Article  CAS  Google Scholar 

  72. 72.

    Farnham, M. W., Stephenson, K. K. & Fahey, J. W. Glucoraphanin level in broccoli seed is largely determined by genotype. HortScience 40(1), 50–53 (2005).

    CAS  Google Scholar 

  73. 73.

    Farnham, M. W., Stephenson, K. K. & Fahey, J. W. The capacity of broccoli to induce a mammalian chemoprotective enzyme varies among inbred lines. J. Amer. Soc. Hort. Sci. 125(4), 482–488 (2000).

    CAS  Google Scholar 

  74. 74.

    Tsuruo, I., Yoshida, M. & Hata, T. Studies on the myrosinase in mustard seed. Agric. Biol. Chem. 31, 18–32 (1967).

    CAS  Google Scholar 

  75. 75.

    Ludikhuyze, L., Ooms, V., Weemaes, C. & Hendrickx, M. Kinetic study of the irreversible thermal and pressure inactivation of myrosinase from broccoli (Brassica oleracea L. Cv. Italica). J. Agric. Food Chem. 47, 1794–1800 (1999).

    Article  PubMed  CAS  Google Scholar 

  76. 76.

    Angelino, D. & Jeffery, E. Glucosinolate hydrolysis and bioavailability of resulting isothiocyanates: Focus on glucoraphanin. J. Funct. Foods 7, 67–76 (2014).

    Article  CAS  Google Scholar 

  77. 77.

    Fahey, J. W. et al. Sulforaphane bioavailability from glucoraphanin-rich broccoli: Control by active endogenous myrosinase. PLOS ONE., (2015).

  78. 78.

    Stephenson, K. K. & Fahey, J. W. Development of tissue culture methods for the rescue and propagation of endangered Moringa spp. germplasm. Econ. Bot. 58, s116–s124 (2004).

    Article  Google Scholar 

  79. 79.

    Thulin, M. Flora of Somalia, Volume 1, 1993, ISBN: 9780947643553.

  80. 80.

    Chase, M. W. & Hillis, H. H. Silica gel: an ideal material for field preservation of leaf samples for DNA studies. Taxon 40, 215–220 (1991).

    Article  Google Scholar 

  81. 81.

    Troyer, J. K., Stephenson, K. K. & Fahey, J. W. Analysis of glucosinolates from broccoli and other cruciferous vegetables by hydrophilic interaction liquid chromatography. J. Chromatogr. A. 919(2), 299–304 (2001).

    Article  PubMed  CAS  Google Scholar 

  82. 82.

    Wade, K. L., Garrard, I. J. & Fahey, J. W. Improved hydrophilic interaction chromatography method for the identification and quantification of glucosinolates. J. Chromatogr. A. 1154(1–2), 469–472 (2007).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  83. 83.

    Fahey, J. W., Wade, K. L., Stephenson, K. K. & Chou, F. E. Separation and purification of glucosinolates from crude plant homogenates by high-speed counter-current chromatography. J. Chromatogr. A. 996(1–2), 85–93 (2003).

    Article  PubMed  CAS  Google Scholar 

  84. 84.

    Egner, P. A. et al. Bioavailability of sulforaphane from two broccoli sprout beverages: Results of a short-term, cross-over clinical trial in Qidong, China. Cancer Prev. Res. 4(3), 384–395 (2011).

    Article  CAS  Google Scholar 

  85. 85.

    Prochaska, H. J., Santamaria, A. B. & Talalay, P. Rapid detection of inducers of enzymes that protect against carcinogens. Proc. Nat. Acad. Sci. USA 89(6), 2394–2398 (1992).

    ADS  Article  PubMed  PubMed Central  CAS  Google Scholar 

  86. 86.

    Liu, H., Dinkova-Kostova, A. T. & Talalay, P. Coordinate regulation of enzyme markers for inflammation and for protection against oxidants and electrophiles. Proc. Nat. Acad. Sci. USA 105(41), 15926–15931 (2008).

    ADS  Article  PubMed  PubMed Central  Google Scholar 

  87. 87.

    Kwak, M.-Y., Kensler, T. W. & Casero, R. A. Jr. Induction of phase 2 enzymes by serum oxidized polyamines through activation of Nrf2: Effect of the polyamine metabolite acrolein. Biochem. Biophys. Res. Commun. 305(3), 662–670 (2003).

    Article  PubMed  CAS  Google Scholar 

  88. 88.

    Zhang, Y. Role of glutathione in the accumulation of anticarcinogenic isothiocyanates and their glutathione conjugates by murine hepatoma cells. Carcinogenesis 21(6), 1175–1182 (2000).

    Article  PubMed  CAS  Google Scholar 

  89. 89.

    R Development Core Team, R: A language and environment for statistical computing. V.3.3.1. (2016).

Download references


We gratefully acknowledge financial assistance and support from the Lewis B. and Dorothy Cullman Foundation, The Missouri Botanical Garden, Trees for Life International, Programa para el Apoyo para la Investigación e Innovación Tecnológica, UNAM, Grant IT200515, and CONACyT Grant 1320404. Paul Talalay encouraged and supported early phases of this work when other financial support was lacking. This project has extended over a number of years, during which many people have lent their intellectual and participatory support to this project, both in the laboratory and in the field, including: Peter Raven, Albena Dinkova-Kostova, W. David Holtzclaw, Cungen Zang, Mark Rogers, Camilo Rojas, Amy Zalcmann, Julieta Rosell, Helen Giles, Belinda Chiu, YuZhu Shi, Hua Liu, Sharadha Dayalan Naidu, Thomas W. Kensler, Balbir Mathur, and Jeffrey Faus. Further, we are most grateful to the International Moringa Germplasm Collection, which houses living material of 12 of the 13 Moringa species, as a resource for scientific research on the basic biology of Moringa and investigation of applied uses such as nutrition, chronic disease prevention and treatment, biofuels, and water clarification. Progeny of some of the materials originally collected and evaluated in this paper are now growing at The Collection. It was established by- and is managed by MEO, with intellectual participation from JWF, KKS, KLW, and their students. Its permanent location is now on the coast of Jalisco, Mexico thanks to a grant from Trees for Life International, and ongoing support from the Instituto de Biología, Universidad Nacional Autónoma de México (UNAM) where MEO is a professor. Trees for Life International has provided the germplasm collection with a research mandate to find the most protein-rich moringa, and to develop a better understanding of moringa protein attributes such as digestibility and amino acid profile.

Author information




J.W.F. and M.E.O. wrote the paper; M.E.O., D.O., and W.N. collected and processed material from the field; M.M. performed and interpreted nuclear magnetic resonance spectroscopy; J.A., P.A.E. and W.C.H. performed mass spectroscopy; J.W.F., K.K.S., G.C. and K.L.W. performed all other laboratory and analytical work. All authors edited and approved the final manuscript.

Corresponding author

Correspondence to Jed W. Fahey.

Ethics declarations

Competing Interests

The authors have no conflicts of interest to disclose with the exception that JWF is a member of the scientific advisory board of Kuli Kuli, (a social benefit corporation that produces and sells Moringa products).

Additional information

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Fahey, J.W., Olson, M.E., Stephenson, K.K. et al. The Diversity of Chemoprotective Glucosinolates in Moringaceae (Moringa spp.). Sci Rep 8, 7994 (2018).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI:

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing