Phytochemistry, 0031~9422/88$3.00+ 0.00 Pergamon Press zyxwvutsrqpo plc. zyxwvutsrq Vol. 27, No. 7, pp. 2343 2346, 1988. Printed in Great Britain. SHORT REPORTS zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPON TRIXANE ELMIRA C. DERIVATIVES DE RISCALA, CESAR A. N. CATALAN, FROM VIRGINIA TRIXIS PRAESTANS E. SOSA*, ALICIA B. GUTI~ZRREZ* and WERNER HERZ* Facultad de Bioquimica, Quimica y Farmacia, Universidad National de TucumBn, Ayacucho 491, 4000 S.M. de TucumBn, Argentina; *Department of Chemistry, The Florida State University, Tallahassee, FL 32306, U.S.A. zyxwvutsrqpo Institute de Quimica Orghica, (Revised received 27 November 1987) Key zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Word lndexp~ mTrixis praestans; Mutisieae; Compositae; trixanes; sesquiterpenes. Abstract-Chemical Investigation of Trixis praestans trixane type as well as common plant constituents. resulted INTRODUCTION Sesquiterpenes based on the unusual carbon skeleton 1 have been isolated from several Trixis species and appear to be characteristic secondary metabolites of Mutisieae, Subtribe Nassauviinae [l-S]. In the present report we describe our study of Trixis praestans (Velloso) Cabrera which has led to the identification of new representatives of the system, i.e. the pairs of C-12 epimers Za,a’, 2b,b’, 2c,c’ and 2d,d’, as well as other well-known plant constituents. Bohlmann and coworkers [I-S] have used the term ‘isocedrane’ for describing carbon skeleton 1. However, it has been pointed out [6] that this term is inappropriate and we therefore propose the trivial name ‘trixane’ for the basic carbon skeleton. RESULTS AND DISCUSSION The, sets of epimers 2a,a’, 2b,b’, 2c,c’ and 2d,d’ which were difficult to purify (2c,c’ was only seen in admixture with 2a,a’ or 2b,b’) and decomposed easily at room temperature, particularly when traces of acid were present, due to cleavage of the acylated hemiacetal functions, were clearly analogues of two sets of C-12 epimers Zg,g’ and 2b,b’ isolated earlier [2] from T. wrightii and T. in& tin theory a distinction between the two possibilities for 2a,a’-2d,d’ and 2g,g’ might be achieved by observing the effect on the C = 0 signals (near 6 170 for acetate carbonyls, near 6 167 for the angeloyl and senecioyl carbonyls and approx. 6 176 for the isovaleryl and 2-methylbutanoyl carbonyl) by selectively irradiating H-9 or H-14. In our case, smallness of sample and the ease with which 2a,a’-2d,d’ decomposed on standing prevented determination of the 13CNMR spectra and thus resolution of the ambiguity. PHYTO 27:7-AA in isolation of several new sesquiterpene diesters of the Extensive decoupling of the ‘H NMR spectra led to the assignments in Table 1 with, for example in the case of 2a,a’, duplicate signals, each of equal intensity, for H-4 (64.41 and 4.33, t’s J=2.5 Hz) H-9 (5.25 and 5.24, f’s, J =6 Hz), H-12 (5.02 and 4.99 br), H-13 (1.24 and 1.20, Me singlets), H-14 (5.57 and 5.55 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQ d, J = 8 Hz) and H-15 (6.35 and 6.33 s). On the basis of models one would also expect duplicate signals for H-2, H-3a,b and H-10 and while accurate chemical shifts of most of the upfield signals were difficult to determine due to superposition of signals (see Table I) we found that two resonances with widely different shifts corresponded to H-10, one at 62.76 for the epimer (or epimers-see below) with H-9 at 5.22, presumably the epimer with C-12 OH oriented toward H-10, and one near 2.4 for the epimer with H-9 at 5.25. The latter epimer has H-4 at 64.33 and H-3b at 2.89, while the former has H-4 at 4.41 and H-3b at 2.90. The remaining source of uncertainty was the distribution of the two different acyl functions (acetate on the one hand and angelate, isovalerate, 2-methylbutanoate and senecioate on the other) over C-9 and C-14. It is not clear on what basis the German authors [2] preferred placement of the 2-methylbutanoyl group of their two epimer pairs on C-14 and the senecioyloxy, resp. isovaleryloxy group on C-9 instead of the reverse.1 The following observation suggests that in fact 2a,a’-2d,d’ may represent not only C-12 epimeric pairs but also mixtures of C9 and C-14 acetates. Thus, acetylation of 2a,a’ and 2b,b furnished unequal amounts of two C-12 acetates, 2e and 2e’, from 2a,a’, and two, 2f and 2f’, from 2b,b’, which could be separated by HPLC, a circumstance which facilitated analysis of their NMR spectra (Table I). Compounds 2e,e’ and 2f,f’ were at first thought to be pairs of C-12 epimers; however, both acetates from 2a,a’, and both acetates from 2b,b’, exhibited almost identical ‘H NMR spectra with the H-10 signal invariably located near 62.80. Hence the C-12 stereochemistry of the components of each pair 2e,e’ and 2fJ must be assumed to be identical. 2343 2344 Short Reports Table H 2a,a’t 1. ‘H NMR spectra 2W-H of compounds 20’ Za,a’-2f,f (270 MHz, CDCI,) 2e (major)? II 2e’ (minor)+ 2f (major): 1.38 hr zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA d 1.32 br d 1.36 hr d I.28 hr d 1.36 br d 1.28 br d n 1.3 1.31 br d 1.32 br d (11, 1.5) la 2f (minor)$ - 1.3 (11) lb 2 3a 3b 2.7 m -2.15 m -1.5 tn 2.90 dt 2.89 dt (11, 2.5) 4.41 f 4.33 r 4 2.7 m -2.15 m -1.5 m 2.90 dt 2.89 dt 4.39 t 4.31 t 2.71 br dd (1 1, 6) -2.2 m 1.56 br dd (14, 3) 2.47 m 2.7 “, -2.14 M -1.5 m 2.90 dr 2.89 dt 4.39 t 4.31 1 2.71 - 2.2 -1.5 -2.45 4.48 r(2) m m M m 4.38 hr 2.71 m - 2.2 m 1.55 hr dd - 2.44 m 4.48 f 2.71 -2.2 - 1.5 -2.42 m M m m 4.35 br (2.5) 7 8a 8b 9 - 2.48 m -2.4 WI -1.95 m 5.22 t 5.25 - 2.42 m - 2.4 m - 1.95 m 5.24 I 5.25 t - 2.46 m -2.4 m -1.95 m 5.23 t 5.22 f -2.44 m - 2.2 m obsc. 5.27 t -2.45 m -2.2 m obsc 5.25 1 - 2.44 m - 2.2 ,?I obsc. 5.27 1 - 2.42 in 2.82 dd (6, 1.5) 2.78 dd 2.80 dd 2.80 dd t - 2.2 ni obsc. 5.23 t (6) 10 2.76 br d 12 13* 14 15 AC* -2.37 m (6) 5.01 s, 4.98 s 1.24 s, 1.20 s 5.57 d, 5.55 d (8) 6.35 s, 6.33 s 2.10 s, 2.09 s *Intensity 2.75 d 5.02 1.24 5.48 6.32 2.09 -2.38 s, 4.98 .s s, 1.19 s d, 5.45 d s, 6.31 s s, 2.08 ‘i m 2.75 d 5.02 1.25 5.52 6.34 2.08 -2.38 m s, 4.98 s s, 1.20 s d, 5.49 d s, 6.32 s s, 2.07 s 5.88 1.13 5.59 6.37 2.15 2.11 s s 5.98 1.15 5.55 63.5 2.11 2.10 d s s s s s d 5 s s 5.88 s 1.12 s 5.50 d 6.33 s 2.11 2.09 5.96 1.14 5.46 6.35 2.11 2.10 s s d s s s three protons. tAng 6.18 qq (7, 1.5) (H-3’), 2.01 dq (7, 1.5) (H-4’)*, 1.92 br (H-5’)*. f i-Val - 2.25 m (H-2’), - 1.5 m (H-3’), 0.98 d (H-4’. 5’) (6~s’). 0 2-MeBu -2.4 m (H-2’), 1.7 ddq (7) (H-3’). - 1.5 m (H-3’). 1.17 d (H-5’)*, 0.93 f (H-4’)*. /ISen 5.70 br t (1) (H-2’) 2.19 d (1) (H-4)*. 1.93 d (1) (H-5’)*. A plausible biogenetic route which utilizes a guaiane precursor The proposal is supported by guaienes, cyperenes and trixanes c41. to the trixane system is outlined in Scheme. the co-occurence of in a Moschuria species EXPERIMENTAL General. For separation of mixtures Waters HPLC equipment (M45 pump, U6K injector with 2 ml loop and R-401 differential refractometer) was used. The column employed was an ALTEX 2a.a 2b.b’ 2c,C’ Ultrasphere ODS column (5 pm. 10 mm, inner diameter x 25 cm). Retention times were measured from the injection point. Plant material. Aerial parts of Trixis praestans (Velloso) Cabrera were collected in July 1985 at the flowering stage at km. 5, Route 1 I, Departamento Parana, Entre Rios Province, Argentina, by Mr. J. M. Retamal, IPNAYS, Facultad de Ingeneria Quimica, Universidad National de1 Litoral and identified by Ing. J. M. Jozami (voucher no. 280. IPNAYS). Flowers and leaves were processed separately. Exfraction qf T. praestans. Flowers (500 g) were extracted R'. R2 = AC, hg, R3 = H R*. K1 = AC. i-Val, Ra = H R', K' = AC. 2-MeBu. R3 = H 2d,d’ Ri, K’ = AC. Sen. R3 = H 2e.e’ RI, R’ = AC. Ang. R3 = Ac 2f,f’ RI, K2 = AC. i-Val. 2&g’ Ri = Sen. 2h.h’ Ri = i-Val. R3 = AC R3 = 2-MeBu. R3 = H RZ = 2-MeBu. K" Short Reports 2345 9(or 14)-Acetoxy-14 (or 9)-angelyloxy-4(12),-14(15)-diepoxywith 2 x 6 1 of CHCI, at room temp for 7 days to give 26 g of 12-hydroxy-trix-5(15)-ene (2a,a’). The ‘H NMR spectrum (Table extract which was suspended in 350ml of EtOH at 5&55”, 1) showed this to be a 1: 1 mixture of C-12 epimeric alcohols diluted with ,250 ml of H,O and extracted successively with which decomposed easily on standing, mp 78-80” (without hexane (3 x 350 ml) and CHCl, (3 x 300 ml). Evaporation of the recrystallization); IR Y,,,,%cm-‘: 3450, 1720-1740, 1665; MS m/z hexane fraction gave 19.6 g of residue, a portion of which (15 g) (rel. int.): 404 [M]’ (0.7), 386 (0.4), 373 (2.7), 358 (0.3), 305 (10.8), was chromatographed over silica gel (460 g) using hexane and 304 (lOO), 303 (43.8), 276 (7), 260 (31), 244 (9.9), 226 (18.6), 198 increasing amounts of Et,O, all fractions being monitored by (1.8); ‘HNMR spectrum in Table 1. TLC. Fractions with higher R, than fl-amyrin (hexane-EtOAc Acetylation (Ac,O, pyridine, 7 hr) gave solid material which 5:2) were combined to yield 8.7 g of waxy material which was on HPLC (MeOH-H,O 23, flow rate 3 ml/min) showed two not processed further. Fractions with the same zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA R, as /?-amyrin peaks in the ratio 7:l. Collected separately they were identified were combined to give 1.6 g of residue which was saponified with as 2e (R, 40 min, major constituent, 8.8 mg, crystalline solid) and dil. KOH. The unsaponified material (1.13 g) was purified by 2e’ (R, 56 min, minor constituent, also a solid, 2.3 mg); ‘H NMR flash chromatography (florisil, hexane-Et,0 5:2). Reversedspectra in Table 1. phase chromatography (eluting solvent MeOH, flow rate 9(or 14)-Acetoxy-14 (or 9)-isovaleryloxy-4(12)-14(15)-diepoxy3 ml/min) of a small portion (75 mg) afforded lupeol(25 mg), B12-hydroxytrix-5(15)-ene(Zb,b’) and 9(or 14)-acetoxy-14(or 9)-(2amyrin (15 mg), germanicol (18 mg) and n-triacontanol (3 mg). methylbutanoyloxy-4 (12,14 (15-diepoxy-12-hydroxy-5 (15)-ene Lupeol and b-amyrin were identified by mp, ‘HNMR and (2c,c’). The solid material from the flowering parts (no mp taken) coinjection of authentic material; germanicol by mp (178-180”), was mainly a 1:l mixture of C-12 epimers Zb,b’ containing a ‘H NMR, MS, and mp of its acetate (275-276”), n-triacontanol small amount of the 2c,c’ epimers, (NMR analysis), which deby ‘H NMR and MS. Fractions with an R, close to cholesterol composed easily on standing: IRv,,, cm-‘: 3440, 1725-1745, were combined and separated by HPLC to give 0.8 mg of 1660; MS m/z (rel. int): 406 [M]’ (l.l), 388 (0.5). 360 (1.4), 260 isofucosterol, 10 mg of stigmasterol and 16 mg of sitosterol, all (24.9), 244 (5.2), 226 (15.4) 216 (1.3), 198 (1.4). The ‘HNMR being characterized by coinjection of authentic material, CC spectrum of the material from the leaves, mp 6162” (without retention time, mp and ‘H NMR. recrystallization) showed it was a mixture consisting of approxiThe CHCl, extract gave a residue (5.1 g) which was purified mately equal parts of 2b,b’ and 2c,c’. by CC (silica gel, 150 g) using CHCI, and increasing amounts of Et,0 (O-35%). Thirty-three fractions were collected. Fr. 18-22 Acetylation of the epimeric mixture and separation in the manner described for 2a,a’ afforded as major product 11 mg of which showed one major spot were combined and separated by Zf, mp 148-150” (without recrystallization) and as minor proHPLC (MeOH-H,O 2: 1, flow rate 2 ml/min) to give 9 mg of duct 2.8 mg of 2f (solid, but mp not determined. ‘HNMR 2n,a’ as a 1: 1 mixture of C-12 epimers (R, 35 min) and 6 mg of spectra which revealed that these fractions were essentially free Zb,b’ as a 1: 1 mixture of C-12 epimers (R, 43 min) containing a of diacetoxy-(2-methylbutanoyloxy) analogues are listed in small amount of 2c,c’ (NMR analysis). Fractions 23-27 were Table 1. combined and separated by HPLC to give 4.6 mg of Zd,d’ as a 9(or 14)-Acetoxy-14 (or 9)-senecioyloxy-4(12),14(15)-diepoxy1: 1 mixture of epimers. 12-hydroxytrix-5( 15)-ene(M,d’). The ‘H NMR spectrum (Table Extraction of leaves of T. praestans (500 g) and work-up in the 1) showed this to be a 1: 1 mixture of C-12 epimeric alcohols same manner furnished 19.9 g of hexane extract and 6.5 g of which decomposed easily on standing; crystalline solid (no mp CHCI, extract. The hexane extract after further purification as taken); MS m/z (rel. int.): 404 [M]’ (1.3), 386 (1.32), 358 (3), 305 described in above gave the same triterpene and sterol consti(50.3), 287 (lOO), 245 (24), 208 (64). tuents as the flowering parts. The CHCl, extract showed one major spot on TLC (R, 0.42, CHCI,-EtOAc 2:l) and only traces of more polar substances. CC (silica gel, CHCI,-Et,0 21) Acknowledgement-Thanks are due to Mr J. M. Retamal for gave 3.9 g of material which HPLC (MeOH-H,O 2: 1, collections of plant material. Work in Tucumin was supported 2 ml/min) resolved into four peaks in the ratio 3 :4: 52: 42. Peak by grants from Consejo National de Investigaciones Cientificas 3 (R, 29 min) was identified as 2a,a’ (1: 1 mixture of C-12 y TCcnicas, Argentina, and Consejo de Investigaciones de la epimers) and peak 4 as a mixture of approximately equal amUniversidad National de TucumBn. ounts of 2b,b’ and 2c,c’. Short Reports zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIH 2346 1. Bohlmann, F. and Zdero, C. (1979) Chem. Ber. 112, 427. 2. Bohlmann, F. and Zdero, C. (1979) Chem. Eer. 112, 435. 3. Bohlmann, F., Zdero, C., King, R. M. and Robinson, H. (1979) Phpchemistry Singh, P., Jakupovic, J. and Bohlmann, F. (1985) Phytodwmistry 24, 1525. 5. Zdero, C., Bohlmann, F., King, R. M. and Robinson, H. (1986) Phytochemistry 25, 2873. 6. Roberts, J. S. (1981) Tcrptwex Specdist Periodid Reports 4. REFERENCES 10, 99. 18, 855. 0031 9422;88 $3.00 i-O.00 zyxwvutsrq Phytochemistry, Vol. 27, No. 7, pp. 2346 2347. 19RR. :$‘ 1988Pergamon Press pla. Printed in Great Britain. A SESQUITERPENE LACTONE, NEZHUN GOREN, SINTENIN, SEVIL ~KSCJZ and FROM AYHAN ACHZLLEA SINTENISZl ULUBELEN Faculty of Pharmacy, University of Istanbul, Istanbul, Turkey (Received 7 December Key Word Index--Achilles sintenisii; Compositae; sesquiterpene Abstract-The aerial parts of Achilles sintenisii afforded in addition whose structure was elucidated by spectral methods. 1987) lactone; costunolide; flavonoids to known compounds a new costunolide, sintenin INTRODUCTION Achilles species have shown antibacterial Cl], anti-inflammatory [Z], antilarval [3], anti-allergic [4] and antiirritant [S] activities. In this paper Achilles sintenisii Hub.-Mor. (Compositae) was chemically investigated, in addition to flavonoids and terpenoid compounds a new sesquiterpene lactone, sintenin was obtained. Table I. ‘H NMR spectral data of sintenin (400 zyxwvutsrqponmlkjihgfedcba M H z, (‘DCI ,) RESULTS AND DISCUSSION The aerial parts of A. sintenisii afforded stigmasterol, c(amyrin, salvigenin, 6-hydroxyluteolin 6,7,3’,4’-tetramethyl ether and sintenin (I) a new sesquiterpene lactone of the costunolide type. The IR spectrum of 1 showed a “Jlactone band at 1760 cm- ‘, ester bands at 1720 and 1240 cm - I, unsaturation at 1660 cm -- ‘. The high resolution mass spectrum gave a molecular ion peak at m/z 350 indicating a molecular formula C, 9H2606. The ‘H NMR ?Ac \ / AC0 H 1 3 5 6 7 SD 9 11 13 14 15 OAc OAc 5.24 hr dd 5.11 dd 4.72 hi- d 4.16 dli 2.31 ddd 2.575 ddd 5.17 dd 2.73 dq 1.24 .S 1.48 S 1.71 s 2.03 s 2.105 s J (Hz): 1,2/I= 12:1,2x=4: 3r,2/1= 9; 3rr,2x = 5.5; 5.6 b = 8: 6.7 = 12; 8[1,9/1= 10.5: * zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 9/I&=5; 11,7==7; 11.13 = 7. 1