Germacrene A–A Central Intermediate in Sesquiterpene Biosynthesis

Abstract This review summarises known sesquiterpenes whose biosyntheses proceed through the intermediate germacrene A. First, the occurrence and biosynthesis of germacrene A in Nature and its peculiar chemistry will be highlighted, followed by a discussion of 6–6 and 5–7 bicyclic compounds and their more complex derivatives. For each compound the absolute configuration, if it is known, and the reasoning for its assignment is presented.


Introduction
With an estimated number of over 80,000 compounds terpenes form the largest class of natural products.T hey are produced by all kingdoms of life and can be classified as mono-(C 10 ), sesqui-(C 15 )o rd iterpenes (C 20 )e tc. according to the number of incorporated isoprenoid units. During the past decades many sesquiterpene synthases have been reported [1][2][3][4][5][6] that catalyse the cyclisation of farnesyl diphosphate (FPP) through diphosphate abstraction to give the reactive farnesyl cation (A,S cheme 1). Attack of the C10 = C11d ouble bond to C1 can yield the (E,E)-germacradienyl cation (B)b y1 ,10-or the (E,E)-humulyl cation (C)by1,11-cyclisation. The alternative reaction by reattack of diphosphate to C3 resultsi nn erolidyl diphosphate (NPP). After ac onformational rearrangement of the vinyl group by rotationa roundt he C2ÀC3 bond, cyclisation reactions mayp roceed to the (E,Z)-germacradienyl cation (D), the (E,Z)-humulylc ation (E), the bisabolyl cation (F), or to cation G,w ith possible formation of either enantiomer for chiral intermediates. Deprotonation of B leads to germacrene A, aw idespread natural product and central intermediate in the biosynthesis of many 1,10-cyclised sesquiterpenes. This review discusses its occurrence in Nature, its chemistry,a nd centrali mportance as an intermediate towards many sesquiterpenes.
vouring the latter hypothesis. In fact, the enantiomeric composition of ac ompound cannotb ec oncluded only from the optical rotation upon its first isolation, or not with certainty if a compound is known to be instable. Methods such as chromatographic separation on ac hiral stationary phase mayb em ore conclusive. Through this approach, Kçnig and co-workers found that 1 from various plants is am ixture of enantiomers, ranging from nearly pure (+ +)-1 in Piper nigrum to mainly (À)-1 in the liverwort Barbilophozia barbata. [14] GermacreneAsynthase (GAS) catalyses a1 ,10-cyclisation of FPP to B,f ollowed by deprotonationt o1 (Scheme 3). Both enantiomers of 1 are accessible through this reaction, depending on whether C10 of FPP is attacked from the Re or the Si face. Since this faceselectivity may be altered by subtle conformationalchanges of FPP in the active sites of GASs,predictions based on amino acid sequences or phylogenetic analyses regardingt he stereochemical implications may be difficult. Many plant GAS have been identified during the past two decades, including two (+ +)-GASs from Cichorium intybus [15,16] and one from Matricaria recutita, [17] with the absolute configurationo f (+ +)-1 established by chiral GC. Sometimes the absolute configuration can be rationally suggested, because 1 is transformed in the same organism into another compound such as (+ +)-costunolide. [18][19][20][21] Further GASs are known from many other plant species, [22][23][24][25][26][27][28][29][30][31][32] but the absolute configurationo f1 has frequently not been determined. While the accumulated literature shows that (+ +)-1 is typical for plants, the recently characterisedb acterial GAS from Micromonosporam arina produces (À)-1, [33] reflectingt he observationt hat terpenes and cationic intermediates towards them from plants and bacteria often represent differente nantiomers. [34][35][36][37] The coinciding absolutec onfiguration of (À)-1 from E. mammosa may point to ab iosynthesis by symbiotic bacteria in the gorgonian. [38]

Chemistry of germacrene A
The isolation and full structural and NMR-spectroscopicc haracterisation of 1 was al ong-standingp roblem significantly hampered by its high reactivity.I ts first isolation from E. mammosa in 1970 was done by extraction and concentration at temperatures below 35 8Ct oa void the Cope rearrangement to 2 (Scheme 2). [7] Chromatographic purification on slightly acidic silica gel induces ac yclisation through cation H1 to a-selinene (4), b-selinene (5), and selina-4,11-diene (6,Scheme 4). [7,11,15] The skeleton of 1 is characterisedb yaconformationally flexible 10-membered ring that shows sufficient ring strain to prevent af ast interconversion between conformers, resulting in broadened signals and multiple signals ets in the NMR spectra. Partial 1 H-and 13 C-NMRd ata were first published for 1 from T. maculata. [12] Later studies improved the NMR data assignments for the main conformers of 1 (recorded at 25 8C), but did not allow for ac ompletion of the data sets. [13,39] Through NOESY the conformers of 1a (UU, Me14 and Me15 up), 1b (UD, updown) and 1c (DU, down-up) in a5 :3:2 ratio were identified (Scheme 5A). [13] The NMR data sets (25 8C) for all three conformers were recently completed using a 13 C-labelling strategy by conversion of all 15 isotopomers of ( 13 C)FPP [40] with GAS from M. marina into (À)-1,r esulting in strongly enhanced 13 C-NMR signals for the labelled carbons. HSQC spectroscopy of enzymatically prepared stereoselectively deuterated and 13 C-labelled 1 allowed the NMR assignment of allh ydrogens. [33,41] The stereoselectively deuterated and 13 C-labelled isotopomers of 1 were also used to study the stereochemical course of its Cope rearrangement (Scheme 5B). According to the Woodward-Hoffmann rules, pericyclic reactions follow as tereochemical course determined by the symmetry of frontier orbitals. [42] For the Diels-Alder reaction this has been verified by stereoselective deuteration, [43,44] while classical experiments for the Cope rearrangement have been performed with meso-a nd rac-3,4-dimethylhexa-1,5-diene. [45] The enzymatic access to labelled 1 allowed to follow the rearrangement to (+ +)-2 that proceeds from ent-1a through ac hair-chair transition state. [33] For many terpene synthase reactions 1 is furtherc yclised in as econd step initiated by reprotonation. This can occur at C1 and lead to the 6-6 bicyclic systemo fH as ap recursor of eudesmane sesquiterpenes (Scheme 6A). The 6-6 bicyclic system could in theory also arise by protonation at C4 leading to the secondary cation I,b ut this reaction is not preferred. Furthermore, 1 can be protonated at C10 with cyclisation to the 5-7 bicyclic skeleton of J,o ra tC 4r esulting in K,r epresenting the precursors to guaiane sesquiterpenes. As an alternative to the formation of neutral 1 and its reportonation also an intramolecular or water-mediated proton transfer in cation B may directly lead to H, J or K,t hus bypassing 1 that would in such cases be better described as as ide product rather than an intermediate. However,e xperimentale vidence to distinguish between these alternatives is difficult to obtain,a nd 1 will preferentiallyb ed iscussed as an intermediate towardsm ore complex sesquiterpenes in this article. Ad etailed discussion of the reactions from 1 will follow in the subsequentsections.

Eudesmanes with ar egular skeleton
The protonation-induced cyclisation of 1 can lead to eight stereochemically distinct cationic intermediates (Scheme7), four of which arise from (+ +)-1 (H1-H4), while the other four stereoisomerso riginate from (À)-1 (H5-H8). For each intermediate, simple deprotonations or nucleophilic attack of water are possible. Also, hydride shifts can occurf irst, which further widens the reachable chemical space of eudesmanes.F or many of these possibilities the corresponding structuresh ave been reported.

Eudesmanes from cation H1
An important intermediate to eudesmanes is H1.D eprotonations from C3 and C15l ead to a-selinene (4)a nd b-selinene (5), two compounds that have been isolated more than 100 years ago from celery oil. [46] Their structures were elucidated in degradation experiments [47] and were correlated to b-eudesmol (7,Scheme 8A). [48][49][50] Based on acomparison of physical characteristics of degradation products to those of other cis-a nd trans-decalins initially a cis-decalin structure was assigned, [51] but al ater conformational re-examination indicated a transfused ring system. [52,53] The absolute configurations of 4 and 5 were determined by chemical correlation through the following arguments. The structure of ketone 8 was established in the classicals ynthesis of steroids by Woodward. [54] Twoy ears later the same group converted 8 into the dicarboxylic acid 10 (Scheme 8B)t hat was the oppositee nantiomer as obtained by degradation of 7 (Schemes 8C) [55] that had previously been correlated with 4 and 5 (vide supra).
The first report about naturallyo ccurring enantiomers of selinane sesquiterpenes identified ent-4 as ac onstituent of the liverwort Chiloscyphus polyanthus in 1973. Its absolute configuration was established by CD spectroscopy in comparison to authentic (À)-4. [60] Compounds ent-4 and ent-6,l ikewise established by CD spectroscopy and accompanied by 2,w ere subsequently reported from the liverworts Diplophyllum albicans and D. taxifolium, [143] while the liverworts Riccardia jackii, Bazzania spiralis and Tylimanthust enellus contain different combinations of ent-4, ent-5 and ent-12. [144][145][146][147] Also insectsw ere reported to contain ent-4 and (+ +)-2,exemplified by their occurrencei nCeroplastes ceriferus,w hich is surprising considering the fact that the "normal" enantiomeric series of compounds is present in the related species C. rubens. [62] In all thesee xamples the absolute configurations wered etermined from the opticalr otations of the isolated compounds. In Penicillium roqueforti also ent-4, ent-5 and ent-12 mayo ccur;i nt his case the absolute configurations were assigned based on their biosynthetic relationship to aristolochene( vide infra)t hat is generated through( À)-1 in this fungus. [148]

Eudesmanes from cation H6
Little is known about eudesmanes arising via cationic intermediate H6.T he compound 7-epi-a-selinene (ent-23, Scheme16A)w as first reported from Amyrisbalsamifera,aspecies from which also 7-epi-a-eudesmol (51,S cheme 16 B) was isolateda nd structurally characterised by NMR spectroscopy. From itsp ositive opticalr otation ([a] D =+10, c 1.8, CHCl 3 )t he authors concluded on the shown absolute configurationf or 51,b ut ac omprehensible explanation for this assignment is missing. Dehydrationo f51 yielded am ixture of two products to which the structures of ent-23 and 52 were assigned by NMR spectroscopy,u nfortunately without separating the obtained materials and determining their optical rotations. The compounds described as ent-23 and 52 also occurred in the essential oil of A. balsamifera. [149] One study reportedt he chro-matographic separation of the compound from A. balsamifera and (+ +)-23 (the latter with am entioned source "provided by Dr.W ilfriedK çnig") on ac hiral stationaryG Cp hase, which represents the only hint in the literature that the structure of ent-23 for the essential oil constituent may be correctly assigned. [150] Compound ent-23 was also reported as major product of at erpene synthase from Vitis vinifera. [150,151] Both enantiomerso f23 have been obtained by synthesis from the enantiomerso f15,b ut optical rotary powers of the products were not measured. [152] However, ent-23 may have an egative optical rotation, as for 23 from Dipterocarpusa latus al ow value of [a] D 20 =+2.1 was determined. [109] This would be consistent with ar eport by Kçnig in which ent-23 was published as the (À)-enantiomer,a lbeit only based on separation by gas chromatography using ac hiral stationary phase without isolation. [153] Compound ent-25 ([a] D 16 =+46.5, c 0.85, CHCl 3 )h as been synthesised using the same strategy as for 6 (Scheme 9C), [78] but has not been isolated from any organism.T he only report about ent-26 from Monactis macbridei by Bohlmann and coworkers [154] gives ar eference to the erroneous "isointermedeol" [128] that was corrected shortly after. [129] Unfortunately, Bohlmann's paper does not give an opticalr otationf or the isolated material so that it is difficult to judge, if the authors of this study were aware of the misassignment of "isointermedeol" at the time of their publication. Overall,t his discussion shows that compounds from H6 are not only rare, buti ft hey occur in the literature,t he assignments of absolute configurations remain unclear.S ince the compounds originate in all cases from higherp lants, they may truly be the usual enantiomers, that is, 23, 25 and 26.

Rearranged Eudesmanes
In this sectionr earranged eudesmanes from H1-H6 will be discussed, while such compounds from H7 and H8 are unknown.

Rearranged eudesmanes from H1
Rearranged eudesmanes can in theory arise from all cations H1-H8 in Scheme 7. An important group of compounds by widespreado ccurrence in Nature originates from H1.S pecifically,t his intermediate can undergo a1 ,2-hydride migration to H1 a that mustp roceed suprafacially and thus determines the configuration at C4 (Scheme 18;1 ,n-hydride or proton migrations as used in this article refer to the distance of nc arbons for the migration, not to positional numbers). As ubsequent 1,2-methyl group migration leads to H1 b (path a) that upon deprotonation yields eremophilene (55)o r4 ,5-diepi-aristolochene (56). Alternatively, H1 a can react in aW agner-Meerwein rearrangement (WMR) with ring contraction to H1 c that results in hinesene (59,path b).
Valencene (68)w as first isolated from orange oil [180] and found to be related to nootkatone (69)b yo xidative conversion, [181] an important value addingt ransformationf or which an artificial enzyme system has been developed. [182] Compound 69 is af lavour constituent of citrus fruits and its structure had previously been established. [183] The opticalr otation of 68 was determined for the material obtained by dehydration of valerianol (76,S cheme 20 B) with NaOAc in refluxing Ac 2 O( [ a] D =+73.4, c 5.3, CHCl 3 ). [184] Asynthesis of (rac)-68 similar to the synthesis of (rac)-55 in Scheme 19 Ah as been developed. [170] The sesquiterpene 68 is ac onstituent of the essential oils from numerous plants, but has rarely been isolated. Bixa orellana is one of the few sources from which its isolationw as mentioned, [185] while it was obtained enriched togetherw ith 2 in asesquiterpene hydrocarbon fraction from the liverwort Porella acutifolia. [186] The combination of 2 and 68 also occurs in the octocoral Plexaurella fusifera, [168] while 68 from bacteria is rare, but has been identified from Streptomyces sp. FORM5. [187] Valencene synthases are known from Citrus sinensis, [188] Vitis vinifera, [150,151] and Callitropsis nootkatensis, [189] in which it occurs together with av alenceneo xidase for the biosynthesis of 69. [190] Besides 68,t he terpenes ynthases from V. vinifera were reportedt op roduce (À)-7-epi-selinene (ent-23, Scheme16) [150,151] that must originate from H6.I tw ould be easier to understand, if one of the two enzymep roducts would represent the opposite enantiomer than reported, so that both could arise through ac ommon intermediate. In fact, the configurational assignment for 68 wasb ased on aG Ca nalysis using ac hiral stationary phase, but without including a (À)-68 standard.

Rearranged eudesmanes from H5
Only af ew reportsa bout rearranged eudesmanes from H5 from Nature are available (Scheme 25). Terpene synthases for ent-55 have been characterised from the myxobacterium Sorangium cellulosum ([a] D 25 =+131.7, c 1.0, CHCl 3 ) [239] and the plant pathogenic fungus Fusarium fujikuroi. [240] The cyclisation mechanism of (+ +)-eremophilene synthase from F. fujikuroi was studied by isotopic labelling experiments that showed selective deprotonation from C12 of FPP in the formation of the intermediate (À)-1,a llowed to follow the 1,2-hydride shift from H5 to H5 a,a nd demonstrated that the final deprotonation from H5 b to ent-55 proceeds with loss of the same protona s incorporated in the cyclisation of (À)-1 to H5 (Scheme7). [240] A crystal structure of ent-55 [239] and full NMR data assignments have been published. [239,240] Only as ynthetic study towards ent-56 ([a] D 25 =+12.5, c 2.5, CHCl 3 )i sa vailable. [241] 4.6. Rearranged eudesmanes from H6 Rearranged molecules from H6 (Scheme 26 A) are (À)-valencene (ent-68)a nd (+ +)-aristolochene( ent-70)t hat has been isolated from Aspergillus terreus ([a] D =+79.4, c 0.0176, hexane), [192,196] and Penicillium roqueforti,i nw hich it occurs to-gether with 2. [148,242,243] The absolute configuration hasbeen established by synthesis of (À)-70 from (+ +)-valencene (68). [192] (+ +)-Aristolochene synthase was first isolated from P. roqueforti (PR-AS) [244] and is also presenti nA. terreus (AT-AS). [245] Subsequent gene cloning and expressiongave efficient access to the recombinant enzymes. [246,247] Ab iphasic flow reactors ystem for the biocatalytic production of ent-70 has been developed. [248] Notably,P R-AS produces am ixture of ent-70 as the main and ent-68 and (À)-1 as side products, while AT-AS yields ent-70 as as ingle product. [249,250] Isotopic labelling experiments demonstrated that the cyclisation of FPP to ent-70 proceeds with inversion of configuration at C1 andt he specific loss of a protonf rom C12. [245] The E252Q variant of PR-AS yielded( À)germacrene A( 1)a st he only product. [250] Further support of (À)-1 as an intermediate was obtained by the observed cyclisation of (R)-5,6-dihydro-FPP (100)tothe germacrene Aanalogue 101 by AT-AS (Scheme 26 B). [251] Similare xperimentsh ave been carriedo ut with fluorinated FPP analogues. [252,253] On the other hand, insteado fatrue pathway intermediate, (À)-1 could only be as hunt product. Allemann and co-workersh ave argued for this view,a s( À)-1 was not accepted as as ubstrate by PR-AS, [249] and ac omputational study showed feasibility of a water-mediated direct protont ransfer from (S)-B to M that could further cycliset oH6 (Scheme 26 C). [254] However,t he same workers later excluded this possibility experimentally,b ecause the incorporation of deuterium from D 2 Oa tC 1o fent-70 proceeded with Re face attack. [255] Based on the crystal structure of PR-AS the actives ite residueT yr92 wass uggested to serve as ag eneral acid in the reprotonation of (À)-1, [256] but also this hypothesis was disfavoured by site-directed mutagenesis. [250] Am ore detailed picture was subsequently obtained by the crystal structure of AT-AS,p roviding evidence that the diphosphate anion is ideally positioned to act as ag eneral acid and base relevant fori )t he deprotonation of (S)-B,w ith the proton taken up by O6, and ii)t he reprotonation of the resulting (À)-1 with donation of ad ifferent proton from O3 (this process may also be concerted with 1 as ah ighly transient species, Scheme 26 D). [257] The results of as ite-directed mutagenesiss uggest that the thus formed eudesmane cation H6 is stabilised by W334 of PR-AS or W308 of AT-AS. [258] Cationic azaanalogues of H6 have been shown to efficiently inhibit catalysis by PR-AS. [259,260] The sesquiterpeneh ydrocarbon ent-70 is the biosynthetic precursor to PR toxin (99), [261] ap otent mycotoxin that targets transcription and protein biosynthesis with al ethal dose of LD 50 = 5mgkg À1 in mice, [262][263][264] and as eries of other oxidation products that are likelyp athway intermediates. [265][266][267][268][269] Surprisingly,d espite the potential of mycotoxin biosynthesis P. roqueforti is traditionally used for the production of blue cheese, which is explainable by the rapid degradation of 99 under cheese fermentation conditions. [270] Biosynthetic hypotheses linking these oxidised metabolites have been investigated by feeding of labelledp recursors [148,269] and discussed on the grounds of the biosyntheticg ene cluster, [271][272][273] but apart from the aristolochene synthase and the poorlyc haracterised eremofortin Co xidase [274] for the installation of the aldehyde function in 99 little is known about the enzymes involved in fungal toxin biosynthesis.

Guaianes formed by C4 protonation of germacreneA
Eight cationic intermediates can be formed from the enantiomers of 1 by protonation atC 4a nd ring closure (Scheme 27). These cations exhibit four stereogenic centres,l eading to a maximum number of 2 4 = 16 possible stereoisomers, but two of the stereogenic centresa re not set independently,s ince the C4/C5 double bond in 1 is E-configured and the ring closure proceeds by anti addition, that is, Me15 and H5 mustb ea rranged trans. Thus, only eight stereoisomers are relevant to this pathway, namely J1-J4 from (+ +)-1,a nd their enantiomers J5-J8 from (À)-1.

Guaianes formedf romcationsJ3a nd J4
Guaianes from J3 and J4 include guaia-1(10), 11-diene (115) that is accessible through both cations by deprotonation, and guaia-9,11-diene (116)o btainable by loss of ap roton from J3 (Scheme30A). Deprotonation of J4 can lead to guaia-10(14), 11-diene (117), ac ompound for which we revise the structure here based on the reason given below,w hile the attack of water to J4 can give 4,5-diepi-pogostol ( 118). For 118 this discussion is hypothetical, because this compound was only obtained in racemic form by synthesis and is not known as natural product. [299] The hydrocarbons( + +)-115 and( + +)-116 were both isolated only from the fruits of Peucedanum tauricum. [311] Their co-occurrencei no ne organisms uggests that they may have the same cationic precursor J3.T he absolute configurations of 115 and 116 were specified by comparison of their hydrogenation products to those obtained from (+ +)-g-gurjunene (120, Scheme 30 B), [312] leading to one common product (119a)f rom all three materials, as judged by GC analysisu sing two different chiral stationary phases.
Guaia-10(14),11-diene (117)i sonly known from Abies koreana. [121] Its absolute configuration was elaborated using the same hydrogenation strategy as for 115 and 116 with chemical correlation to aciphyllene (122,S cheme 30 C). At the stage of this work the structure of 123 with 7S stereochemistry was assigned fora ciphyllene, [284] which would have led to the hydrogenationp roducts 119f and 119i,a nd therefore the structure of 121 was concluded for the naturalp roduct from A. koreana expectedt og ive the hydrogenation products 119f and 119g. However, shortlya fter thes tructureo fa ciphyllene underwenta revision to (7R)-122. [313] In conclusion,t he trulyo btainedh ydrogenation products from aciphyllenew ere 119c and 119h,w ith theconsequence that then atural product from A. koreana must be revisedh erewitht o117,e xpectedt og ive 119c and 119d.
The synthetic compound 1-epi-aciphyllene (124)h as been prepared from guaiol (114), [314] but has not been discovered from Nature so far.I ndeed, its biosynthesis is not easily understood, as its formation through the K series (Scheme 32, Secion 5.5) of cations cannotl ead to a cis-orientation of H1 and Me14. If 124 exists at all as an atural product, two sequential 1,2-hydride migrations from J4 to J4 a and deprotonation could explain its formation (Scheme 31). Full 1 H-and 13 C-NMR data for 124 were reported, [314] but unfortunately no optical rotation that would be useful for comparison in case of its future isolation.

Guaianes formed fromcationsJ5-J8
Despite the fact that for 103 the absolute configuration has not been determined and this compound could in principle arise through J6,noguaianes from J5-J8 are known.The absolute configuration of 1,4-diepi-g-gurjunene (109)f rom C. hooperi would be most interesting to know,a ss pongesm ay produce the optical antipodes of plant compounds.

Cyclised and Rearranged Guaianes
Further cyclisations eventually with skeletal rearrangements are important for two groups of compounds originating from J1 and J3,w hile no examples from the other cations of the J series or from cations of the K series are known.

Compounds from J3
The biosynthesis of rotundene (136), isorotundene( 137)a nd cyperene ( 138)c an be understood from J3 (Scheme 37 A). Its cyclisation to J3 a (path a) and deprotonation yields 136 and 137,w hile a1 ,2-hydride shift to J3 b (path b) followed by a 1,5-proton shift to J3 c,c yclisation to J3 d and deprotonation result in 138.T his common biosyntheticp athway nicely explains the co-occurrence of 136-138 in Cyperusr otundus. [347] Compound 136 ([a] D = À16.3) was first reported from C. rotundus and C. scariosus, [348] and later also from C. alopecuroides, [349] but at this stage only with the planar structure. (À)-Isorotundene (137)w as isolated from C. rotundus whoser elative configuration was determined by NOESY. [347] This allowed to demonstrate that 136 has the same skeleton by conversion into rotundol( 139)t hrough oxymercuration and dehydration with POCl 3 (Scheme37B). The absolute configuration of 136,a nd thus also of 137,w as determined by ozonolysis to 140,d ecar-boxylation to am ixture of epimers 141 ab,W ittig methylenation to 142 ab and catalytic hydrogenation to 119ab (Scheme 37 C). One of these hydrocarbons wasi dentical to 119a obtained by hydrogenation of 120 (Scheme 30 C). Complete 1 H-and 13 C-NMR data for 137 have been reported, [347] but are lacking for 136.

Conclusions
Germacrene As hows au nique and interesting chemistry mainlyc haracterisedb yi ts reactivityt owards acid-catalysed cyclisationsa nd its thermall abilityi naCope rearrangement to b-elemene. Similarobservations have been made for other germacrenes, [369] suggesting that the high ring strain associated with the 10-membered ring in these systems may be as trong drivingf orce for the observed reactions leadingt om uch less strained compounds with 6-membered rings. The reactivity built up by the ring strain is also used in enzymatic reactions towardss esquiterpenes for which germacrene As erves as an important intermediate.I ne nzyme reactions not only the formationo f6 -6 bicyclic compounds, but also of 5-7 bicyclic derivatives can be achieved, and for both cases follow-up chemistry by skeletal rearrangements can further increaset he structural variability.S ubsequents teps include oxidativea nd other modifications after terpene cyclisation, leading to numerous derivatives for each compound presented in this review,w hich furtheru nderlines the central importance of germacrene Ai n sesquiterpene biosynthesis.