Stemodane Diterpenes and Diterpenoids: Isolation, Structure Elucidation, Biogenesis, Biosynthesis, Biological Activity, Biotransformations, Metabolites and Derivatives Biological Activity, Rearrangements

The scientific activity carried out over forty-five years on stemodane diterpenes and diterpenoids structure elucidation, biogenesis, biosynthesis, biological activity and biotransformations was reviewed.

The structure of (+)-3 was deduced by means of chemical work (highlighted below), IR and 1 H-NMR.
The location of the secondary hydroxy group followed the observation that (−)-5 could be converted into an α,β-unsaturated ketone. It followed that the carbonyl should have been located at C(1) or at C (3). No other location of the carbonyl group could have accommodated such unsaturated functions (5c or 5d) within the molecule ( Figure 5). Applying the octant rule to the corresponding saturated ketones 5a and 5b, it was possible to eliminate the location of the carbonyl function at C(1). In fact, for a carbonyl at C(1), the Cotton effect would have been positive, while for a carbonyl at C(3) it would have been negative. The recorded Cotton effect for (−)-5 was negative. The secondary hydroxy group was therefore placed at C(3). The stereochemistry of the HO-C(3) was established β (equatorial) in view of the large geminal H-C(3) coupling constant (1H, dd, J = 11; 5 Hz), indicating that the latter proton was axial ( Figure 4). On these grounds, the (+)-maritimol structure (+)-3 ( Figures 6 and 7) was therefore attributed [2].
The same group then addressed the structure elucidation work of what was supposed to be (+)-stemodinol (4). The investigation on this new diterpenoid (m.p. 182-183 • C, [α] D = +13.8) was carried out by means of chemical work (highlighted below, Scheme 2), IR and 1 H-NMR. The 1 H-NMR spectra revealed the presence of a primary alcohol function that was located at C(18) on the basis of a detailed 1 H-NMR analysis. In order to confirm the stemodane skeleton, (+)-4 was oxidized to 7 ([α] D not reported as in all cases where the rotation sign is not given) and the latter deoxygenated by the Huang-Minlon procedure to (+)-6. This resulting material was "identical in all respects" with (+)-2-deoxystemodinone (6), previously obtained from (+)-2 and (+)-3 (vide supra), thus confirming the stemodane skeleton and the HO-C (13). Unfortunately, the comparison was misleading. Some years later (vide infra), it was, in fact, demonstrated that the structure (+)-4 had been erroneously attributed to (+)-stemodinol and that the true structure of the latter was that of (+)-stemarin (8) (m.p. 183-184 • C, [α] D = +17.8), a structural diastereoisomer with a new and unique skeleton, the isolation from S. maritima and structure elucidation of which had been reported by Manchand and Blount in 1975. The paper [2] by Doorenbos and colleagues was submitted to the J. Pharm. Sci. at the end of March 1975 and was accepted for publication on Feb. 1976. On Aug. 1975, Manchand and colleagues submitted the paper to J. C. S. Chem. Commun. [3], and published in the same year, in which the isolation, X-ray structure and absolute configuration determination of (+)-stemarin (8), "a diterpene with a new skeleton" isolated, along with (+)-2-deoxystemodinone (6), from S. maritima was described. [3] ( Figure 7).

Scheme 2.
Chemical work for the establishment of (+)-stemodinol (4) structure [2]. Some years later, in 1982, R.B. Kelly (University of New Brunswick, St. John, N.B.) and colleagues, who had previously described the total synthesis of (±)-stemarin (8) (vide infra), and Manchand, in the frame of their work on stemodane diterpenes, described the preparation of (±)-stemodinol (4). The comparison of the synthetic compound with the nominal identical compound isolated by Doorenbos and colleagues [2] showed the two materials were not the same. On the contrary, the comparison with an authentic sample of (+)-stemarin (8) showed its identity with the latter [4,5].
We decided to highlight the above work because it nicely represents the structure elucidation of new and complex natural products, such as stemodane diterpenoids, by the tools available in the middle of the seventies.
As will be reported in the sequel, also (+)-stemodinol (4) was eventually isolated from S. maritima. Its structure was established by HR-ESI-MS and 2D-NMR [6].
Later, in the frame of their work on stemodane diterpenoids, P.B. Reese (University of West Indies, Mona) and colleagues isolated a new compound from S. maritima (−)-stemodinoside D (stemodin-α-L-arabinofuranoside) (17). Its structure was established by HRMS(EI) and 2D-NMR [12]. Furthermore, in 2010, Arriaga and colleagues (Universidad Federal do Ceará) isolated stemodin 1 from the leaves of S. maritima collected in the Ceará coast in the northeast region of Brazil [13]. Its structure was established by EIMS, IR, 1 H-NMR and 13 C-NMR. The specific rotation was not reported. Later, in 2014, the above Brazilian group described the isolation of stemodinol (4), stemodin (1) and stemodinoside B from S. maritima (15) [6]. This was the first report about the isolation of (+)-stemodinol (4) from natural sources. The structures of 4, 1 and 15 were elucidated by means of HR-ESI-MS, 1 H-, 13 C-NMR and 2D-NMR. In this case, the specific rotations were also not reported. Finally, in 2018 Fu and colleagues (Hainan Normal University and Chinese Academy of Sciences) described the isolation from the stems and leaves of Trigonostemon heterophyllus Merr., among other products, of (+)-trigoheterone A (18), a new stemodane diterpenoid. The structure of the latter was established on the basis of extensive spectral analyses [14] ( Table 1).
Later, Wiesner and his group exploited this rearrangement for the synthesis of several diterpene alkaloids whose C/D ring system is constituted by a bicyclo[3.2.1]octane system [21][22][23][24][25][26][27][28]. This goal was achieved by locating, in a substituted bicyclo[2.2.2]octane system, at a specific position, a suitably configurated leaving group the departure of which promoted the migration of the antiperiplanar C-C bond. Remarkable examples of such specificity were reported in the past in the frame of biogenetic syntheses of stemodane [4,5,9,[29][30][31][32][33], aphidicolane [30,31,[34][35][36][37][38] and stemarane [39][40][41][42][43][44], diterpenes and diterpenoids. As can be seen (Scheme 4), the bonds indicated by a green arrow migrate in one case (entry a) through the lower face of the molecule, while in entry b, the same bond migrates through the lower face owing to the different location of the leaving group. From entry c in Scheme 4, it can also be appreciated (entry c vs. a) how a different orientation of the leaving group induces the migration of a different C-C bond.   [45], H.L. Goering and colleagues (University of Wisconsin, Madison) [46][47][48] and Kraus and colleagues (Tübingen University) [49]. They also observed that the nucleophilic group attacks the intermediate carbonium ion in such a way that the incoming nucleophile results afterward as exo configurated to the bicyclo[3.2.1]octane two C-bridge.
Having observed that in diterpenoids, whose C/D ring system is constituted by a bicyclo[3.2.1]octane moiety, the HO group at the former bridge-head C-atom is always exo configurated with respect to the two C-atom bridge and in view of the above-quoted Walborsky and Goering work [45][46][47][48], a biogenetic synthesis of (±)-2-deoxystemodinone (6), in which the desired (natural) substitution at C(13) was obtained in a 7:1 ratio over the C(13) epimer, was achieved in the past by our group [33].
These types of rearrangements have also been recently the object of a quantum chemistry investigation by Y.J. Hong and D.J. Tantillo (University of California-Davis) [50]. Their work clarified the mechanisms of formation from syn-CPP of stemodane diterpenes and related diterpenes. The complex network of reaction pathways involving concerted but asynchronous dyotropic and triple shift rearrangements was disclosed.

Biosynthesis
Differently from structurally related tetracyclic diterpenoid (+)-aphidicolin (28), an antimitotic and antiviral metabolite produced by the fungus Cephalosporium aphidicola Petch [51,52], whose biosynthesis was investigated by radioisotope and 13 C-NMR methods, revealing the origin of the carbon skeleton [53] and by incorporation of labelled intermediates [54], to our knowledge, similar studies were not carried out for stemodane diterpenes and diterpenoids.
Nevertheless, the synthases responsible for their biosynthesis, which occurs in plastids, were investigated. As known, diterpene synthases are encoded by nuclear genes, which are imported into the plastids after being synthesized in the cytoplasm thanks to cleavable N-terminal transit peptides [55,56]. Geranylgeranyl diphosphate (GGPP) (VII) is the precursor of diterpenes and diterpenoids in their biosynthesis. By means of this linear 20carbon isoprenyldiphosphate precursor, diterpene synthases generate diterpenes through a series of electrophilic cyclizations and/or rearrangements, starting with a carbocation formation and often ending by deprotonation [56].
GGPP (VII) is converted into syn-CPP (I) by class II terpenoid cyclases (Scheme 5). Several class I terpenoid cyclases are responsible for the further cyclization of CPP stereoisomers. Ent-kaurene synthase-like 8 and 11 (OsKSL8 and OsKSL11) are closely related (88.6% amino acid identity) as well as their encoding genes (92.0% nucleotide identity) [57]. Both enzymes are syn-CPP specific (Scheme 5). Unlike other OsKSL that produce only a single diterpene, OsKSL8 is able to generate two different diterpenes from syn-CPP, namely (+)stemod-12-ene (19) (~20%) and (+)-stemar-13-ene (20) (~70%). The enzymatic product of OsKSL11 is represented, on the contrary, almost entirely by stemod-13(17)-ene (31) [58]. The gene encoding for OsKSL11 has been first discovered in 2006 by R.J. Peters (Iowa State University), R.M. Coates (University of Illinois, Urbana) and colleagues in rice (Oryza sativa) [58]. Functional characterization of OsKSL11 showed that it was a syn-CPP specific exo-stemodene synthase (Scheme 5). This finding was surprising, first because the OsKSL11 sequence was not present in the rice genome database, then because stemodane type diterpenoids had not been yet identified in rice [58]. Even today, despite extensive phytochemical investigations, such compounds have not been identified in Oryza species. It should be noted that almost all of these studies focused on leaf phytoalexins. The Authors hypothesized that rice stemodane diterpenes could be produced and accumulated in other organs and/or in response to different stimuli (e.g., viral infections). Genes and enzymes involved in the biosynthesis of stemodane diterpenes and diterpenoids have not yet been identified so far in Stemodia species. Scheme 5. Proposed biosynthetic pathways for stemodane and stemarane diterpenoids in Oryza sativa L. (rice). OsCPS4, syn-copalyl diphosphate synthase; OsKSL8, ent-kaurene synthase-like 8; OsKSL11, ent-kaurene synthase-like 11.

Biological Activity
As reported by D.F. Austin (University of Arizona), in the folk medicine of the Dutch Antilles, an infusion of leafy branches of sea-lavender (Lavandula) and S. maritima, mixed with Epsom salts, is used by men against venereal diseases [59]. Plants of Calceolaria genus are used in Central and South America popular medicine as stomach tonics, bactericidal agents and sweeteners [60]. In the frame of a study on plants toxic to ruminants and equine in Rio Grande do Norte western and eastern Seridó, Riet-Correa and co-workers (Universidade Federal de Campina Grande, PB) reported that some farmers consider S. maritima (melosa), which grows in areas rich of mineral salts, toxic and responsible, inter alia, of aborts. Nevertheless, the Authors report that S. maritima has not been described as toxic, nor have experiments proven its toxicity [61].
Essential oils obtained by the Arriaga group from S. maritima leaves and stems, collected in the state of Ceará, showed larvicidal properties against the Aedes aegypti mosquito larvae, responsible for the transmission of yellow fever in Central and South America and in West Africa and a vector of dengue hemorrhagic fever [62]. However, the main compounds obtained by hydrodistillation from leaves and stems were β-caryophyllene, 14-hydroxy-9-epi-β-caryophyllene and caryophyllene oxide. This biological activity cannot, therefore, be attributed to the stemodane diterpenes.
Besides, in 2017, the University of Ceará group, headed in this case by M.M. Bezerra, also observed that S. maritima extracts decreased inflammation, oxidative stress and alveolar bone loss in an experimental periodontitis rat model. In the frame of this work "no signs of systemic illness, adverse pharmacological events or changes in behavior were observed throughout the experimental period" [63].
Recently a phytochemical analysis and the gastro-protective effects of S. maritima hexane extract were also described [64].
Hypotheses about the hydroxylation of the above compounds by the involved enzymes in relation to the binding sites are also proposed [69].
The regio-and stereochemistry of the hydroxylation was also discussed. The structures of the above metabolites and those obtained in the following studies by the group were established by HRMS and 1 H-and 13 C-NMR [71,72].
Hypotheses on xenobiotic biotransformations vs. biosynthetically-directed transformations in relation to the hydroxylated sites and the role of nearby hydroxyls were also put forward.
In recent years, numerous derivatives of stemodane diterpenes have been obtained by means of biotransformation and some of them have shown remarkable biological activities, highlighting their potential as active ingredients for pharmaceutical, agricultural and industrial applications. In our opinion, the research activity in this area is currently underdeveloped and it would be beneficial to improve efforts for the discovery of new molecules and for preclinical and clinical experimentation, mainly focusing on the development of new antiviral, antibacterial and anticancer drugs.
As it can be seen, only a few of the above compounds limited the growth of some human tumor cell lines. On the contrary, most of them induced proliferation of such cells ( Figure 19) [83]. As far as analogues/derivatives 80-87 are concerned, in the LPO three compounds showed prominent activity; besides all derivatives displayed higher COX-1 enzyme inhibition than COX-2; finally, only few of the latter compounds limited the growth of some human tumour cell lines while most of them induced the cell proliferation.

Conclusions
In this review, the work which involved various research groups for about fortyfive years on stemodane diterpenes and diterpenoids structure elucidation, biogenesis, biosynthesis, biological activity and biotransformations is described.
The efforts which led to the structure elucidation of tetracyclic diterpenoids with a new skeleton, at a time in which 13 C-NMR and advanced analytical techniques were not available yet, were described.
The probable biogenesis of stemodane diterpenes, as well as the work which allowed identification of the enzymes responsible for stemodane diterpenes biosynthesis was reviewed.
The biological activity of (+)-stemodin (1), congeners, metabolites and derivatives was also taken into account, as well as the biotransformations operated by a number of microorganisms, which permit to hydroxylate most of the stemodane nucleus C-atoms, allowing easier access, compared to total synthesis, to hydroxylated derivatives.
This class of structurally intriguing compounds was also the target of a large number of syntheses. Owing to the vastness of this aspect of stemodane diterpenoids chemistry, they will be considered at another time.