Isolation and characterisation of irinans, androstane-type withanolides from L.

2003 Isola ion and cha ac e isa ion o i inans, and os ane- ype wi hanolides om Physalis pe u iana L. Annika S ein‡1, Da e Compe a‡1, Bianka Ka ge2, Ma k B öns up1,2 and Jakob F anke*1 Full Resea ch Pape Open Access Add ess: 1Cen e o Biomolecula D ug Resea ch, Leibniz Uni e si y Hanno e , Schneide be g 38, 30167 Hanno e , Ge many and 2Helmhol z Cen e o In ec ion Resea ch, Inho ens asse 7, 38124 B aunschweig, Ge many Email: Jakob F anke* - [email p o ec ed] * Co esponding au ho ‡ Equal con ibu o s Keywo ds: and os anes; Physalis pe u iana; s e oids; s uc u e elucida ion; wi hanolides Beils ein J. O g. Chem. 2019, 15, 2003–2012. doi:10.3762/bjoc.15.196 Recei ed: 31 May 2019 Accep ed: 07 Augus 2019 Published: 23 Augus 2019 This a icle is pa o he hema ic issue "Te penes". Gues Edi o : J. S. Dickscha © 2019 S ein e al.; licensee Beils ein-Ins i u . License and e ms: see end o documen . Abs ac Wi hanolides a e s e oidal lac ones widesp ead in Nigh shade plan s wi h o en po en an ip oli e a i e ac i i ies. Addi ionally, he s uc u al di e si y o his compound class holds much po en ial o he disco e y o no el biological ac i i y. He e, we epo wo newly cha ac e ised wi hanolides, named i inans, om Physalis pe u iana wi h highly unusual unca ed backbones ha esemble mammalian and os ane sex ho mones. Based on biomime ic chemical eac ions, we p opose a model ha links hese compounds o wi hanolide biosyn hesis. I inans ha e po en an ip oli e a i e ac i i ies, ha a e howe e lowe han hose o 4ß-hyd oxywi hano- lide E. Ou wo k es ablishes and owi hanolides as a new subclass o wi hanolides. 2003 In oduc ion T adi ional medicine has long been a sou ce o inspi a ion o mode n d ug esea ch. An impo an example is Wi hania somni e a, also known as ashwaghanda o Indian ginseng, which has been used in Ayu edic medicine o ea a la ge a ie y o ailmen s [1]. Ex ensi e s udies e ealed wi hanolides, a class o s e oidal lac ones, o be p ima ily esponsible o he medicinal e ec s [1,2]. A la ge ange o pha macological p op- e ies has been assigned o wi hanolides, wi h an ip oli e a i e ac i i ies being he mos po en ones [1]. Wi hanolides ha e been also disco e ed in nume ous gene a o he han Wi hania, o example Da u a, Dunalis, Ioch oma, Jabo osa, Lycium and Physalis [3], esul ing in mo e han 300 known ep esen a i es [3]. Physalis pe u iana is a wi hanolide p oduce o pa icula ele ance as i is widely cul i a ed o i s edible be ies [4]. So a , se e al wi hanolides ha e been epo ed om P. pe u iana and o he Physalis species, mos p ominen ly physalins, pe u- lac ones and 4ß-hyd oxywi hanolide E (1) [5-16]. As pa o ou ongoing p og amme ocussed on he biochemis y o wi hano- lides, ou aim was o gain u he insigh s in o he wi hanolide p o ile o P. pe u iana. He e we epo i inans A (2) and B (3), wo unusual unca ed wi hanolides wi h and os ane backbones. We show ha oxida i e, bu no acidic o basic condi ions Beils ein J. O g. Chem. 2019, 15, 2003–2012. 2004 Figu e 1: Wi hanolides om Physalis pe u iana. A) S uc u es o he newly cha ac e ised unca ed wi hanolides i inan A (2) and B (3) wi h an and os ane backbone, oge he wi h he known wi hanolides 1 and 4–6 isola ed om P. pe u iana. B) P. pe u iana plan . enable con e sion o he pu a i e p ecu so 4ß-hyd oxywi h- anolide E (1) o i inan A (2). Based on his in insic eac i i y we p opose a biosyn he ic model ha will se e as u he guid- ance o elucida ing he enzyma ic basis o and os ane o ma- ion in plan s in he u u e. Resul s and Discussion To isola e wi hanolides om P. pe u iana, we used a pu i ica- ion s a egy based on p e ious epo s [17-19]. Nine weeks old whole P. pe u iana plan s (140 g) we e ex ac ed wi h H2O/MeOH (3:1) and di ided in o ac ions soluble in pe o- leum e he , chlo o o m, and n-bu anol, espec i ely. The chlo o o m ac ion was u he sepa a ed by lash ch oma og a- phy on a C18 s a iona y phase, esul ing in h ee majo sub ac- ions F1–F3. Final pu i ica ion by p epa a i e HPLC ollowed by NMR analysis e ealed 4β-hyd oxywi hanolide E (1) as he majo compound (50 mg) as well as he known me aboli es wi hanolide E (4), wi hanolide F (5) and pe ulac one H (6) by compa ison o li e a u e da a (Figu e 1) [9,20]. Two addi ional compounds a ac ed ou a en ion based on hei unusual 1H NMR spec a (Table 1). Bo h showed wo mul iple s in he ole inic egion, which a e highly cha ac e is ic o wi hanolides wi h A- ing Michael accep o s. Howe e , com- pa ed o o he wi hanolides, se e al signals we e missing. Typi- cally, wi hanolides show i e single s o me hyl g oups in he alipha ic egion, as well as he H-22 oxyme hine p o on o he lac one moie y. Su p isingly, bo h compounds showed only wo pu a i e me hyl signals, and no signal which migh co espond o H-22. Thus, we easoned ha bo h unknown compounds migh be unca ed wi hanolide-like compounds. HRESIMS sugges ed a sum o mula o C19H24O5 o he i s compound, which was suppo ed by he 13C spec um (Table 1). By compa ing he spec um o NMR da a o o he wi hanolides, we quickly iden i ied he Michael sys em in ing A based on wo ole inic p o ons (δH 6.94 and 6.22 ppm), a seconda y alcohol a C-4 (δH 3.79 ppm), a 5,6-epoxide (δH 3.37 (H-6)), and a e ia y alcohol a C-14 (δC 80.9 ppm). COSY co ela- ions suppo ed by HMBC analysis (Figu e 2A) e ealed an in ac ABCD ing sys em wi h a subs i u ion pa e n iden ical o 4β-hyd oxywi hanolide E (1). Only a single, s iking di e ence was no ed: C-17 was shi ed om 87.8 o 218.0 ppm, s ongly sugges ing he p esence o a ke one ins ead o an alcohol. In ag eemen wi h he p edic ed sum o mula and he absence o all side chain ca bons, his comple ed he s uc u e o he i s unknown compound, which we named i inan A (2, Figu e 1). The second unknown compound had a sum o mula o C19H24O3 based on HRESIMS and 13C NMR (Table 1). In con as o he i s compound, no epoxide and no seconda y alcohol a C-4 was p esen , in ag eemen wi h he di e en elemen al composi ion. Ins ead, 13C NMR indica ed a double bond a C5–C6 (δC 135.7 and 124.3 ppm). O he wise, all spin sys ems and co ela ions indica ed a ypical wi hanolide ABCD Beils ein J. O g. Chem. 2019, 15, 2003–2012. 2005 Table 1: 13C and 1H NMR da a (CDCl3, 500 MHz, 298 K) o i inans A (2) and B (3) in compa ison o he known compound 4β-hyd oxywi hanolide E (1, CDCl3, 400 MHz, 298 K, δ in ppm, J in Hz). Fo ca bon numbe ing see Figu e 1 and Figu e 2. 13C1H posi ion 1 2 3 1 2 3 1 201.9 202.0 203.8 – – – 2 133.2 132.3 128.1 6.22 (1H, d, 9.9) 6.22 (1H, d, 10.0) 5.91 (1H, ddd, 10.0, 3.1, 1.2) 3 141.4 142.1 145.3 6.92 (1H, dd, 9.9, 6.1) 6.94 (1H, dd, 10.0, 5.8) 6.79 (1H, ddd, 10.0, 5.0, 2.6) 4 70.4 69.9 33.6 3.74 (1H, d, 6.1) 3.79 (1H, dd, 5.8, 2.4) 3.31 (1H, dddddd, 21.3, 2.8, 2.8, 2.8, 2.8, 2.8)a 2.88 (1H, dd, 21.2, 4.9) 5 64.2 63.9 135.7 – – – 6 63.1 63.1 124.3 3.28 (1H, b s) 3.37 (1H, m) 5.64 (1H, d , 5.7, 2.0) 7 26.0 24.9 24.1 2.03 (2H, m) 2.11 (1H, d , 14.2, 3.1) 1.84 (1H, ddd, 14.1, 11.7, 1.4) 2.08 (1H, m) 1.95 (1H, m) 8 34.3 32.6 35.5 1.83 (1H, m) 1.90 (1H, m) 1.88 (1H, m) 9 36.7 38.1 37.1 1.69 (1H, m) 1.51 (1H, m) 2.10 (1H, m) 10 47.9 47.8 50.9 – – – 11 21.5 20.4 21.6 1.72 (1H, m) 1.56 (1H, m) 1.91 (1H, m) 1.46 (1H, m) 2.34 (1H, m) 1.52 (1H, m) 12 29.8 24.3 25.0 2.25 (1H, m) 1.28 (1H, m) 1.66 (1H, d, 13.2) 1.55 (1H, m) 1.86 (1H, m) 1.63 (1H, m) 13 54.6 52.6 52.5 – – – 14 81.9 80.9 81.0 – – – 15 32.5 30.0 29.9 1.66 (1H, m) 1.59 (1H, m) 1.92 (2H, m) 1.96–1.85 (2H, m) 16 38.0 33.1 33.1 2.72 (1H, m) 1.45 (1H, m) 2.44 (1H, ddd, 18.9, 7.6, 4.1) 2.33 (1H, d , 18.8, 8.8) 2.35–2.46 (2H, m) 17 87.8 218.0 218.5 – – – 18 20.4 17.9 18.1 1.07 (3H, s) 1.01 (3H, s) 1.05 (3H, s) 19 16.9 17.8 19.2 1.42 (3H, s) 1.45 (3H, s) 1.27 (3H, s) 20 79.2 – – – – – 21 19.8 – – 1.42 (3H, s) – – 22 79.7 – – 4.88 (1H, dd, 11.8, 5.3) – – 23 34.4 – – 2.51 (2H, m) – – 24 150.8 – – – – – 25 121.6 – – – – – 26 166.0 – – – – – 27 12.5 – – 1.88 (3H, s) – – 28 20.8 – – 1.94 (3H, s) – – 14-OH n.d. 1.41 (1H, b s) 1.41 (1H, b s) 4-OH n.d. 2.57 (1H, d, 2.50) - aAppa en dsex . See Figu e S19 (Suppo ing In o ma ion File 1) o de ails. n.d. no de ec ed. ing sys em. Again, a ca bon wi h a dis inc down ield shi o 218.4 ppm was ound, demons a ing he p esence o a ke one a C-17. Highly unusually, H-4β appea ed as a double o sex e s (1:5:10:10:5:1) (dsex ) in he 1H NMR spec um. This mul iple was explained as “dddddd” by a o al o six COSY co ela ions (Figu e S19, Suppo ing In o ma ion File 1). The esul ing compound was named i inan B (3, Figu e 1). A com- pound o pu a i ely iden ical s uc u e was isola ed om P. pe u iana be o e, bu only agmen a y physicochemical da a has been epo ed so a [14]. To elucida e he ela i e s e eochemis y o i inans A (2) and B (3), we analysed NOESY da a (Figu e 2). In he case o i inan A (2), he β con igu a ion o OH-4 was deduced by he NOESY co ela ion OH-4/CH3-19. OH-14 was assigned as α based on he co ela ions OH-14/H-12α and H-9/H-12α. The 5,6-epoxide was de e mined as β by a co ela ion om H-6 o H-3. These assignmen s a e in comple e ag eemen wi h he ela i e s e eo- chemis y o 4ß-hyd oxywi hanolide E (1). In i inan B, he con- igu a ion o OH-14 could no be unambiguously in e ed om NOE da a due o he signal o e lap o H-15 wi h H-12 and Beils ein J. O g. Chem. 2019, 15, 2003–2012. 2006 Figu e 2: Key NMR co ela ions. (A) COSY and HMBC co ela ions o i inan A (2). (B) COSY and HMBC co ela ions o i inan B (3). (C) Key NOESY co ela ions o i inan A (2). (D) Key NOESY co ela ions o i inan B (3). o he p o ons. As an al e na i e, OH-14 α con igu a ion was deduced om he chemical shi s o C-12 and C-9, which expe- ience a s ong shielding γ-gauche e ec o OH-14α con igu a- ions [21]. These da a indica e a ela i e s e eochemis y o i inan B ma ching wi hanolide F (5). I inans ep esen highly unusual wi hanolide de i a i es, as hey lack he side-chain lac one ing ha is a common s uc- u al ea u e o i ually all known wi hanolides [3], bu pos- sess an and os ane backbone ins ead. While and os anes such as and os e one (7) a e well-known human sex ho mones (Figu e 3A) [22], hei occu ence in plan s is a e [23-26]. Only a single wi hanolide and os ane has been ully cha ac- e ised be o e, cinedione (8), isola ed om Physalis cine as- cens (Figu e 3A) [23]. We p opose he name and owi hano- lides o his wi hanolide subclass, which so a appea s o be cha ac e is ic o Physalis species. The biosyn hesis o and os anes in mammals equi es h ee enzyma ic s eps s a ing om choles e ol (9, Figu e 3B) [27]. Choles e ol (9) is con e ed o p egnenolone (10) by he cy och ome P450 choles e ol side-chain clea age enzyme (P450scc), which clea es he C20–C22 bond [27]. Then, he bi unc ional P450c17 ac s as a 17α-hyd oxylase and 17,20-lyase o gi e ise o and os anes [27]. Rela ed enzymes ha e no been epo ed om plan s. We sea ched ansc ip ome da a o P. pe u iana o pu a i e homologues o hese enzymes [28]. The bes hi s only had amino acid sequence iden i ies o 22–28%, indica ing ha no P450 enzymes o hese clans exis in P. pe u iana. Al hough enzymes wi h simila ca aly ic ac i i y migh ha e e ol ed con e gen ly in plan s, he di e en subs i- u ion pa e n in he side chain sugges s ha a side-chain clea age mechanism dis inc om mammals is in ol ed. While he o de o oxida i e s eps in wi hanolide biosyn hesis is s ill comple ely elusi e [29], we p opose ha his agmen a ion occu s a a la e s age, when mos ypical wi hanolide unc ional- isa ions ha e al eady been in oduced. Indeed, i inan A (2), i inan B (3) and cinedione (8) can be di ec ly linked o he known wi hanolides 4ß-hyd oxywi hanolide E (1), wi hanolide F (5) and wi hanolide S [23], espec i ely (Figu e S20, Suppo - ing In o ma ion File 1). I he agmen a ion occu ed ea ly in he biosyn hesis, his would imply ha se e al biosyn he ic en- zymes ha e o ole a e subs a es wi hou he lac one side chain. We he e o e p opose ha he side-chain clea age enzyme in wi hanolide biosyn hesis ac s a a la e s age, using common pa hway end p oduc s such as 4ß-hyd oxywi hanolide E (1) as i s subs a es. Two mechanisms a e concei able o his ans- o ma ion (Figu e 3C): A non-oxida i e G ob agmen a ion could make use o a push–pull mechanism be ween C-17 and C-22, building on acid–base ca alysis. Al e na i ely, an en- zyme could clea e he C17–C20 diol oxida i ely. Se e al P450 enzymes ha e been epo ed o be capable o clea ing diols, p esumably ia a e ic pe oxo in e media e (Figu e 3C) [30,31]. Beils ein J. O g. Chem. 2019, 15, 2003–2012. 2007 Figu e 3: S uc u es and biosyn hesis o and os anes. (A) And os ane backbone and and os e one (7) as a ypical mammalian sex ho mone. Cine- dione (8) is he only o he ully cha ac e ised and owi hanolide known. (B) Biosyn hesis o and os anes in mammals. (C) Possible clea age mecha- nisms in ol ed in and owi hanolide biosyn hesis in plan s. To gain u he insigh s in o he biosyn he ic ou e and o exclude ha and owi hanolides a e isola ion a e ac s [32], we exposed 4ß-hyd oxywi hanolide E (1) as he likely p ecu so o i inan A (2) o a ious chemical condi ions (Figu e 4). In gene al, 1 was s able in all sol en s es ed, namely chlo o o m, me hanol, DMSO and ace oni ile (da a no shown). T ea men wi h acid a pH 3 caused no eac ion a all when hea ing up o 70 °C (Figu e 4A). A pH 0, se e al uniden i ied compounds appea ed, bu no i inan A (2). In basic condi ions, only a single uniden i ied p oduc was o med a pH 11 and 70 °C. Nex , we es ed whe he 1 could be oxida i ely clea ed [33]. Incuba ion o 1 wi h NaIO4 a oom empe a u e did no esul in any eac- ion (da a no shown). Howe e , al hough i has been epo ed ha pe ioda es a e no capable o clea ing di e ia y glycols Beils ein J. O g. Chem. 2019, 15, 2003–2012. 2008 Figu e 4: In insic eac i i y o 4ß-hyd oxywi hanolide E (1) unde acidic/basic and oxida i e condi ions, espec i ely. (A) LC–MS ch oma og ams (ELS de ec ion) o 1 incuba ed a di e en pH alues. (B) LC–MS ch oma og ams (UV de ec ion a 200–400 nm) o 1 ea ed wi h di e en oxida i e eagen s. The o ma ion o 2 in he NaIO4 eac ion was con i med by NMR analysis. See also Figu e S21 in Suppo ing In o ma ion File 1 o ex ac ed ion ch oma og ams (EICs). Table 2: An ip oli e a i e ac i i ies in di e en cell lines. Da a indica e EC50 alues ± SD in µM. A549 = human lung ca cinoma; L929 = mouse ib o- blas ; KB-3-1 = human ce ix ca cinoma; MCF-7 = human b eas cance cell line. Compound A549 L929 KB-3-1 MCF-7 4ß-hyd oxywi hanolide E (1) 3.74 ± 0.50 0.27 ± 0.30 1.11 ± 0.98 10.65 ± 6.18 i inan A (2)a5.01 ± 5.27 2.29 ± 0.88 4.62 ± 5.76 17.88 ± 7.27 i inan B (3)b3.45 ± 1.91 1.68 ± 1.78 2.40 ± 2.32 13.56 ± 9.18 s au ospo ine (posi i e con ol) 1.19 ± 0.99 <0.003 0.04 ± 0.01 0.16 ± 0.02 au ano in (posi i e con ol) >7.03 2.35 ± 0.83 1.59 ± 0.37 2.06 ± 0.60 aEs ima ed 90% pu i y based on 1H NMR. bEs ima ed 80% pu i y based on 1H NMR. [33,34], we no ed o ma ion o small quan i ies o i inan A (2) when pe o ming he eac ion a 70 °C (Figu e 4B and Figu e S21 in Suppo ing In o ma ion File 1). The iden i y o i inan A (2) was e i ied by isola ion o he co esponding compound by p epa a i e HPLC (4% yield) ollowed by NMR analysis. This esul con i ms ou NMR-based s e eochemical assignmen and unambiguously links i inan A (2) o 4ß-hyd oxywi hanolide E (1). We also pe o med an oxida i e clea age eac ion wi h ca - aly ic amoun s o MoO2(acac)2 in DMSO as desc ibed by Ga cía e al. [34], which also led o he o ma ion o ace amoun s o i inan A (2). Ou expe imen s sugges ha i inan A (2) and mos likely all and owi hanolides a e no isola ion a e- ac s bu ue na u al p oduc s, which equi e an oxida i e en- zyme o acili a e he C–C bond clea age. Fu u e s udies will shed ligh on he enzyma ic basis o and owi hanolide o ma- ion. Conside ing he po en bioac i i ies o and os anes as well as wi hanolides, we wonde ed whe he he loss o he side-chain lac one would nega i ely impac he an ip oli e a i e ac i i y. I inan A (2) and B (3) oge he wi h 4ß-hyd oxywi hanolide E (1) as a posi i e con ol we e e alua ed agains a panel o ou cell lines (Table 2). In ou assays we obse ed dec easing ac i - i ies du ing he hi d and ou h eplica es, esul ing in la ge s anda d de ia ions and po en ially indica ing limi ed s abili y o hese compounds. None heless, EC50 alues o 4ß-hyd oxy- wi hanolide E (1) we e in good ag eemen wi h p e iously published alues [14,35,36]. I inans A (2) and B (3) we e 1.3 o 10- old less ac i e han 4ß-hyd oxywi hanolide E (1), wi h he excep ion o i inan B (3) in A549 cells, which was equipo en . Howe e , i inan A (2) and B (3) samples had a pu i y o 90% and 80%, espec i ely. We he e o e canno exclude ha uniden i ied impu i ies, which could no be emo ed by epea ed p epa a i e HPLC, obscu e he ue EC50 alues o i inans. We conclude ha i inans possess po en an ip oli e a- i e ac i i y, ha is howe e educed compa ed o 4ß-hyd oxy- wi hanolide E (1). Ou esul s demons a e he impo ance o he lac one side chain o bioac i i y. Conclusion We ha e disco e ed and cha ac e ised i inans A and B, wo new wi hanolides om P. pe u iana wi h unca ed back- bones. They esemble mammalian sex ho mones o he and os ane class. The ela i e s e eochemis y was elucida ed based on NOESY analysis. Chemical s udies suppo a model ha hese compounds a e o med by an oxida i e p ocess. We Beils ein J. O g. Chem. 2019, 15, 2003–2012. 2009 p opose he name and owi hanolides o his wi hanolide subclass. Expe imen al Gene al expe imen al p ocedu es Seeds o Physalis pe u iana we e ob ained om Flo aSel , Spe li and Quedlinbu ge Saa gu . Plan s we e ini ially g own in seed s a e soil (Kölle’s Bes e Anzuch e de) and la e ans- e ed o po ing soil (Kölle’s Bes e P lanze de). Plan s we e g own unde LED illumina ion (SANligh S2W) a 350 µmol s−1 m−2 PPFD wi h a 12 h pho ope iod and a 18–25 °C wi hou empe a u e and humidi y con ol. Plan s we e wa e ed wice pe week wi h ap wa e as needed. NMR spec a we e eco ded using B uke AscendTM 400 o DRX 500 MHz spec ome e s ope a ing a 400 and 500 MHz o 1H NMR and a 100 and 125 MHz o 13C NMR whe e CDCl3 was used as sol en . Chemical shi s we e e e enced ela i e o he esidual sol en signal o CDCl3 (δH = 7.26 ppm, δC = 77.16 ppm) and exp essed in δ alues (ppm), wi h cou- pling cons an s epo ed in Hz. Analysis was conduc ed wi h TopSpin (Ve sion 4.0.6, B uke ). ATR-IR analysis was pe - o med o he ange o 400–4000 cm−1 using a Shimadzu IRA ini y 1S spec ome e wi h samples dissol ed in chlo o- o m. Op ical o a ions we e measu ed wi h a Pe kin Elme 341 pola ime e . Using me hanol as sol en , he wa eleng h o maximum abso p ion was de e mined on a Jasco V-630 spec- opho ome e . Flash pu i ica ion was pe o med on a Bio age Isole a One using columns desc ibed below. HRMS measu e- men s we e ca ied ou on a Wa e s Alliance 2695 HPLC coupled o a Mic omass LCT P emie mass spec ome e . Fo analy ical and p epa a i e LC–MS a Wa e s ins umen was used consis ing o a Wa e s 2767 au osample , Wa e s 2545 pump sys em, Wa e s 2998 diode a ay de ec o , Wa e s 2424 ELS de ec o , and a Wa e s SQ De ec o 2 o mass spec ome- y in ESI+ and ESI– modes be ween m/z 150 and 1000. In ana- ly ical mode, a Phenomenex Kine ex column (2.6 µm, C18, 100 Å, 4.6 × 100 mm) was used wi h a g adien o [sol en A: H2O + 0.05% o mic acid; sol en B: ace oni ile + 0.045% o mic acid; g adien : 10% o 90% B o e 10 min, 1 mL/min]. Samples we e dissol ed o a concen a ion o 10 mg/mL in MeOH and 20 µL injec ed. In p epa a i e mode, a Phenomenex Kine ex Axia column (5 µm, C18, 100 Å, 21.2 × 250 mm) equipped wi h a Phenomenex Secu i y Gua d p ecolumn (Luna, C5, 300 Å) was used in combina ion wi h he sepa a ion g adien desc ibed below. Ex ac ion and isola ion o wi hanolides 140 g o 9 weeks old, whole Physalis pe u iana plan s we e ozen in liquid ni ogen and g ound o a ine powde . The powde was ex ac ed wi h 500 mL H2O/MeOH (3:1) a oom empe a u e o 3 h. A e il a ion and e apo a ion o he sol- en unde educed p essu e, he c ude ex ac was esuspended in 300 mL H2O and de a ed wi h 300 mL pe oleum e he . The emaining aqueous laye was u he ex ac ed wi h 2 × 300 mL CHCl3 ollowed by 2 × 300 mL n-BuOH. This esul ed in a 660 mg pe oleum e he ac ion, 386 mg CHCl3 ac ion and 1174 mg n-BuOH ac ion. The CHCl3 ac ion was sepa a ed ia e e sed-phase lash ch oma og aphy (Bio age SNAP KP-C18-HS 30 g column) wi h a H2O/MeOH g adien . Samples we e adso bed on o Celi e unde educed p essu e o d y loading. A g adien om 30% o 95% MeOH was used. F ac ions we e pooled guided by UV maximum abso bance o o m main ac ion F1 (sub ac ions 1–21, 102 mg), F2 (sub ac ions 22–31, 12 mg) and F3 (sub ac ions 32–45, 48 mg). No wi hanolides we e de ec ed in F2 based on LC–MS analysis and he e o e disca ded. F ac ion F1 was u he sepa a ed by p epa a i e LC–MS. The sample was dissol ed in MeOH o a concen a ion o 15 mg/mL. 100 µL was injec ed pe un. A sepa a ion g adien was used [sol en A: H2O + 0.05% o mic acid; sol en B: ace oni ile + 0.045% o mic acid; g adien : 10% o 90% B o e 10 min, 20 mL/min]. The pos -column low was spli (100:1) and he mino i y low made up o 1 mL/min wi h MeOH + 0.045% o mic acid o in-line analysis by UV, ELSD and MS. The majo i y low was collec ed. The ollowing peaks we e collec ed and iden i ied by NMR: R = 5.8–6.0 min (i inan A (2), 6 mg); 7.2–8.0 min (4β-hyd oxywi hanolide E (1), 49 mg); 9.0–9.2 min (i inan B (3), 1 mg). The collec ed ac- ions we e e apo a ed unde educed p essu e using a Ch is RVC 2-25 CDplus o a ional acuum concen a o . Main ac ion F3 was also sepa a ed by p epa a i e LC–MS as desc ibed abo e, yielding he known compounds wi hano- lide E (4) ( R = 7.3–7.6 min, 6 mg), pe ulac one H (6) ( R = 7.6–7.8 min, 9 mg) and wi hanolide F (5) ( R = 7.8–8.5 min, 9 mg) which we e iden i ied by NMR [20]. Analy ical da a 4β-Hyd oxywi hanolide E (1) was isola ed as a whi e c ys- alline powde . NMR da a o 1 is lis ed in Table 1. All spec os- copic p ope ies ma ched li e a u e da a [20]. I inan A (2): whi e c ys alline powde ; [α]D20 +10.48 (β = 0.62; MeOH); UV (MeOH) λmax (log ε) 239 nm (3.93); IR (ATR, CHCl3) νmax: 3460, 2967, 2930, 1734, 1674, 1454, 1373, 1092, 1036, 986, 922, 754 cm−1; o 1H and 13C da a see Table 1; HRESIMS m/z: [M + Na]+ calcd o C19H24O5Na+, 355.1516; ound, 355.1519. Beils ein J. O g. Chem. 2019, 15, 2003–2012. 2010 I inan B (3): whi e c ys alline powde ; [α]D20 –10.00 (β = 0.06; MeOH); UV (MeOH) λmax (log ε) 251 nm (3.94); IR (ATR, CHCl3) νmax: 3402, 2955, 2930, 1682, 1383, 1259, 1215, 1136, 1088, 1016, 966, 806, 748 cm−1; o 1H and 13C da a see Table 1; HRESIMS m/z: [M + Na]+ calcd o C19H24O3Na+, 323.1618; ound, 323.1626 BLAST sea ch o known and os ane biosyn hesis enzymes The known and os ane biosyn hesis enzymes Homo sapiens P450scc (UniP o KB accession P05108) and Homo sapiens P450c17 (P05093) we e used o sea ch epo ed Physalis pe u- iana ansc ip ome da a [28] ia he BLASTn algo i hm. Bo h enzymes yielded se e al ull-leng h hi s wi h amino acid se- quence iden i ies o 22–28%. Oxida i e clea age o 4β-hyd oxywi hanolide E (1) o i inan A (2) by NaIO4 57.8 mg o NaIO4 (270.2 µmol, 7.0 equi ) in 400 µL ho H2O was added o 19.4 mg 4ß-hyd oxywi hanolide E (1, 38.6 µmol, 1.0 equi ) in 1 mL MeOH. The eac ion was incuba ed a 70 °C o 72 h in a hea block wi h shaking a 1000 pm. A e ha ime a peak wi h m/z 315 co esponding o [M + H − H2O]+ wi h a e en ion ime o 3.5 min was obse ed by LC–MS, co-elu ing wi h au hen ic i inan A (2). The eac ion mix u e was sepa a ed by p epa a i e LC–MS as desc ibed abo e o gi e a whi e c ys alline powde (0.5 mg, 4%), which was con i med o be i inan A (2) by 1H NMR spec oscopy. Oxida i e clea age o 4β-hyd oxywi hanolide E (1) o i inan A (2) by MoO2(acac)2 2 µL o a MoO2(acac)2 s ock solu ion in DMSO (100 µg/µL, 0.6 µmol, 0.02 equi ) was added o 14.6 mg hyd oxywi hano- lide E (1, 29.0 µmol, 1.0 equi ) in 100 µL DMSO. The eac ion was incuba ed a 130 °C o 3 h in an oil ba h wi h s i ing a 400 pm. A e ha ime a peak wi h m/z 315 co esponding o [M + H − H2O]+ wi h a e en ion ime o 3.5 min was obse ed, co-elu ing wi h au hen ic i inan A (2). H2O (5 mL) and CHCl3 (5 mL) we e added o he eac ion mix u e. The laye s we e sepa a ed and he aqueous phase was ex ac ed wi h chlo o o m (3 × 5 mL). Combined o ganic laye s we e washed wi h wa e (5 mL), d ied o e MgSO4, il e ed and he sol en was e apo a ed unde educed p essu e. The esul ing c ude eac- ion p oduc was hen analysed by analy ical HPLC as de- sc ibed abo e. An ip oli e a i e assays The e ec o compounds on cell iabili y was p obed wi h a WST-1 es using he p ocedu e o Ishiyama e al. [37] as modi- ied by Sasse e al. [38]. The ollowing cell lines we e used: mouse ib oblas cell line L929 (DSM ACC 2), human ce ix ca cinoma cell line KB-3-1 (DSM ACC 158), he human lung ca cinoma cell line A549 (DSMZ ACC 107) and human b eas cance cell line MCF-7 (DSM ACC 115). The subcon luen cells we e b ie ly washed wi h Ea le’s Balanced Sal Solu ion (Gibco) wi hou Ca and Mg, ypsinized and e-suspended in Dulbecco’s modi ied eagle’s medium ha con ained 5% e al bo ine se um (FBS; L929, KB-3-1, A549) o Roswell Pa k Memo ial Ins i u e medium ha con ained 5% FBS, 0.5% Minimum Essen ial Medium Non-Essen ial Amino Acids, Gibco (MEM NEAA), 0.5% Glu aMAX (Gibco) and insulin a 5 μg/mL (MCF-7). 25 µL o se ial dilu ions o he es com- pounds (64–0.06 µg/mL, ha we e made wi h a pipe ing obo (epMo ion, Eppendo , Hambu g, Ge many), we e added o 25 μL aliquo s o a cell suspension (1500 cells o KB-3-1, L929 and A549, 3000 cells o MCF-7) in 384 well mic o i e pla es. Blank and sol en con ols we e incuba ed unde iden- ical condi ions. A e an incuba ion pe iod o 5 days, 3 μL WST-1 ( eady o use solu ion by Roche) was added. The incu- ba ion ime o he pla es a 37 °C a ied be ween he cell lines om 20 min o KB-3-1 and A549, L929 o 30 min, and 2 h o MCF-7 be o e measu ing abso bance a 450 nm ( e e ence 600 nm) wi h an In ini e 200 PRO pla e eade (Tecan, Männedo , Swi ze land). As posi i e con ol compounds, Au ano in and S au ospo in we e applied. The abso bance o he sol en con ol was se o 100%. The EC50 alues we e de- e mined wi h Sigma Plo . All da a a e a e age alues om ou biological eplica es. Suppo ing In o ma ion Suppo ing In o ma ion File 1 NMR, MS, UV and IR spec a o i inan A (2) and i inan B (3). NMR da a o wi hanolide E (4), wi hanolide F (5) and pe ulac one H (6). [h ps://www.beils ein-jou nals.o g/bjoc/con en / supplemen a y/1860-5397-15-196-S1.pd ] Acknowledgemen s This wo k has been ca ied ou wi hin he amewo k o he SMART BIOTECS alliance be ween he Technische Uni e si ä B aunschweig and he Leibniz Uni e si ä Hanno e . This ini ia i e is suppo ed by he Minis y o Science and Cul u e (MWK) o Lowe Saxony, Ge many. LC–MS (INST 187/621) and NMR (INST 187/686-1) ins u- men s unded by he Deu sche Fo schungsgemeinscha (DFG) we e used. We hank P o . Russell Cox o his suppo and help ul discussions, and Ma cel A nd o p elimina y wo k. We hank D . Jö g Foh e and colleagues o suppo wi h NMR measu emen s and Ka ja Kö ne and colleagues o excellen lab suppo . Beils ein J. O g. Chem. 2019, 15, 2003–2012. 2011 ORCID® iDs Ma k B öns up - h ps://o cid.o g/0000-0002-8971-7045 Jakob F anke - h ps://o cid.o g/0000-0002-7603-6232 Re e ences 1. Rai, M.; Jogee, P. S.; Aga ka , G.; San os, C. A. d. Pha m. Biol. (Abingdon, U. K.) 2016, 54, 189–197. doi:10.3109/13880209.2015.1027778 2. Mau ya, R. J. Pha m. Pha macol. 2010, 62, 153–160. doi:10.1211/jpp.62.02.0001 3. Chen, L.-X.; He, H.; Qiu, F. Na . P od. Rep. 2011, 28, 705–740. doi:10.1039/c0np00045k 4. Fische , G.; He e a, A.; Almanza, P. J. Cape Goosebe y (Physalis Pe u iana L.). In Pos ha es Biology and Technology o T opical and Sub opical F ui s; Yahia, E. M., Ed.; Woodhead Publishing Se ies in Food Science, Technology and Nu i ion; Else ie : Ams e dam, Ne he lands, 2011; pp 374–397. doi:10.1533/9780857092762.374 5. Sang-nge n, M.; Youn, U. J.; Pa k, E.-J.; Kond a yuk, T. P.; Simmons, C. J.; Wall, M. M.; Ru , M.; Lo ch, S. E.; Leong, E.; Pezzu o, J. M.; Chang, L. C. Bioo g. Med. Chem. Le . 2016, 26, 2755–2759. doi:10.1016/j.bmcl.2016.04.077 6. Pa k, E.-J.; Sang-Nge n, M.; Chang, L. C.; Pezzu o, J. M. J. Na . P od. 2019, 82, 492–499. doi:10.1021/acs.jna p od.8b00861 7. Ki son, I.; Ab aham, A.; Se hi, P. D.; Sub amanian, S. S.; Glo e , E. Phy ochemis y 1976, 15, 340–342. doi:10.1016/s0031-9422(00)89029-9 8. F olow, F.; Ray, A. B.; Sahai, M.; Glo e , E.; Go lieb, H. E.; Ki son, I. J. Chem. Soc., Pe kin T ans. 1 1981, 1029–1032. doi:10.1039/p19810001029 9. Saku ai, K.; Ishii, H.; Kobayashi, S.; Iwao, T. Chem. Pha m. Bull. 1976, 24, 1403–1405. doi:10.1248/cpb.24.1403 10.Neogi, P.; Sahai, M.; Ray, A. B. Phy ochemis y 1986, 26, 243–247. doi:10.1016/s0031-9422(00)81520-4 11.Dinan, L. N.; Sa ke , S. D.; Šik, V. Phy ochemis y 1997, 44, 509–512. doi:10.1016/s0031-9422(96)00553-5 12.Fang, S.-T.; Liu, J.-K.; Li, B. S e oids 2012, 77, 36–44. doi:10.1016/j.s e oids.2011.09.011 13.Ahmad, S.; Malik, A.; Yasmin, R.; Ullah, N.; Gul, W.; Khan, P. M.; Nawaz, H. R.; A za, N. Phy ochemis y 1999, 50, 647–651. doi:10.1016/s0031-9422(98)00567-6 14.Lan, Y.-H.; Chang, F.-R.; Pan, M.-J.; Wu, C.-C.; Wu, S.-J.; Chen, S.-L.; Wang, S.-S.; Wu, M.-J.; Wu, Y.-C. Food Chem. 2009, 116, 462–469. doi:10.1016/j. oodchem.2009.02.061 15.Fang, S.-T.; Li, B.; Liu, J.-K. Hel . Chim. Ac a 2009, 92, 1304–1308. doi:10.1002/hlca.200900005 16.Zhang, W.-N.; Tong, W.-Y. Chem. Biodi e si y 2016, 13, 48–65. doi:10.1002/cbd .201400435 17.Xia, G.; Li, Y.; Sun, J.; Wang, L.; Tang, X.; Lin, B.; Kang, N.; Huang, J.; Chen, L.; Qiu, F. S e oids 2016, 115, 136–146. doi:10.1016/j.s e oids.2016.09.002 18.Xu, Y.-m.; Bun ing, D. P.; Liu, M. X.; Banda anayake, H. A.; Guna ilaka, A. A. L. J. Na . P od. 2016, 79, 821–830. doi:10.1021/acs.jna p od.5b00911 19.Chen, L.-X.; Xia, G.-Y.; He, H.; Huang, J.; Qiu, F.; Zi, X.-L. RSC Ad . 2016, 6, 52925–52936. doi:10.1039/c6 a07031k 20.Ozawa, M.; Mo i a, M.; Hi ai, G.; Tamu a, S.; Kawai, M.; Tsuchiya, A.; Oonuma, K.; Ma uoka, K.; Sodeoka, M. ACS Med. Chem. Le . 2013, 4, 730–735. doi:10.1021/ml400144e 21.Zhang, H.; Timme mann, B. N. J. Na . P od. 2016, 79, 732–742. doi:10.1021/acs.jna p od.5b00648 22.Kicman, A. T. B . J. Pha macol. 2008, 154, 502–521. doi:10.1038/bjp.2008.165 23.Maldonado, E.; Al a ado, V. E.; To es, F. R.; Ma ínez, M.; Pé ez-Cas o ena, A. L. Plan a Med. 2005, 71, 548–553. doi:10.1055/s-2005-864157 24.Siddiqui, B. S.; Usmani, S. B.; Begum, S.; Siddiqui, S. Phy ochemis y 1993, 33, 925–928. doi:10.1016/0031-9422(93)85306-c 25.Sanogo, R.; Ge mano, M. P.; de Tommasi, N.; Pizza, C.; Aquino, R. Phy ochemis y 1998, 47, 73–78. doi:10.1016/s0031-9422(97)00477-9 26.Pupo, M. T.; Viei a, P. C.; Fe nandes, J. B.; das G.F. da Sil a, M. F.; Fo, E. R. Phy ochemis y 1997, 45, 1495–1500. doi:10.1016/s0031-9422(97)00167-2 27.Mille , W. L.; Auchus, R. J. Endoc . Re . 2011, 32, 81–151. doi:10.1210/e .2010-0013 28.Fukushima, A.; Nakamu a, M.; Suzuki, H.; Yamazaki, M.; Knoch, E.; Mo i, T.; Umemo o, N.; Mo i a, M.; Hi ai, G.; Sodeoka, M.; Sai o, K. F on . Plan Sci. 2016, 7, 1883. doi:10.3389/ pls.2016.01883 29.Dha , N.; Razdan, S.; Rana, S.; Bha , W. W.; Vishwaka ma, R.; La oo, S. K. F on . Plan Sci. 2015, 6, 1031. doi:10.3389/ pls.2015.01031 30.S ushke ich, N.; MacKenzie, F.; Che keso a, T.; G abo ec, I.; Usano , S.; Pa k, H.-W. P oc. Na l. Acad. Sci. U. S. A. 2011, 108, 10139–10143. doi:10.1073/pnas.1019441108 31.O iz de Mon ellano, P. R. Subs a e Oxida ion by Cy och ome P450 Enzymes. In Cy och ome P450: S uc u e, Mechanism, and Biochemis y; O iz de Mon ellano, P. E., Ed.; Sp inge In e na ional Publishing: Cham, Swi ze land, 2015; pp 111–176. doi:10.1007/978-3-319-12108-6_4 32.Capon, R. J. Na . P od. Rep. 2019. doi:10.1039/c9np00013e 33.Z iely, M.; Goldman, A.; Ki son, I.; Glo e , E. J. Chem. Soc., Pe kin T ans. 1 1986, 229–231. doi:10.1039/p19860000229 34.Ga cía, N.; Rubio-P esa, R.; Ga cía-Ga cía, P.; Fe nández-Rod íguez, M. A.; Ped osa, M. R.; A náiz, F. J.; Sanz, R. G een Chem. 2016, 18, 2335–2340. doi:10.1039/c5gc02862k 35.Yen, C.-Y.; Chiu, C.-C.; Chang, F.-R.; Chen, J. Y.-F.; Hwang, C.-C.; Hseu, Y.-C.; Yang, H.-L.; Lee, A. Y.-L.; Tsai, M.-T.; Guo, Z.-L.; Cheng, Y.-S.; Liu, Y.-C.; Lan, Y.-H.; Chang, Y.-C.; Ko, Y.-C.; Chang, H.-W.; Wu, Y.-C. BMC Cance 2010, 10, 46. doi:10.1186/1471-2407-10-46 36.Xu, Y.-M.; Wije a ne, E. M. K.; Babyak, A. L.; Ma ks, H. R.; B ooks, A. D.; Tewa y, P.; Xuan, L.-J.; Wang, W.-Q.; Saye s, T. J.; Guna ilaka, A. A. L. J. Na . P od. 2017, 80, 1981–1991. doi:10.1021/acs.jna p od.6b01129 37.Ishiyama, M.; Tominaga, H.; Shiga, M.; Sasamo o, K.; Ohku a, Y.; Ueno, K. Biol. Pha m. Bull. 1996, 19, 1518–1520. doi:10.1248/bpb.19.1518 38.Sasse, F.; S einme z, H.; Schupp, T.; Pe e sen, F.; Memme , K.; Ho mann, H.; Heusse , C.; B inkmann, V.; Ma , P. V.; Hö le, G.; Reichenbach, H. J. An ibio . 2002, 55, 543–551. doi:10.7164/an ibio ics.55.543