Diterpenes from Chrysoma pauciflosculosa: Effects on Florida sandhill species

Phytochemistry, Vol. 34, No. 1, pp.97-105, 1993 Printed in GreatBritain.

DITERPENES

MARIOS

003l-9422/93 $6.00+ 0.00 0 1993Pergamon PressLtd

FROM CHRYSOMA PAUCIFLOSCULOSA: FLORIDA SANDHILL SPECIES

EFFECTS ON

A. MENELAOU, JEFFREYD. WEIDENHAMER,* G. BRUCE WILLIAMSON,? FRANK R. FRONCZEK, HELGA D. FISCHER, LEOVIGILDO QUIJANO~ and NIKOLAUS H. FISCHER$ Departments

of Chemistry

and tBotany,

Louisiana

State University,

Baton

Rouge, LA 70803, U.S.A.

(Received6 July 1992) IN HONOUR

OF PROFESSOR

Key Word Index-Chrysoma pathy.

JEFFREY

HARBORNE’S

SIXTY-FIFTH

paucifosculosa; Asteraceae; Astereae; clerodane

BIRTHDAY

diterpenoids;

allelo-

Abstract-The aerial parts of Chrysoma paucijlosculosa (syn. Solidago paucifosculosa) gave the known diterpcnes, 17oxygrindelic acid, 17-oxogrindelic acid and 17-carboxygrindelic acid, as well as the new diterpene, 17-hydroxy-7a&epoxygrindelic acid. 17-Oxogrindelic acid decomposes in the presence of air giving three products. The structural data of the diterpenes as well as those of degradation products and derivatives were elucidated by chemical and spectroscopic methods. 17-Oxygrindelic acid and 17-oxogrindelic acid were tested for their effect on the germination and radicle growth of three Florida sandhill species and Luctuca satiua. At concentrations of 12 to 48 ppm, 17oxogrindelic acid reduced the germination and radicle growth of Schizachyrium scoparium and Leptochloa dubia, two native sandhill grasses, but had no significant effects on germination and only a slight stimulatory effect on radicle growth of Rudbeckia hirta and Luctuca sativa. Mixtures of 17-oxogrindelic acid with an equimolar mixture of three C. pauciflosculosa sesquiterpenes, (+)-curlone, (+)-sesquiphellatidrene and (-)-cc-trans-bergamontene, did not enhance activity. 17-Oxygrindelic acid was in general less active in reducing germination and growth than 17oxogrindelic acid.

and growth of two native grasses of the Florida sandhill community, little bluestem (Schizachyrium scoparium) and green sprangletop (Leptochloa dubia), the native dicot, blackeyed Susan (Rudbeckia hirta), and commercial lettuce (Lactuca sativa).

INTRODUCTION

We have been testing the hypothesis [l] that allelochemicals produced and released by fire-sensitive members of the Florida scrub community deter the invasion of grasses and pines from the adjacent sandhill community which provide the fuel for frequent surface fires [ 1,2]. As a continuation of the search for allelochemicals from members of the Florida scrub community which affect the germination and growth of native herbs and pines, we investigated C. paucijlosculosa, a common shrub of the Florida scrub with alleged allelopathic potential [3]. Water extracts of fresh leaves of C. paucijlosculosa contained a mixture of diterpene acids, three known derivatives of grindelic acid and a new diterpene epoxide, as well as previously reported sesquiterpenes [4]. The two major diterpene constituents, 17-oxygrindelic acid (1) and 17oxogrindelic acid (2) and their derivatives, and the sesquiterpenes, (+)-curlone, (+)-sesquiphellandrene, and (-)-cc-trans-bergamotene, as well as a diterpene- sesquiterpene mixture were tested for inhibition of germination

RESULTSAND DISCUSlON

Chemical data

The aerial parts of C. pauciflosculosa afforded the known diterpenes oxygrindelic acid (l), oxogrindelic acid (2) and 17-carboxygrindelic acid (9), which were identified as the previously described methyl esters [S, 63. In addition, the new 17-hydroxy-7cr,8a-epoxygrindelic acid (13) was isolated. Diterpenes 1 and 2 were also correlated by allylic oxidation of 1 with MnO, in methylene chloride providing 17-oxogrindelic acid (2), the NMR and mass spectral data of which were in agreement with literature values for the methyl esters [7]. Since 13C NMR data of the three diterpenes had not been previously reported, they are presented in Table 1. Acetylation of 1 with acetic anhydride in pyridine afforded acetate 3 and lactone 4. Diterpene 3 is known and was previously identified as its methyl ester 11 [S]. Compound 4, C,,H,,O,, is a gum with an IR band at 1753 cm-’ indicating the presence of a y-lactone. The

*Permanent address: Department of Chemistry, Ashland University, Ashland, OH 44805, U. S. A. $Permanent address: Departamento de Quimica Organica, Facultad de Ciencias, UNAM, Mexico. §Author to whom correspondence should be addressed. 97

M. A. MENELAOUet al.

RI

Rz

1 2

H H

CH, OH CHO

3

H

CHpOAc

5 6 7 9

CH, H CH3 CH3

CO,CHs CO, CH,

9 10

H CH,

11 12

CH, CH,

13 H 14H 15 CH,

the assigned structure. The formation of the lactone ring under acylation conditions involves a mixed anhydride intermediate which acylates the primary alcohol to form the nine-membered lactone ring [9]. 17-Oxogrindelic acid (2) decomposed in air to give the dicarboxylic acid 9 as the major product and compounds 14 and 16 as minor products. Air oxidations of aldehydes to the corresponding carboxylic acids are well known reactions [lo]. Diacid 9 and its dimethyl ester derivative (5) gave ‘HNMR spectra in agreement with previously reported data for compounds 9 and 5 [6]. 17-Norgrindelic acid (14) and 7,8-dihydro-9-oxo-17-norgrindelic acid (16) represented new compounds which are most probably formed by decarbonylation of oxogrindelic acid (2) via a free radical process [lo]. Formation of 14 and 16 from diacid 9 could be excluded since 9 did not show any signs of decomposition even after extended light exposure. Compound 14, C,,H,,O,, showed IR bands at 3750 and 1707 cn- ’ typical of a carboxylic acid and a carbonyl, respectively. The ‘HNMR spectrum exhibited four methyl signals as well as two olefinic signals at 65.85 (H-7,ddd,J=9.8,4.7,2.5 Hz)and65.54(H-8,ddd,J=9.8, 2.1, 2.1 Hz) coupled to each other (J=9.8 Hz). The mass spectral data confirmed the assigned structure and the position of the double bond. The molecular ion (m/z = 306) was in agreement with the empirical formula of 14 and the RDA fragment (m/z= 182) required that the double bond is at the 7(8)-position. 7(8)-Dihydro-8-oxo17-norgrindelic acid (16), a gum, exhibited an IR band at 1715 cm - ’ indicative of a ketone and/or a carboxylic acid. Its ‘HNMR spectrum exhibited four typical methyl

CHlOH CHO CO2 H COzH CHpOAc CH, OAc: 7a. &I - epoxlde CH> OH. 7a. &I- epoxlde 16 R-H

H H

17 R-CH,

absence of an acetate group (NMR, MS) and the decrease in the polarity of this compound, when compared with oxygrindelic acid, further suggested that 4 is a lactone. Its ‘HNMR spectrum exhibited signals for four methyl groups, as did its precursor 1, and two signals at 64.83 (dt, J= 12.7, 1.5, 1.5 Hz) and 4.71 (d, J= 12.6 Hz) which were assigned to the lactonic C-17 methylene protons. This was supported by the presence of a methylene carbon signal at 671.1 (C-17) and a carbonyl carbon absorption at 6 174.4 (C-15) in the 13C NMR spectrum. The molecular ion (m/z = 3 18) as well as the ion due to a retro-DielsAlder (RDA) fragmentation (m/z= 194) also agreed with

Table 1. 13C NMR spectral data of compounds l-57-11, C

1

1 2

39.4 t 18.6t 41.8 t 33.3 s 42.4 d 24.3 t 133.4d 137.9s 91.6s 40.9 s 27.2 t 32.5 t 81.6s 46.8 t 172.0s 26.9 q 65.6 t 32.8 q 21.8q 16.9q -

3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18. 19 20 OAc(Me) OAc(C0) 15-OMe 17-OMe

2 39.1 t 18.2 t 41.5 t 33.0s 41.5d 26.1 t 160.8d 141.8 s 88.5 s 41.2s 26.3 t 31.5 t 82.9 s 45.8 t 173.4 s 26.8 q 194.7d 32.5 q 21.6q 16.6q

3 39.3 t 18.6t 41.8 t 33.3 s 42.4 d 24.5 t 136.3 d 133.1 s 91.4s 40.9 s 27.3 t 35.5 t 81.5s 47.1 t 171.6s 26.8 q 66.3 t 32.8 q 21.8q 16.8 q 21.1 q 170.7s

4 41.5t 18.4 t 41.8 t 32.9 s 42.0 d 24.0 t 131.9d 135.8s 90.6s 41.2s 28.5 t 31.7t 85.1 s 48.8 t 174.4s 29.5 q 71.1 t 32.7 q 22.2 q 17.oq

“Signals may be interchangeable in any column.

5 37.7 t 18.7 t 41.8 t 33.2 s 41.8 t 24.4 t 137.6d 135.6 s 88.5 s 40.7s 27.9 t 31.8 t 82.1 s 47.2 t 169.9s 27.2 q 172.1 s 32.7 q 22.0 q 16.8q

13 and 16 (50 MHz, CDCl,, CDCl, as int. standard)

7

8

38.6 t 18.8t 42.3 t 33.3 s 42.0 d 24.2 t 130.2 d 139.4s 89.9 s 40.9s 28.0 t 32.6 t 81.9s 47.6 t 172.0 s 27.4 q 65.5 t 32.8 q 22.2 q 16.9q

38.4 t 18.5 t 41.8t 33.2 s 41.4d 26.4 t 159.1 d 143.0s 87.6 s 41.2s 26.5 t 31.6t 83.0 s 46.2 t 172.0s 27.2 q 195.1 q 32.5 q 21.8q 16.8q

51.4q

51.2q

51.2qa 51.7q”

9 38.9 t 18.6 1 41.7t 33.2 s 41.7d 25.1 t 145.3 d 133.0s 89.7 s 41.4s 27.9 t 32.0 t 82.7 s 46.8 t 174.4 s= 27.6 q 172.5 $ 32.6 q 21.8q 16.7q

10

11

13

16

38.4 t 18.7 t 41.8 t 33.2 s 41.5d 24.8 t 141.5d

38.0 t 18.7 t 41.9 t 33.2 s 42.1 d 24.4 t 133.61 134.3 s 89.5 s 40.7 s 27.9 t 32.5 t 81.7s 47.1 t 171.6s” 27.2 q 66.8 t 32.7 q 22.1 q 16.7q 21.1 q 170.8 S= 51.3q -

4O.Of 18.2 t 41.6 t 33.2 s 37.7 d 22.1 t 59.1 d 61:9s 88.7 s

38.1 t 18.6 t 41.7 t 33.6s 45.6d 23.2 t 37.5 t 212.3s 94.5 s 43.7 s 23.5 t 32.0 t 82.8 s 45.8 t 173.9s 23.2 q 33.4 q 22.0 q 15.8q -

88.5 s 41.0s 28.1 t 31.9t 82.4 s 47.1 c 171.9se 27.4 q 172.3s” 32.6 q 21.9q 16.8q 51.3q

26.9 t 31.5 t 81.4s 48.1 t 171.4s 26.3 q 63.4 t 32.6 q 22.1 q 16.3q

Diterpenes from Chrysoma paucijlosculosa absorptions as well as two one-proton signals at 62.96 and 2.28 assigned to two protons (H-7) next to the carbonyl. The i3CNMR spectrum (Table 1) showed the presence of only 19 carbons with one absorption at 6212.3 typical of a non-conjugated ketone (C-8). The placement of the ketone function at C-8 was based on the multiplicity of the two proton signals (H-7) next to the carbonyl. Compounds 15 and 17, which are the respective methyl esters of 14 and 16, were formed by oxidative degradation of methyl oxogrindelate (8). Their structures were established by mass spectral and ‘H NMR spectral comparison with data of the parent compounds. Reaction of a portion of the water-methylene chloride extract containing diterpenes 1 and 9 was reacted with diazomethane to give the methyl esters 5, 6 and 7. Compound 5, the dimethyl ester of 17carboxygrindelic acid (9), and monoester 7 had been previously described as derivatives of 9 isolated from Grindelia robustn [S]. Since the X-ray structure and i3C NMR spectral data of 7 had not been previously reported, they are included in this paper. The position of the carbomethoxy group in 6, either C-15 or C-17, was established by ‘H NMR spectral comparison. Compound 6 is the monomethyl ester derivative from the dicarboxylic acid 9 with a carbomethoxy methyl absorption at 63.71. In 5, the two carbomethoxy methyl signals appeared at 63.63 and 3.73 and compounds 7 and 8 exhibited carbomethoxy methyl absorptions (C-15) at 63.65 and 3.61, respectively. Based on the above chemical shift comparison, the methyl signal at 63.73 in compound 6 must be due to the conjugated methyl ester (C-17). This assignment was supported by comparing the ‘HNMR spectrum of compound 6 with the data for 10, which was obtained, besides compounds 15 and 17, as decomposition product of methyl oxogrindelate (8). In 10, the C-15 ester methyl absorbs at 63.63, which is nearly identical with the carbomethoxy methyl shifts of esters 7 and 8. From the above data it can be concluded that compound 6 must represent the C-17 monoester of diacid 9. Compounds 11and 12 were obtained from the reaction with diazomethane of a chromatographic fraction of the crude plant extract rich in diterpene 3. Compound 11 had been previously described as a derivative of a Grindelia stricta constituent [8]. Compound 12 was also obtained by acetylation of 13 which is most probably a minor component in the crude extract. Signals for epoxide 13 were exhibited in the ‘HNMR spectrum of fractions of the crude extract but it could not be isolated from the crude extract due to losses in chromatographic procedures of this minor constituent. Therefore, it was prepared by reacting the crude water-methylene chloride extract, rich in 1, with m-chloro-peroxybenzoic acid (m-CPBA). The stereochemistry of the epoxide was determined by comparison of its 13CNMR spectrum with that of the epoxide of grindelic acid [S, 71. The chemical shift of C-5 in 13 shifted upfield (A64.6) in comparison to C-5 of 1, thus indicating that the epoxide is cl-orientated [7]. The ‘HNMR spectrum of 12 showed a signal at 63.20 assigned to H-7. The methylene protons at C-17 appeared as AB-doublets at 64.41 and 3.95 (J= 12.1 Hz) which is

99

due to restricted rotation of the acetate moiety, a result that was also observed in acetates 3 and 11. The 7a,8aepoxide group also restricts free rotation of the primary alcohol moiety as evidenced by the two nonequivalent protons at C-17 in 13 which exhibited doublets (J= 12.3 Hz) at 63.83 and 63.78, instead of a broad singlet at 64.13 in 1. X-Ray data of methyl oxygrindelate (7)

The two six-membered rings have the trans-decalin skeleton. They are trans-fused at C-5/C-10 and adopt two different conformations. Ring A has the chair conformation while ring B is distorted to a half chair due to the double bond at C-7/C-8. The furan ring is also in the halfchair conformation The data for methyl oxygrindelate (7) are in very close agreement with those obtained for methyl grindelate [ 111. The hydroxy group O-4 forms an intermolecular hydrogen bond with the furan oxygen O1, having distance 2.813 (3) A. The positional parameters of 7 are listed in Table 2 and the molecular structure is shown in Fig. 1. Further crystallographic data are deposited at the Cambridge Crystallographic Data Centre. Bioassay data

Results of the aqueous solution bioassays of alcohol 1 and aldehyde 2 are summarized in Fig. 2. Germination of Table 2. Positional parameters of methyl 17-oxygrindelate (7) and their estimated standard deviations. Atom

x

Y

Z

B,c#)

01

0.7276 (2) 0.8561(4) 0.6637 (3) 09090(3) 0.7823 (4) 0.7155(6) 0.5673 (5) 0.5870(5) 0.6735 (4) 0.7119(5) 0.7876 (4) 0.8471(4) 0.8482 (3) 0.8140(4) 0.9935 (4) 0.9578 (4) 0.7841(4) 0.6967 (4) 0.749 l(4) 0.7550(5) 0.9130(4) 0.4260 (6) 0.6605 (7) 0.9604 (4) 0.7040 (7)

0.1891(2) 0.4531(2) 0.5186(2) 0.2983 (2) -0.0071(3) -0.1101(3) -0.1150(3) -0.0942(3) 0.0038 (2) 0.0248 (3) 0.1216(3) 0.1732(2) 0.1362(3) 0.0243 (2) 0.1662(3) 0.2641(3) 0.2601(3) 0.3537 (3) O&42(3) 0.2250 (3) 0.2726 (3) -0.0816(3) -0.1816(3) -0.0303 (3) 0.6126(3)

0.6232(l) 0.7463 (2) 0.6809(2) 0.3935 (1) 0.6707(2) 0.6767(3) 0.6312(3) 0.5415(3) 0.5303 (2) 0.4427 (2) 0.4336(2) 0.4934(2) 0.5796(2) 0.5839 (2) 0.6262 (2) 0.6631(2) 0.6805 (2) 0.6617 (2) 0.7021(2) 0.7648 (2) 0.4763 (2) 0.5057(3) 0.4999(3) 0.5545(3) 0.7139(3)

3.66 (4) 9.53 (9) 6.74 (7) 5.00 (6) 5.03 (9) 7.0(l) 6.6(l) 5.23 (9) 3.95 (7) 4.90 (9) 4.16(7)

02 03 04 Cl c2 c3 c4 c5 C6 c7 C8 c9 Cl0 Cl1 Cl2 Cl3 Cl4 Cl5 Cl6 Cl7 Cl8 Cl9 c20 c21

3.60(7) 3.60(7) 3.91(7) 4.44 (8) 4.66 (8) 4.09 (8) 4.84(8) 5.51(9) 5.9(l) 4.65 (8) 8.0(l) 7.9(l) 5.71(l) 8.9(l)

The equivalent isotropic thermal parameter, for atoms defined anisotropically, is defined by the equation: B,, =4/3[a2B 1, +b2B2,+cZB,,+abB,, cos y+acB,, cos fl+bcB,, cos a].

100

M. A. MENELAOU et al.

Fig. 1. The molecular structure of methyl 17-oxygrindelate (7). R. hirta and L. satioa was minimally affected by the two diterpenes. Significant reductions of germination were observed with both sandhill grasses, with the aldehyde 2 being slightly more active. The higher activity of the aldehyde was expected due to its greater alkylating power of the a&unsaturated aldehyde. 17-Oxogrindelic acid (2) also had a greater effect on radicle elongation of the grasses, reducing radicle growth 35-40% at 48 ppm. Surprisingly, the aldehyde 2 had no effect on R. hirta and a stimulatory effect on lettuce radicle elongation, while the alcohol 1 did show activity, reducing lettuce radicle growth 15% and R. hirta growth 41% at 120 ppm. The low activity observed with lettuce confirms earlier observations with allelochemicals obtained from other scrub species that lettuce is less sensitive to those compounds than are the native sandhill species. Significant activity of 17-oxogrindelic acid (2) and its decomposition products appeared in only seven cases from the 100 tests (four species x five solutions x five concentrations). These cases were not common to any test solutions, nor was higher inhibition found at higher concentrations. Therefore, no pattern of biological activity is evident. Comparisons of radicle lengths revealed 11 cases of inhibition and nine of stimulation, but all the itimulation occurred in lettuce, whereas inhibitions were found in three native species. The aldehyde 2 reduced radicle growth of L. dubia to 60% of the control at 27 and 14 ppm, whereas the dicarboxylic acid 9 at 95 ppm and the decomposition mixture at 86 ppm reduced radicle growth of R. hirta to 70% and S. scoparium to 90% of their control (data not shown). At the concentrations tested, the C. pauc$osculosa sesquiterpenes had only minor effects on the germination

and radicle growth of the four target species [4]. Schizachyrium radicle growth was significantly stimulated by 10m4 M solutions of (+)-curlone and (-)-Etrans-bergamotene but Leptochloa was not affected by any of the three sesquiterpenes. Rudbeckiu radicle grotvth was reduced to about 80% of the control by 10m4 M solutions of (+)-curlone and (-)-m-transbergamotene. Lettuce germination was inhibited by aqueous saturated (65% of control) as well as 10m4 and 10m5 M solutions (49% and 77% of control) of (+)curlone, and was stimulated significantly by a 10e4 M solution of (-)-or-trans-bergamotene (data not shown). Bioassays of aqueous solutions of compound 9 as well as the mixture of the decomposition products of (+)epicurlone, which had been obtained by bubbling oxygen through its solution under irradiation, showed no significant effects on the germination or radicle growth of the four test species. This suggested that decomposition products of (+)-curlone including its major decomposition product do not seem to be involved in allelopathic effects of C. pauciflosculosa [4]. In the assays of oxogrindelic acid (2) in combination with the previously obtained [4] sesquiterpene mixture [(+)-curlone. (+)-sesquiphellandrene and (-)a-trans-bergamotene], there was no indication of a synergistic effect between the diterpene 2 and the sesquiterpene mixture. The sesquiterpene mixture alone significantly only inhibited little bluestem germination. In general, only the highest concentration (maximal 167 ppm) of 17-oxygrindelic acid (2) was active in reducing germination and growth. Inhibitory effects were greatest on the two native grasses. At 0.5 mM, the germination of S. scoparium and L. dubia was reduced to 56% and 40% of control, respectively, and radicle growth to 52% and 43% of control, respectively. Aqueous solutions of 2 were somewhat less active (Fig. 2). Little or no increase in activity was observed when the sesquiterpene mixture was added to oxogrindelic acid (2). Strong inhibitory effects were caused by polyacetylenic root constituents of C. pauciJlosculosa [12].

EXPERIMENTAL

Plant matwial. Aerial parts of Chrysoma pa&jlosculo~ (Michx.) Greene were collected in June 1987,2 km west of the entrance of Hwy. 292 in Perdido Key, Florida by G. B. Williamson; voucher No. 70385 deposited at the Louisiana State University Herbarium. Isolation of diterpenes 1 and 2. Fresh leaves (1.5 kg) were extracted as described previously [4]. Both compounds are present in all three extracts, the water extract which was re-extracted with CH,Cl, (termed H,O-CH,Cl, extract), hexane and CH,Cl,. The hexane extract was rich in compound 2 and the CH,Cl, extract in compound 1. The hexane extract (3.46 g after removal of fatty acids) was chromatographed on silica gel (TLC-7GF) using VLC [13] and hexane-CH,Cl,

Diterpenes from Chrysoma pauciflosculosa

101

17-OXOGRINDELIC ACID EFFECTS % OF CONTROL 100

80

80

Q-SC

E-SC

Q-Ld

E-Ld G-Rh E-Rh CONCENTRATION (ppm)

12

i%@i24

n38

G-La

E-L8

-48

17-OXYGRINDELIC ACID EFFECTS % OF CONTROL

80

80

Q-SC

E-SC

m20

G-Ld E-Ld Q-Rh E-Rh CONCENTRATION (ppm)

-40

080

G-La

E-La

120

Fig. 2. Germination (G) and radicle elongation (E-) of Schizachyrium scoparium (-SC), Leptochloa dubia (-I-d), Rudbeckia hirta (-Rh) and Luctuca sativa (-Ls) at four concentrations of 17-oxogrindelic acid (2) and 1‘I-oxygrindelic acid (1). Responses are shown as a per cent of the control germination and radicle elongation in distilled water. Percentages greater than 100% are shown only as 100%. An asterisk at the top of a column indicates statistically significant (P=O.O5) inhibition.

by CH,Cl,-EtOAc mixts of increasing polarity yielding 24 (20 ml) fractions. The combined frs 15 and 16 (148 mg) were chromatographed by prep. TLC (CH,CI,-EtOAc, 7 : 1) to yield 2 (47 mg) and fr. 23 (183 mg) was chromatographed by prep. TLC (hexane-EtOAc, 2: 1; x 2) to give 1 (36 mg). Acetylation. The CH,Cl, extract (44.5 g) was chromatographed on silica gel (TLC-7GF) using VLC (hexane, then CH,Cl, and CH,CI,-EtOAc mixts of increasing polarity). Frs 13 and 14 were combined (5.17 g) and rechromatographed on silica gel yielding mixts followed

171 fractions (22 ml each). Frs 48-66 were combined (0.90 g)

and 510 mg was dissolved in 1.5 ml of pyridine plus 0.5 g of Ac,O and left overnight at room temp. After removal of pyridine by azeotropic evapn with C,Hs, the residue was chromatographed on silica gel (CH,Cl,-EtOAc mixts of increasing polarity) yielding 47 fractions (22 ml each). Frs 4-18 were combined (8 mg) and chromatographed by prep. TLC (CH,Cl,) to yield 6 mg of pure lactone 4. Combined frs 29-36 (21 mg) were chromatographed by prep. TLC (CH,Cl,-EtOAc, 4: 1) to yield 3 (7 mg).

102

M.

A.

MENELAOU et al.

Esterification with diazomethane. The HzO-CHzClz extract (894 mg) in 10 ml of dry Et,0 was reacted for 30 min with a freshly prepared ethereal CH,N, soln. Evapn and chromatography of the residue by VLC on silica gel (TLC-7 GF hexane-CHzCl, mixts of increasing polarity) gave 21 fractions of 20 ml each. Fr. 11 (9 mg) was rechromatographed by prep. TLC (CHzCI,, x 2) to give four bands, the least polar of which contained 5 (1 mg). Prep. TLC (CHzClz-EtOAc, 6 : 1) of fr. 16 (95 mg) gave 3 bands the second of which contained 6 (27 mg). Fr. 17 (66 mg), upon prep. TLC (CH,Cl,-EtOAc, 5 : 1; x 2) provided 4 bands, 1 and 2 of which were combined (21 mg) and rechromatographed by prep. TLC (hexane-EtOAc, 2 : 1, x 3) giving 4 bands, the least polar of which contained pure 7 (5 mg). Part of fr. 16 obtained from the VLC of the hexane extract was methylated with CHzN, and the reaction mixt. chromatographed by prep. TLC (hexane-EtOAc, 3: 1) to yield 8 (7 mg). Compound 8 (11.3 mg) was exposed to air and light to allow for oxidative decomposition. The most polar fraction from the prep. TLC of the decomposition mixture contained 10. A chromatographic fraction of the crude plant extract rich in compound 3, after reaction with CH,N, and subsequent separation by prep. TLC, gave 11 plus a compound exhibiting ‘H NMR signals identical with those of epoxide 12. MnO,-oxidation of 1. VLC of the combined frs 15 and 16 of the CH,Cl, extract exhibited ‘HNMR signals diagnostic of 1. This fraction (0.4 g) was refluxed for 1.5 hr in 40 ml of dry hexane and 1.5 g of activated MnO,. After suction filtration over a bed of celite the solvent was evapd in vacua and the residue chromatographed by prep. TLC (CH,Cl,) yielding 2 (72 mg). When exposed to air in sunlight, aldehyde 2 decomposed to give, after chromatography by prep. TLC (CHzCl,), compounds 9,14,16 and unreacted 2. A mixt. (12 mg) of decomposed 2 was reacted with CH,N, giving, after prep. TLC (CHzCl,, x 2), compounds 15, 17 and 6. Compound 11 (45 mg) in 3 ml of CHzCl,, was reacted with m-CPBA (24 mg) at room temp. for 4 hr, to provide after chromatography pure 12. The HzO-CH,Clz extract (188 mg) in CHzCl, (20 ml), mainly containing 1, was reacted with mCPBA(150 mg) at room temp. for 1 hr. Subsequent VLC and prep. TLC yielded 13. 17-Oxygrindelic acid (1). C,,H,,O,, gum; [a];” -50.8” (CHCl,; c 0.017); IR ~2’; cm-‘: 3416 (OH), 1711 (COOH); EIMS (probe) m/z (rel. int.): 336 [M]’ (O.l), 212 [RDA]+ (lOO), 194 [212-H,O]+ (5), 176 [194-H,O]+, 134(17), 81 (19) 69(20), 55(26),43(23),41 (33). 17-Oxogrindelic acid (2). C,,H,,O,, gum; [a];” -43.9” (CHCI,; c 0.04); IR v”,“,p cm-‘: 1701 (CHO, CO,H); EIMS (probe), m/z (rel. int.): 334 [M]’ (0.4), 210 [RDA]+ (lOO), 182 (lo), 150 (17), 110 (49), 91 (20), 79 (18), 77 (13), 65 (8), 47 (5). 17-Acetoxygrindelic acid (3). C,,H,,O,, gum; [ali -75.7“ (CHCI,; c 0.06); IR vN_“c’cm-‘: 1734 (OCOMe): II ...P^

EIMS (probe), m/z (rel. int.): 3 18 [M - AcOH] + (8), 254 [RDA]+ (38), 194 [254-AcOH]+ (78), 176 Cl94 -H,O]+ (72), 161 [176-15]+ (7), 152(74), 134(65),91 (68), 55 (35), 53 (49), 41 (100). Diterpene &tone (4). C,,H,,O,, gum; [a]:” -61.2” (CHCI,; c 0.02); IR vp$’ cm-‘: 1753; EIMS (probe) m/z (rel. int.): 318 CM]’ (4), 258 (3), 215 (5), 201 (ll), 194 [R,DA]+(9),152(19),109(15),105(18),91(32),79(18),77 (16), 69 (16), 55 (34), 43 (lOO), 41 (24). Methyl 17-carboxymethylgrindekzte (5). C,,H,,O,, 1738 gum; CalA -41.2” (CHCI,; c 0.004); IR vfi:‘cm-‘: (CO,Me), 1722 (conj. COzMe); EIMS (probe) m/z (rel. int.): 378 [M]’ (0.2), 254 [RDA]+ (54), 222 [254 -MeOH]+ (lOO), 190 [222-MeOH]+ (19), 194 [222 -CO]+ (7), 162 [190-CO]’ (8). 148 [164-14]+ (15), 133 [148-Me]+ (4), 105 (16) 91 (27), 69 (25) 59 (33), 55 (25), 43 (44), 41 (29). 17-Carboxymethylgrindelic acid (6). C21H3405, gum; [a]i3-46.5” (CHCI,; ~0.04); IR ~2:’ cm-‘: 3372 (COzH), 1682 (COzH, COzMe); EIMS (probe) m/z (rel. int.): 364 [Ml’ (0.4), 240 [RDA]+ (65), 208 [RDA -MeOH]+ (lOO), 190 (5), 162 (10.5), 115 (13), 91 (36), 79 (20), 70 (19), 59 (24), 43 (73), 41 (36). Methyl 17-oxygrindelate (7). C21H3404, crystals; mp 104-l 10”; [a];” - 76.2” (CHCI,; c 0.008); IR ~2:’ cm- i: 3451 (OH), 1738 (COzMe); EIMS (probe) m/z (rel. int.): 350 [M] + (O.Ol), 226 [RDA] + (36), 208 [226- 18]+ (19), 176 [208-MeOH]+ (lOO), 161 [176- 151’ (3), 148 (19), 135 (47), 134 (84), 105 (29), 91 (43), 81 (64), 79 (36), 55 (68), 43 (64) 41 (56). Methyl 17-oxogrindelate (8). C,,H,,O,, gum; IR YE:’ cm-‘: 1739 (COzMe, CHO); EIMS (probe) m/z (rel. int.): 348 [Ml+ (0.6), 224 [RDA]+ (43), 192 [224-MeOH]+ (7), 150(29), 128 (ll), 123 (25), llO(lOO), 105(22), 97 (50), 91 (36), 77 (27). 69 (39), 55 (41), 43 (34). 17-Carboxygrindelic acid (9). Cz0H3,0,, gum; IR ~2: -l: 3144 (COOH), 1699 (COOH); EIMS (probe) m/z ii. int.): 350 [M]’ (0.5), 226 [RDA]+ (87), 208 [226 -28]+(100), 190[208-18]+(13),91 (15),86(11),84(15), 77 (lo), 67 (lo), 55 (7), 41 (11). Methyl 17-carboxygrindelate (10). C21H3405, gum; EIMS (probe) m/z (rel. int.): 364 [M] + (0.3), 240 [RDA]+ (49.5), 222 [RDAIS]’ (lOO), 190(18), 162 (II), 148 (23), 139(12),121(16.6),105(11),91(19),81(10.5),69(14),55(12). Methyl 17-acetoxygrindelate (11).CZ3HJ605, gum; IR v:~:’ cm- ‘: 1731 (AC, COzMe); EIMS (probe) m/z (rel. int.): 332 [M -AcOH]+ (0.4), 268 [RDA]+ (29), 208 [RDA-AcOH]+ (53.3), 176 [208-MeOH]+ (lOO), 148 (18), 134(69), 119(13), 105(16),91 (28),81 (23),69(16),55 (16), 43 (5). Methyl 17-acetoxy-7a,8a-epoxygrindelate(12). C,,H,,O,, gum; IR ~2:’ cm- ‘: 1741 (AC, COzMe); EIMS (probe) m/z(rel. int.):408 [M]’ (0.2), 348 [M-AcOH]+ (7.6), 333 [348- 151’ (I), 69 (26), 59 (27), 55 (46), 43 (lOO), 41 (19). 7a,8a-Epoxy-oxygrindekc acid (13). C,,H,,O,, gum; EIMS (probe) m/z (rel. int.): 352 [MI’ (I), 322 (22), 321 (lOO), 210 (24), 202 (15), 197 (24), 183 (80), 165 (25), 143 (26), 124 (65), 109 (35), 95 (32), 79 (25), 43 (4). 17-Norgrindelic acid (14). C,,H,,O,, gum; IR vky cm-‘: 3750, 1707 (CO,H); EIMS (probe) m/z (rel. int.):

Diterpenes from Chrysoma paucijosculosa

103

gum; IR v!$tl en- ‘: 1715 (CO,H, GO); EIMS (probe), m/z (rel. int.): 322 [M]’ (41), 185 (63), 169 (lOO), 151 (28), 123 (IS), 107 (19), 95 (24), 67 (36), 55 (42), 43 (40); ‘H NMR in Tables 3 and 4, 13CNMR in Table 1.

306 [M]’ (O.l), 182 [RDA]+ (lOO), 164 [182-18]+ (3.4), 146 (2.5),_l36 (4.3), 105 (5.3), 122 (10.3), 95 (10.5), 91 (9.5), 79 (7), 41 (5). Methyl 17-norgrindelute (15). C&,H,,O,, gum; EIMS (probe), m/z (rel. ht.): 320 [M]’ (0.6), 196 [RDA]+ (lOO), 164 [196-MeOH]+ (13), 149 [164-Me]+ (40), 136(16), 122 (47), 119 (26), 105 (44), 91 (58), 81 (28), 79 (52), 69 (51), 59 (35), 55 (67), 43 (27), 41 (39). 7,8-Dihydro-8-oxo-17-norgrindek acid (16). C19HS004,

Methyl

7,8-dihydro-8-oxo-17-norgrindelic

1

2

3*

5 7 14 14 16 17 17 18 19 20” OAc SOMe 17-OMe

1.7 dd 5.99 dd 2.96 d 2.50 d 1.41 s 4.13 bs 4.13 bs 0.92 s 0.89 s 0.82 s

1.81 dd 7.00 dd 3.00 d 2.41 d 1.42 s 9.37 s

1.72 dd 6.07 dd 2.79 d 2.57 d

lection on an Enraf-Nonius

CAD4 diffractometer equip-

0.94 s 0.92 s 0.79 s

as int.

4*

s

6

7

st

9

ObSC.

obsc. 6.45 dd 2.80 d 264d 1.33 s

1.78 dd 6.81 dd 3.22 d 2.44 d 1.41 s -

1.72 d 5.88dd 2.75 d 2.63 d 1.36 s 4.15 bs 4.15 bs 0.90 s 0.88s 0.80s -

1.89 dd 6.86 dd 2.68 d 2.50 d 1.39 s 9.38 s 0.92s 0.91s 0.76s

1.81dd 7.08 dd 3.00 d 2.58 d 1.40s

3.65s

3.61s

5.87 dd 2.67 d 2.54 d

1.4OS

1.4OS

4.65 d 4.51 d 0.91 s 0.88 s 0.81 s 2.06 s

4.83 dt 4.7ld 0.89 s 0.87 s 0.84 s -

-

0.91 s 0.90 s 0.84 s -

0.90 s 0.88 s 0.83 s 3.63 s 3.73 s

0.91s 0.89s 0.82s -

-

3.71 s

J (Hz): 1: 5=11.7, 5.7, 7=3.6, 3.6, 14=15.2=15.2; 2: 5=11.4, 5.7, 7=4.2, 3.2, 14=15.2, 14’=15.6; 3: 5 =11.7,5.6,7=3.7,3.7,14=15.3,14’=15.3,17=12.7,17’=12.6;4:7=5.2,14=13.1,14’=12.9,17=12.7,1.5, 17’=12.6; 5: 7=4.9, 2.6, 14=14.4, 14’=14.2; 6: 5=11.4, 5.8, 7=3.8, 14=15.4, 14’=15.4; 7: 5=11.7, 5.0, 7 =3.6,3.6,14=13.9,14.0;8:5=11.6,5.6,7=5.4,3.2,11=12.1,9.7,6.3,11’=12.6,9.9,6.4,14=14.2,14’=14.2; 9: 5=11.4, 5.7, 7=4.2, 3.3, 14=15.0, 14’=14.9. TAdditional signals of 8: 62.50 (H-6, dt), 2.19 (H-6’, ddd), 2.54 (Hll, ddd), 1.80 (H-l 1’, ddd). 2.86 (H-12, ddd). Table 4. ‘H NMR

H

10

5 6 6

spectral data of compounds 10-17 (200 MHz, CDCI,; standard) 11

12

1.71 dd -

-

13

CDCI,

as int.

14

15

16

17

-

1.69 dd 2.21 ddd

-

-

5,69ddd 5.48dt 2.66 d 2.54 d 1.26 s 0.91 s 0.88 s 0.86 s -

2.96ddd 2.28 ddd 2.47 d 238 d 1.37 s 0.95 s 0.83 s 0.75 s

2.39 dd -

1.72 ddd

-

7 7 8 14 14 16 17 17’ 18 19 20 OAc

6.81 dd

5.91 dd -

3.20dd

3.52 bd

2.80 d 2.62 d 1.36 s

2.67 bs 2.67 bs 1.33 s 4.61 d 4.51 d 0.88 s 0.85 s 0.78 s 2.05 s

275 s 2.75 s 1.35 s 4.41 d 3.95 d 0.87 s 0.86 s 0.83 s 2.07 s

2.74 d 2.49 d 1.37 s 3.83 dd 3.78 d 0.88 s 0.86s 0.85 s -

5.85 ddd 5.54ddd 2.58 s 2.58 s 1.32 s 0.93 s 0.90 s 0.90 s -

15-OMe 17-OMe

3.63 s -

3.62 s

3.66 s -

0.90 s 0.89 s 0.82 s -

(17).

gum; IR ~22 en-‘:

Table 3. ‘H NMR spectral data of compounds 1-9 (200 MHz and 400 MH.q* CDCI,, CDCl, standard) H

acid

1733 (COzMe). X-Ray data of methyl oxygrindelate (7). A crystal of dimensions 0.08 x 0.30 x 0.45 mm was used for data col-

C,,H,,O,,

-

2.71 s 2.71 s 1.30s 1.02 s 0.88 s 0.87 s -

3.65 s -

-

3.66 s

J (Hz): lo: 7=4.6, 2.9, 14=14.2, 14’=14.3; 11: 5=11.7, 5.1, 7=3.0, 17=12.5, 17’=12.6; 12: 7=2.9, 1.2, 17=12.1, 17’=12.1; 13: 6=15.9, 12.7, 3.1, 7=2.5, 14=13.6, 14’=13.7, 17=12.3, 17 =12.4; 14: 7=9.8, 4.7, 2.5, 8=9.8, 2.1, 2.1; 15: 5=11.8, 5.1, 6=11.8, 11.4, 7.4, 7=9.8, 4.9, 2.3, 8 =9.7,2.0, 14= 14.0, 14’=14.Q l&7=13.4, 13.4,6.9,14= 14.4,14’= 14.6; 17:7= 12.3,5.3,7’= 12.3, 6.5.

104

M.

A.

MENELAOU et

ped with CuKcl radiation (A= 1.54184 A) and a graphite monochromator. Crystal data are: C2iHJ404, M, = 350.50, orthorhombic space group P2,2,2i, a= 8.7677 (13), b= 13.737 (2), c= 14624 (2) A, V=2002.2 (8) A3, 2 =4, d,=1.163 g cmA3, T=24”. Intensity data were measured by w-26 scans of variable rate designed to yield I =25a(I) for all significant reflections. Two octants were collected within the limits 2” 30 (I) and were used in the refinement. The structure was solved by direct methods and refined by full matrix least squares, treating non-hydrogen atoms anisotropically, using the Enraf-Nonius SDP [14]. Hydrogen atoms were located in difference maps and included as fixed contributions. Convergence was achieved with R = 0.05511 and R, =0.06266. The mirror-image structure was refined under identical conditions, yielding R = 0.05523 and R, = 0.06287. The X-ray structure of the former enantiomer is illustrated in Fig. 1, and its coordinates are tabulated in Table 2. Diterpene aqueous solution bioassays. In order to determine whether oxygrindelic acid (1) and oxogrindelic acid (2) were active at maximum concentrations in aq. soln without the aid of solubilization agents, satd aq. solns of each were prepared by sonication for 2 hr. After 24 hr, solns were filtered and the concns of each determined by UV absorbance measurements. Maximum concn of alcohol 1 was 120 ppm, and 48 ppm for aldehyde 2. Three dilutions of the satd soln of each compound were prepared for determination of dose-response relationships. Assays were conducted under the same conditions as before except that 100 x 15 mm plastic Petri dishes sealed with plastic film were used. Assays were conducted in 100 x 15 mm plastic Petri dishes lined with 1 sheet of Whatman No. 1 filter paper and sealed with parafilm. Each dish contained 25 seeds of one of the four test species: two native sandhill grasses, little bluestem cv ‘Cimarron’ Schizachyrium scoparium (Michx.), Nash and green sprangletop, Leptochloa dubia (H.B.K.) Nees and the native aster, blackeyed Susan, Rudbeckia hirta L., and commercial lettuce cv ‘Great Lakes 118’, Lactuca sativa L. Treatments were replicated 6 times for little bluestem and 3 times for the other target species. Assays were carried out in the dark at room temp. (23-25”), and terminated after 3 days for lettuce and 5 days for the other species. Dishes were frozen to terminate growth prior to measurement of radicle length and germination. Seeds were considered to be germinated if the radicle protruded at least 1 mm. 17-Oxogrindelic acid (2) and its decomposition product bioassays. Bioassays of 17-oxogrindelic acid and its decomposition products were performed as above. Solutions of 2 (27 ppm), its dicarboxylic acid 9 (95 ppm), ketone 16 (36 ppm), alkene 14 (18 ppm) and a mixt. (86ppm) of the latter 3 were prepared in 5 concnssaturated (X), 0.75 X. 0.50 X, and 0.125 X. There were 6 replicate dishes for S. scoparium and 4 for each of the

al.

other species, except only 3 for the alkene (14) soln, and controls of distilled H,O were run at double the replication levels of the test solutions. Oxidized sesquiterpene bioassays. Bioassays were performed on the oxidized products of a satd ( + )-curlone soln after bubbling oxygen through the soln for 2 hr under irradiation using a 150 W, 120 V reflector spot lamp. Activities of (+)-curlone, its oxidation products and two (20 and 50 ppm) mixts of (+)-curlone decomposition products [4] were compared to an H,O control and a 0.1% EtOH control, as the latter was used to help dissolve the 20 and 50 ppm decomposition products. Still the 50 ppm soln was cloudy. Satd solns were prepared by sonication followed by filtration. Assays were conducted, as above, except that 480 ml, 8 cm diameter glass jars, with foil-lined lids were used instead of Petri dishes. Diterpene, sesquiterpene mixture bioassays. The sesquiterpene test mixt. was an equimolar mixt. (0.1 mM total concentration; ca 20 ppm) of the 3 sesquiterpenes (+)curlone, (+)-sesquiphellandrene. and (-)-a-transbergamotene previously isolated from C. paucijlosculosa [4]. Oxogrindelic acid (2) was assayed at 35 and 175 ppm, both separately and in combination with the sesquiterpene mixt. Because of the difficult solubilization of the diterpene aldehyde 2, sesqui- and diterpenes were applied to the centre of the filter paper in Me&O. The Me,CO was allowed to evap. completely, and 5 ~1 DMSO added as solubilizing agent, followed by 5 ml H20. Both Hz0 and H,O-0.1% DMSO controls were run. Concns are expressed assuming complete solubilization.

Acknowledgements-This material is based upon work supported by the Cooperative State Research Service, U. S. Department of Agriculture and Rangeland Renewable Resources under Agreement No. 88-33520-4077 of the Competitive Research Grants Program for Forest Biology. We thank Dr T. Manimaran, Ethyl Corporation, for obtaining optical rotations and Dr Rafael Cueto for FTIR spectra.

REFERENCES 1.

2.

3.

4. 5. 6.

Richardson, D. R. and Williamson, G. B. (1988) For. Sci. 34, 592. Fischer, N. H., Tanrisever, N. and Williamson, G. B. (1988) in Biologically Active Natural Products; Potential Use in Agriculture (Cutler, H. G., ed.), p. 233. Symposium Series 380, ACS, Washington, D.C. Eleuterius, L. N. (1979) Final Report for the Coastal Field Research Laboratory, Southeast Regional Office, National Park Service, U. S. A., pp. 101-110. Menelaou, M. A., Macias, F. A., Weidenhamer, J. D. and Fischer, N. H. (1993) Pbytochemistry (submitted). Panizzi, L., Mangoni, L. and Belardini, M. (1961) Gazz. Chim. Ital. 522. Brunn, T., Jackman, L. M. and Stenhagen, E. (1962) Acta Chem. Stand. 16, 1675.

Diterpenes

from Chrysomn paucijfoscuha

7. Guerreiro, E., Kavka, J., Saad, J., Oriental, M. and Giordano, I. (1981) Rev. Latinoam. Quim. 12, 77. 8. Bohlmann, F., Ahmed, M., Borthakur, N., Wallmeyer, M., Jakupovic, T., King, R. and Robinson, H. (1982) Phytochemistry 21, 167. 9. March, J. (1985) Advanced Organic Chemistry, 3rd (edn, p. 356. John Wiley. 10. Vinogradov, M. G. and Nikislim, G. I. (1971) Russ. Chem. Rev. 40,916.

PHYTO

34:1-H

105

11. O’Connell, A. M. (1973) Acta. Cryst. Sect. B 2!8,2232. 12. Menelaou, M. A., Foroozesh, M., Williamson, G. B., Fronczek, F. R., Fischer, H. D. and Fischer, N. H. (1992) Phytochemistry 31, 3769. 13. Coil, J. C. and Bowden, B. F. (1986) J. Nat. Prod. 49, 934. 14. Frenz, B. A. and Okaya, Y. (1980) Enraf-Nonius Structure Determination Package. Enraf-Nonius, Delft.