Phyrochemwry, 1976, Vol. 15, pp. 309-311 Pergamon Press. Printed in England VERGATIC ACID, A NEW PENTACYCLIC TRITERPENE FROM SALI/IA VIRGATA AYHAN ULUBELEN and F&XC AYANOCLU Faculty of Pharmacy, University of Istanbul, Turkey zyxwvutsrqponmlkjihgfedcbaZYXWVUTS (Revised receioed 28 June 1975) Key Word Index-S&a uirgata Jacq.; Labiatae; a new pentacyclic triterpene; virgatic acid. AIrstract-A new pentacyclic triterpene acid was isolated from the arial parts of Salvia virgatu and its constitution was established as 3/I-hydroxy-l-oxo-olean-12-en-28-oic acid and named virgatic acid. INTRODUCX’ION Pentacyclic triterpenoids are common in Sulviu species. Ursolic and oleanolic acids were isolated from S. triloba [l-4], S. o# icinalis [3,5-111, S. calycina [3], S. glutinosa [12], S. pratensis and S. sclarea [13]. Micromeric acid was obtained from S. horminum [14], a- and fl-amyrin were found in S. apiana [lS] and S. o~cinalis [16], ana- gadiol was isolated horn S. broussuneti [lfl. This paper describes a new triterpene acid which was obtained from the chloroform extracts of S. virgutu. The compound has been named virgatic acid and its structure has been established. RFSULTS The petrol extract of the aria1 parts of S. uirgata yielded a small amount of a new triterpene together with flavonoids [18]. The chloroform extract of the mark on column chromatography gave the same compound in a high yield. This was named virgatic acid, CsOH460e IR indicated a hydroxyl (345Ocm-‘), a six membered ring carbonyl (1722 cm-‘) and a carboxyl (1697 cm-‘). The NMR spectrum showed methyl singlets at 60.89, 0.96, 1.03 and 1.25 corresponding to seven methyl groups and a hydroxyl proton at 1.72 (DzO exchange). A well divided triplet at 342 showed a hydrogen next to hy- droxyl group and indicated the presence of -CSi(OH)CH2-, a one hydrogen triplet at 546 suggested a Al2 double bond. Acetylation of the compound gave the corresponding monoacetate. The above mentioned data and the MS of the methyl ester acetate of virgatic acid suggested a derivative of an olean-12-en-oic acid. The MS (see Scheme 1) showed diagnostically important peaks at 526 (M+), 262 (a), 203 (c), 133 (r) and 249 (e). The peaks at m/e 262 (a) and 203 (c) showed that the D and E rings were the same as in oleanolic acid. The peak at m/e 203 is especially characteristic for 28-oic acids and seems to have arisen by the loss of the -COOMe group from fragment (a). The base peaks were at m/e 262 and 203. A Retro-Diels- Alder fragmentation pattern is characteristic for A’ 2 pen- tacyclic triterpenes [19]. The small peak at m/e 249 (e), comprising of the D and E rings arose by a fragmen- tation involving one hydrogen transfer and cleavage of an allylically activated band. 0 HO 0 The MS fragmentation pattern restricted the allocation of the hydroxyl and carbonyl groups to rings A and B. The position of the acid group was confnmed by the fact that pentacyclic triterpenic acid methyl esters give in their NMR spectra the carboxy methyl peak at 6 359 or less, if the acid group is at C-28. In the cases where the acid group is at another location the NMR peak appears between 6 3.59 and 3.62. Virgatic acid gave this peak at 6 3.59 in its methyl ester acetate and 6 3.56 in its methyl ester derivative [20]. The Ai2 position of the double bond and the C-17 position ofthe carboxyl group were supported by the formation of a lactone with bro- mine. There are several publications describing the effects of hydroxyl and/or 0x0 groups on the shifts of methyl peaks in NMR spectra [21-241. The shifts in the methyl peaks of virgatic acid and its acetyl derivative were calcu- lated according to the above mentioned papers and they veritled the C-3/l position for the hydroxyl group. The 3&hydroxyl is in accord with biogenetic considerations and is also in agreement with the observed NMR triplet at 6 342. In order to further confirm the 3/I-hydroxyl and C-28 carboxyl groups the Huang-Minlon reduction was performed [25]. The IR spectrum, the R, values and mmp of the resulting compound showed that it was iden- tical to authentic oleanolic acid. NaBH, reduction of the acid and Jones oxidation of the resulting diol caused an enolization (Scheme 2) to give a compound with a &,., at 254 nm indicating that the carbonyl group should be at the C-l position. The formation of the enol ruled out the possibility of ring B having the carbonyl group. The alternative position of the carbonyl group could be at C-2, but in this case the L,,__ of the enol should be around 270nm. Also in 309