Keywords

1 Introduction

Intrigued with previous reports of growth regulating properties of pollen extracts, Mitchell and Whitehead (1941) examined the growth responses and histological changes that resulted from the application of ethereal extracts of corn pollen on intact bean plants or on the cut surfaces of decapitated stems. They observed that the first internode of the plants where these extracts were applied grew significantly more and faster than the untreated ones or treated with some known auxins, as well as gained more fresh and dry weights than the controls. They demonstrated that these were light dependent phenomena, and due to cell elongation rather than cell division. When applied to tap roots, these extracts inhibited root elongation and provoked the appearance of small tumors distal to the application point. When these pollen extracts where applied to the cut surfaces of decapitated stems they caused pronounced radial elongation of epidermal, cortical parenchyma, and endothelial cells. Later Mitchell et al., reported that immature bean seeds also contained plant growth-stimulating hormones (Mitchell et al.1951) and that Brassica napus pollen contained new, yet unknown, hormones they called brassins (Mitchell et al.1970), all of them with properties similar to those reported earlier (Mitchell and Whitehead 1941).

About 60 kinds of pollen were then screened for plant growth activity in the bean second internode assay, and “a few samples, notably the pollen from rape plant (Brassica napus L.) and alder tree (Alnus glutinosa L.), produced an unusual response that combined elongation (the typical gibberellin response) with swelling and curvature” (Mandava 1988). At the same time, some experiments showed that application of brassins to young bean and Siberian elm tree plants promoted overall plant growth (Mitchell and Gregory 1972), what led United States Department of Agriculture (USDA) to initiate an effort aimed to explore the agricultural perspectives of brassins and to isolate their component(s). After processing 500 libers of rape pollen, finally, the USDA team announced the isolation and structure elucidation of the active principle, brassinolide (1) (Grove et al.1979), the first plant hormone of steroidal nature, presenting, unlike animal steroidal hormones, (i) a 22α,23α-dihydroxylated campestane side chain, (ii) a B-ring lactone, and, (iii) a 2α,3α-dihydroxylated ring A. Bean second internodes exhibited elongation, curvature, swelling and even splitting when treated with increasing amounts of brassinolide (1) (Grove et al.1979) (Fig. 1.1), a very distinct effect never observed with any other known plant hormone. Its isolation was followed by its partial synthesis (Fung and Siddall 1980; Ishiguro et al.1980) and of its analogues (Thompson et al.1979, 1981, 1982; Mori 1980; Takatsuto et al.1981; Sakakibara and Mori 1982; Sakakibara et al.1982; Mori et al.1982), some later recognized as plant hormones themselves.

Fig. 1.1
figure 1figure 1figure 1

Natural brassinosteroids

The early synthetic work furnished many compounds with similar or weaker brassin activity, what prompted natural products chemists to search for brassinolide related compounds in plant species other than rape. To the first of them, the 6-ketosteroid castasterone (2) (Yokota et al.1982a), the putative biosynthetic precursor of brassinolide (1), followed that of dolicholide (3) (Yokota et al.1982b), dolichosterone (4) (Baba et al.1983), both with a 24-methylene-5α-cholestane structure, and 28-homodolichosterone (11) (Baba et al.1983), with a 24(E)-ethylidene-5α-cholestane skeleton instead of a 5α-campestane basis as in brassinolide (1) and castasterone (2), and then a multitude of brassinosteroids (BRs) of different side chain structures and oxygenation patterns were isolated, giving rise to the class of brassinosteroids phytohormones, the components of which will be described ahead.

2 Natural Brassinosteroids

About sixty compounds with structures related to that of brassinolide (1) were isolated from or detected in plant materials in the last forty years (see Table 1.1 and Fig. 1.1). They were found in 26 species of 6 families of Algae, in 2 species of 2 families of Bryophyte, in 15 species of 8 families of Pteridophyte, in 6 species of 4 families of Gymnospermae, in 74 species of 35 families of Angiospermae (in 18 species of 6 families of Monocotyledoneae and 56 species of Dicotyledoneae), and in some plant derived products. About 15 biosynthetic precursors of brassinosteroids, some presenting brassinosteroid activity themselves, were found in many plant species.

Table 1.1 First report of a natural brassinosteroid

The finding that the rice lamina inclination assay, developed by Maeda (1965), to test for auxin activity could be used to detect the activity of brassinosteroids at even nanomolar or subnanomolar concentrations (Wada et al.1981) and the development of a microanalytical method for the quantification of 22α, 23α-dihydroxybrassinosteroids (Takatsuto et al.1982), allowed a rapid expansion of the number of known brassinosteroids. The first brassinosteroids isolated presented, as common features, (i) a 5α-cholestane or a 6,7-seco-5α-cholestane derived skeleton, (ii) ring A with one to three oxygen functions (one always at carbon 3), (iii) ring B fully saturated or with varying degree of oxidation at carbon 6, (iv) all-trans ring junctions and (v) 22α,23α-dihydroxylation. In this sense, 3-oxygenated (20β)-5α-cholestane-22α,23α-diols of plant origin, bearing additional alkyl or oxy substituents, were considered as natural brassinosteroids (Zullo and Adam 2002). A more restricted definition states that, in the biosynthetic route to a brassinosteroid lactone, “one would consider as brassinosteroids only those compounds originated after the 22α,23α-dihydroxylation (i.e., those between teasterone or 6-deoxoteasterone and brassinolide), and hence as brassinosteroid precursors those before dihydroxylation occurs (i.e., those compounds up to cathasterone and 6-deoxocathasterone)” (Zullo et al.2003; Zullo and Kohout 2004). After then some other brassinosteroids presenting 2,3-epoxy, 23-dehydro, 23-glycosidic, 23-ester functions, or 26-nor side chain or even 2,3-unsaturation were isolated, “allowing to consider as natural brassinosteroids the 3-oxygenated (20β)-5α-cholestane-22α,23α-diols or their derived compounds isolated from plants, bearing additional alkyl or oxy substituents” (Zullo 2018).

The unconjugated brassinosteroids so far isolated present 27 (C27), 28 (C28) or 29 (C29) carbons, with 5α-cholestane or 26-nor-5α-campestanes (= 26-nor-24α-methyl-5α-cholestane) structures for the C27 series, 5α-campestane (= 24α-methyl-5α-cholestane), 5α-ergostane (= 24β-methyl-5α-cholestane) or 24-methylene-5α-cholestane skeletons for the C28 series, and 5α-sitostane (= 24α-ethyl-5α-cholestane), 24(Z)-ethylidene-5α-cholestane, 25-methyl-5α-campestane and 24-methylene-25-methyl-5α-cholestane structures for the C29 series (Fig. 1.2). Only one of the side chains of isolated brassinosteroids is of a 26-nor sterol, although C26-demethylation of brassinosteroids have been demonstrated in metabolic studies with some species (Joo et al.2012, 2015; Kim et al.2000a, b). From the 12 different side chains of natural brassinosteroids, 9 of them present 22α,23α-dihydroxylation, while one presents a 22α-hydroxy-23-oxo group, another one presents conjugation of one glucose unit at the 23α-hydroxyl, and a last different side chain shows phosphorylation at the 23α-hydroxyl. Feeding studies shows that side chain glucosylation can occur at either C-23 (Poppenberger et al.2005) or C-22 (Soeno et al.2006), and also at C-25 or C-26 after hydroxylation at these carbons (Hai et al.1996). Phosphorylation (Kim et al.2015) and sulfonation (Rouleau et al.1999) have been demonstrated to occur at the side chain of brassinosteroids, but while the first occurs at C-23, the second occurs at C-22, at least with the actual experimental data available.

Fig. 1.2
figure 2

Brassinolide and structural variations of brassinosteroids

It is known that the bioactivity of brassinosteroids is dependent on the structure of the side chain and of the A/B rings (Takatsuto et al.1983b; Takatsuto et al.1983a; Brosa et al.1996; Takatsuto et al.1987; Mandava 1988; Liu et al.2017; Zullo and Adam 2002). Regarding to the side chain, as general rules, employing the rice lamina inclination assay on any of its versions (Maeda 1965; Wada et al.1981; Fujioka et al.1998a), for the same A/B ring structures, 22α,23α-dihidroxybrassinosteroids of the brassinolide series are so active as of the 28-homobrassinosteroids series (Takatsuto et al.1983a), and more active than those of 24-epi- or 28-norbrassinosteroids (Takatsuto et al.1983a; Wada et al.1983), which are more active than 26-norbrasssinosteroids (Kim et al.2000a; Watanabe et al.2001). 23-Dehydrogenation (Watanabe et al.2001), or conjugation at one of the side chain hydroxyls (Suzuki et al.1993b; Kim et al.2015; Rouleau et al.1999), diminishes (Yokota et al.1998; Suzuki et al.1993b) or abolishes the biological activity (Kim et al.2015; Rouleau et al.1999), an effect contrary to that observed with 25-methylation (Mori and Takeuchi 1988). It is to note that the relative biological activity of brassinosteroids vary according to the biological assay performed for their evaluation, not only in relation to the side chain but also to the other active sites of their molecules (Takatsuto et al.1983b; Watanabe et al.2001; Zullo and Adam 2002; Liu et al.2017).

A greater structural variation is observed in ring A, with 15 different structures reported, ranging from Δ2,3-unsaturated to trioxygenated and conjugated brassinosteroids: even so, this variation still does not reflect all the possible substructures at this ring, presumed either by efforts of large scale isolation of brassinosteroids (Kim 1991; Fujioka 1999), or by the study of the metabolism of brassinosteroids (Zullo 2018). The biological activity for brassinosteroids with A/B trans ring junctions increases as substitution in ring A changes in the order 3β-hydroxy < 3-oxo < 3α-hydroxy < 2α,3α-dihydroxy, and diminishes as deviates from these patterns (Mandava 1988; Zullo and Adam 2002; Liu et al.2017; Takatsuto et al.1987; Fujioka et al.1995a).

The structural variations in ring B reflect the main steps in the biosynthesis of brassinosteroids (Vriet et al.2013), being more active as its oxidation state increases (Mandava 1988) sequentially from the 6-deoxo to the 6α-hydroxy to the 6-oxo and to the 7-oxalactone types. Therefore, brassinosteroids can be classified, according to the B ring structure, as: (a) 6-oxo-7-oxalactonic brassinosteroids: (i) 2α,3α-dihydroxylated: brassinolide (1), dolicholide (3), 28-homodolicholide (10), 28-norbrassinolide (14), 28-homobrassinolide (17), 24-epibrassinolide (27), cryptolide (54); (ii) 2α, 3β-dihydroxylated: 3-epibrassinolide (51); (iii) 3α-hydroxylated: 2-deoxybrassinolide (7-oxatyphasterol, 43); (iv) 3β-hydroxylated: 3-epi-2-deoxybrassinolide (7-oxateasterone, 58); (b) 6-oxo (or 6-keto) brassinosteroids: (i) 2α,3α-dihydroxylated: castasterone (2), dolichosterone (4), 24-epicastasterone (9), 28-homodolichosterone (11), 28-homocastasterone (12), 28-norcastasterone (15), 25-methyldolichosterone (16), 25-methylcastasterone (33), 26-norcastasterone (59); (ii) 2β,3α-dihydroxylated: 2-epicastasterone (20), 2-epi-25-methyldolichosterone (24), 23-dehydro-2-epicastasterone (55); (iii) 2α,3β-dihydroxylated: 3-epicastasterone (21), 3,24-diepicastasterone (23); (iv) 2β,3β-dihydroxylated: 2,3-diepicastasterone (22), 2,3-diepi-25-methyldolichosterone (25); (v) 3α-monohydroxylated: typhasterol (7), 2-deoxy-25-methyldolichosterone (18), 28-homotyphasterol (37), 28-nortyphasterol (49); (vi) 3β-monohydroxylated: teasterone (8), 3-epi-2-deoxy-25-methyldolichosterone (19), 28-homoteasterone (34), 28-norteasterone (62); (vii) 1β,2α,3α-trihydroxylated: 1β-hydroxycastasterone (28); (viii) 1α,2α,3β-dihydroxylated: 1α-hydroxy-3-epicastasterone (29); (ix) 2α,3α-epoxide: 2,3-diepisecasterone (52); (x) 2β,3β-epoxide: secasterone (38), 24-episecasterone (46); (xi) Δ2-olefin: secasterol (53); (xii) 3β-conjugates: teasterone-3-myristate (35), teasterone-3-laurate (44), 3-O-β-D-glucopyranosylteasterone (48); (xiii) 23α-conjugates: 23-O-β-D-glucopyranosyl-25-methyldolichosterone (26), 23-O-β-D-glucopyranosyl-2-epi-25-methyldolichosterone (32), castasterone 23-phosphate (60); (xiv) 3-dehydro: 3-dehydroteasterone (36); (c) 6α-hydroxybrassinosteroids: 6α-hydroxycastasterone (47); (d) 6-deoxobrassinosteroids: (i) 2α,3α-dihydroxylated: 6-deoxocastasterone (5), 6-deoxodolichosterone (6), 6-deoxo-28-homodolichosterone (13), 6-deoxo-25-methyldolichosterone (31), 6-deoxo-28-norcastasterone (41), 6-deoxo-24-epicastasterone (42); (ii) 2α,3β-dihydroxylated: 3-epi-6-deoxocastasterone (30); (iii) 3α-monohydroxylated: 6-deoxotyphasterol (39), 6-deoxo-28-nortyphasterol (50), 6-deoxo-28-homotyphasterol (61); (iv): 3β-monohydroxylated: 6-deoxoteasterone (45), 6-deoxo-28-norteasterone (56); (v) 3-dehydro: 3-dehydro-6-deoxoteasterone (40), 3-dehydro-6-deoxo-28-norteasterone (57).

3 Brassinosteroids Precursors

A series of papers revealed the main steps of brassinosteroids biosynthesis, from the plant sterols to the brassinosteroid lactones, especially that from campesterol (CR) or campestanol (CN) to brassinolide (1). From these studies it became clear that, if the natural brassinosteroids can be easily recognized from their chemical structures, similar observation does not happen with their precursors (see Fig. 1.3 and Table 1.2). The first experiments established the biosynthesis of brassinolide (1) from teasterone (8) via, sequentially, 3-dehydroteasterone (36), typhasterol (7), and castasterone (2) (Suzuki et al.1993a, 1994a, c) (follow by Fig. 1.4). Soon after it was found that campesterol (CR) was converted to campestanol (CN) and to 6α-hydroxycampestanol (63), 6-oxocampestanol (64), 22α-hydroxy-6-oxocampestanol (65), named cathasterone, and this one to teasterone (8) (Fujioka et al.1995b). The complete biosynthetic sequence of brassinolide starting from campesterol (CR) via cathasterone (65) is known as the early C-6 oxidation pathway (a route in which C-6 oxidation occurs earlier than 22α,23α-dihydroxylation).

Fig. 1.3
figure 3

Brassinosteroids precursors

Table 1.2 Brassinosteroid precursors
Fig. 1.4
figure 4figure 4

Biosynthesis of brassinolide (1) from campesterol (red: sterols; green: brassinosteroids precursors; blue: brassinosteroids). Adapted from Zullo 2018

The frequent isolation or detection of 6-deoxobrassinosteroids brought the suspicion that another biosynthetic route to brassinosteroid lactones could exist. Feeding experiments with labeled precursors established the sequence 6-deoxoteasterone (45), 3-dehydro-6-deoxoteasterone (40), 6-deoxotyphasterol (39), 6-deoxocastasterone (5), castasterone (2), brassinolide (1), which was called the late C-6 oxidation pathway (a route in which C-6 oxidation occurs later than 22α, 23α-dihydroxylation) (Choi et al.1997). It was further demonstrated the conversion of campestanol (CN) to 6-deoxoteasterone (45) through 6-deoxocathasterone (66) (Bishop et al.1999), and the presence of 3-epi-6-deoxocathasterone (67), a putative brassinosteroid precursor, in cultured cells of Catharantus roseus (Fujioka et al.2000b).

A thorough examination of the sterols present in cultured cells of C. roseus and in Arabidopsis seedlings, conjugated with metabolic studies with deuterated substrates, revealed that the conversion of campesterol (CR) to campestanol (CN) occurs through campest-4-en-3-one (4en3one) and campestan-3-one (3one) (Fujioka et al.2002). Moreover, it revealed the operation of intermediates in the conversion of campesterol (CR) to 6-deoxocathasterone (66), originating 22α-hydroxycampesterol (68), 22α-hydroxycampest-4-en-3-one (69), and 22α-hydroxy-5α-campestan-3-one (70) from, respectively, campesterol (CR), campest-4-en-3-one (4en3one) and campestan-3-one (3one). In the same extracts were found also the 28-norhomologues 22α-hydroxycholesterol (71), 22α-hydroxycholest-4-en-3-one (72), 22α-hydroxy-5α-cholestan-3-one (73), 6-deoxo-28-norcathasterone (74) and 3-epi-6-deoxo-28-norcathasterone (75). Later, studying the action of Arabidopsis CYP90C1 and CYP90D1, it was found that these enzymes act on 3-epi-6-deoxocathasterone (67), 22α-hydroxycampesterol (68), 22α-hydroxy-5α-campestan-3-one (70), and 22α-hydroxycampest-4-en-3-one (69) to yield, respectively, 6-deoxotyphasterol (39), 22α, 23α-dihydroxycampesterol (76), 3-dehydro-6-deoxoteasterone (40), and 22α, 23α-dihydroxycampest-4-en-3-one (77), revealing a new shortcut in the biosynthesis of brassinosteroids. Compounds 63-77, isolated from plant material, present side chains with no oxygen function or 22α-monohydroxylated or 22α,23α-dihydroxylated and rings A/B typical of common plant sterols (as 3β-hydroxy-Δ5-sterols or 3β-hydroxy-5α-stanols) or less usual ones [like Δ4-sten-3-ones (Franke et al.2004; Georges et al.2006; Pinto et al.2002) or 5α-stan-3-ones (Guillen and Manzanos 2001)] or reflecting the steps for the construction of typical A/B rings of brassinosteroids (5α-stan-3β,6α-diols, 5α-stan-3β-ol-6-one, 5α-stan-3α-ol) (Fig. 1.5). None of these fragments, per se, can be attributed exclusively to brassinosteroids (Zullo 2018).

Fig. 1.5
figure 5

Fragments found in brassinosteroids precursors

It is to note that only brassinosteroids precursors of campestane and cholestane skeletons had been isolated to date, what does not exclude the possibility of similar biosynthetic reactions can occur at the remaining skeletons (ergostane, sitostane, 24-methylenecholestane, 24-ethylydenecholestane, 25-methylcampestane and 24-methylene-25-methylcholestane), for all the possible sequences in the grid (as shown in Fig. 1.4) or through conversions of skeletons while functionalizing them towards the synthesis of castasterone-like or brassinolide-like brassinosteroids, what could explain the isolation or detection of brassinosteroids of different skeletons in the same plant materials. The fact that total sterols usually comprise 2–3 × 10−3 g/g of plant dry weight (Benveniste 2004) and that brassinosteroids are present usually in 10−12–10−9 g/g fresh weight in plant material (Bajguz and Tretyn 2003; Takatsuto 1994), immersed in a matrix of tens of compounds of similar structure (and, hence, of similar polarity and similar chromatographic behavior), turns a very difficult task to determine the brassinosteroids profile of a given plant material, including the compounds of transient existence, like their precursors, can explain why precursors of different skeletons have not been isolated yet.

4 Brassinosteroids with Partially Elucidated Structure

A few natural brassinosteroids were isolated in pure state in enough amount to identify them by the usual spectroscopic methods, but usually they are detected by comparison with authentic compounds prepared by synthesis. Sometimes, due to small amounts of samples, to similar spectroscopic characteristics but different chromatographic behavior, it is not possible to determine the structure of all compounds present in a given brassinosteroids extract. Eventually the complete structure of one of these compounds is correctly elucidated.

One of the richest sources of brassinosteroids, the seeds of kidney beans, presents about 60 compounds of partially known structure (Hwang et al.2006), for which some of them were described (Yokota et al.1987c) (see Fig. 1.6). Among them is cited 1 isomer of 6-deoxo-28-homodolichosterone (78), 4 isomers of castasterone (79), 1 isomer of a hydroxylated castasterone (80), 2 isomers of 28-homocastasterone (81), 3 isomers of a homologue of dolichosterone (82), 1 isomer of a brassinolide derivative with 14 atomic units higher (83), 1 isomer of a brassinolide derivative with 44 atomic units higher (84), 1 isomer of dolicholide (85), 1 isomer of dolicholide with an extra oxygen (86), another one with an extra hydroxyl (87), a dolicholide derivative 28 atomic units higher (88), and another one with a carboxy group (89), an isomer of 28-homobrassinolide (90), an homologue of dolicholide (91) and its carbonyl derivative (92), a carbonyl homologue of dolicholide (93) (Yokota et al.1987c). Two other brassinosteroids were reported in Phaseolus vulgaris, ξ-epi-23-dehydrocastasterone (94) and an homologue with a carbonyl group (95) (Kim 1991). Three isomers of 28-homobrassinolide (90), four isomers of 23-dehydrobrassinolide (96), and one isomer of 28-homodolicholide (97) were reported in pollen and anthers of Cryptomeria japonica (Yokota et al.1998). 25-Methyldolichosterone (16) was later identified as one of the isomers of (82) (Kim et al.1987), as well as cryptolide (54) as one of the four isomers of 23-dehydrobrassinolide (96) (Watanabe et al.2000).

Fig. 1.6
figure 6

Brassinosteroids with partially elucidated structure

5 Occurrence of Brassinosteroids

Brassinosteroids have been isolated from different plant organs such as pollen, anthers, seeds, leaves, stems, roots, flowers, and grain as well as in insect and crown galls. The endogenous level of brassinosteroids varies from plant’s organ and the age of the plant. Pollen and immature seeds are found to have the highest concentration of brassinosteroids, however, young growing tissues contain higher levels of brassinosteroids than mature tissues. The presence of some bioactive brassinosteroids viz., castasterone (2, BR2), brassinolide (1, BR1), 6-deoxocastasterone (5, BR5), teasterone (8, BR8), typhasterol (7, BR7) and 3-dehydro-6-deoxoteasterone (40, BR40) was confirmed in at least 103, 71, 40, 34, 28 and 28 plant species, respectively. Brassinolide (1) and castasterone (2) are widely distributed in algae and flowering plants, but only castasterone (2) was detected in lower non-flowering plants (liverwort, moss, lycophytes and ferns). Their presence in so many species, from the simplest algae to the more complex phanerogams, as well as the increasing detection in many new species indicates their ubiquitous distribution in the plant kingdom, what is expected from their role as plant hormones.

Table 1.3 lists the occurrence of brassinosteroids in plant species and Table 1.4 the occurrence of the established brassinosteroids precursors. It does not discriminate from which organ they were isolated or detected, or the concentration which they were found, so, primary source of information must be retrieved for proper use of their data.

Table 1.3 Occurrence of natural brassinosteroids
Table 1.4 Occurrence of brassinosteroids precursors

Brassinosteroids were also found in plant derived products, as 24-epibrassinolide (27) in biodiesel cakes of Brassica carinata A. Braun or Brassica napus L. (Bardi and Rosso 2015); brassinolide (1), castasterone (2), typhasterol (7), teasterone (8) and 28-homocastasterone (12) in a vermicompost leachate (Aremu et al.2015); and brassinolide (1), castasterone (2), 28-norbrassinolide (14) and 28-norcastasterone (15) in date (Phoenix dactilifera L.), medlar (Eryobotrya japonica Lindl.), milkvetch (Astragalus sp.), rape (Brassica napus L.) and robinia (Robinia pseudo-acacia L.) honeys, and also 28-homobrassinolide (17) in the last four honeys (Wang et al.2017).