Antimycobacterial Plant Terpenoids

Table of Contents

Abstract

Tuberculosis (TB), mainly caused by Mycobacterium tuberculosis, is the leading killer among all infectious diseases worldwide and is responsible for more than two million deaths annually. For over thirty years no antitubercular agents with new mechanisms of action have been developed. The recent increase in the number of multi-drug resistant clinical isolates of M. tuberculosis has created an urgent need for the discovery and development of new antituberculosis leads. This review covers recent reports on plant-derived terpenoids that have demonstrated moderate to high activity in in vitro bioassays against M. tuberculosis. In this review, mono-, sesqui-, di- and triterpenes, and sterols, their structural analogs and semisynthetic derivatives will be discussed, with particular emphasis on the structural features essential for antimycobacterial activity.

Key words

Terpenoids - Mycobacterium tuberculosis - antituberculosis - tuberculostatic - antimycobacterial

Introduction

It is estimated that one-third of the world’s population is infected with the tubercle bacillus [1]. While only a small percentage of infected individuals will develop clinical tuberculosis, each year there are approximately eight million new cases and two million deaths. M. tuberculosis is thus responsible for more human mortality than any other single bacterial species. The HIV pandemic has exacerbated the problem by providing a large reservoir of highly susceptible individuals [2].

A number of efficacious antitubercular agents were discovered in the late 1940’s and 1950’s with the last, rifampin, introduced in the 1960’s [3], [4]. These agents have reasonable efficacy and when used in combination should preclude the development of drug resistance. The use (or in most cases misuse) of these drugs has lead over the years to an increasing prevalence of multiple-drug resistant (MDR) strains and there is now an urgent need to develop new effective agents [5], [6].

There have been a number of practical obstacles to the development of new anti-TB agents, among them a lack of economic incentive due to the predominance of disease in the developing world. The very slow growth and highly contagious nature of M. tuberculosis have also served to discourage the drug discovery effort.

Nonetheless, several drugs with interesting anti-TB activity have been identified in the past few years; three such compounds are the rifamycins. Rifabutin, with less P450 activating activity than rifampicin, has activity against a small percentage of rifampin-resistant strains [7], [8] and is active clinically [9], [10], [11], [12]. Rifapentene provides a much longer serum half-life and thus holds the promise of allowing for intermittent dosing but exhibits complete cross-resistance with rifampin [13]. KRM-1648 appears to be the most potent among the rifamycins and has demonstrated activity in phase II trials in tuberculosis [14]; however, it also exhibits partial cross-resistance with rifampin. Fluoroquinolones such as ofloxacin and levofloxacin have demonstrated clinical activity [15], [16], [17], [18]; however, the recently approved (in the U.S.A. for non-TB respiratory indications) moxifloxacin appears to be the most active flouroquinolone against M. tuberculosis in vitro and in the mouse [19], [20], [21]. Two members of a new class of antimicrobials, the oxazolidinones, showed modest in vivo activity against M. tuberculosis [22]. Finally a nitroimidazopyran has demonstrated interesting in vitro and in vivo activity [23]. While all of these compounds show potential, none have yet to demonstrate activity in clinical trials (with the exception of KRM-1648 and the older quinolones), thus there continues to be a need to identify new agents.

The following review surveys the literature for plant-derived terpenoids that have demonstrated moderate to significant biological activity against M. tuberculosis. It covers active compounds of five major groups of terpenoids: monoterpenes, sesquiterpenes, diterpenes, triterpenes, as well as phytol and its derivatives and structural analogs. Nearly all of the compounds discussed were screened using the BACTEC 460 system [24], [25] except where indicated otherwise. Compounds that did not demonstrate a minimum inhibitory concentration (MIC) of 64 μg/ml or lower and were unrelated to active terpenoids, have been omitted. Comparisons between chemical structures and their biological activities must be treated with caution, especially when different bioassay techniques have been used.

Monoterpenes

Due to the small number of monoterpenes with reported biological data and their structural similarity to phytol and its analogs, a more detailed structure-activity analysis will be discussed together with the results on phytol. It should be briefly pointed out, however, that among the open-chain monoterpenes, citronellol (1), nerol (2), and geraniol (3) gave MICs of 64, 128 and 64 μg/ml, respectively (Table [1]) [26], [27]. Compounds 2 and 3 represent respective Z- and E-isomers, and 1 being saturated at C-2; this suggests that the presence of a C-2 double bond has only minor influences upon the activity.[]

Table 1 Minimum inhibitory concentrations (MICs) of selected plant terpenoids against M. tuberculosis (H₃₇Rv)

Compound Class Name	Source	MIC (μg/ml) a	Reference
Antituberculosis Drugs
isoniazid	Sigma-Aldrich Co.	0.05	[40]
rifampin	-	0.25	[40]
streptomycin	Sigma-Aldrich Co.	2.0	[40]
ethambutol	Sigma-Aldrich Co.	3.8	[40]
pyrazinamide	Sigma-Aldrich Co.	100	[40]
Monoterpenes
citronellol (1)	Sigma-Aldrich Co.	64	[26]
nerol (2)	Sigma-Aldrich Co.	128	[26]
geraniol (3)	Sigma-Aldrich Co.	64	[27]
Sesquiterpenes
costunolide (4)	Saussurea lappa	32	[28]
11βH-dihydrocostunolide (5)	synthetic analog	128	[28]
parthenolide (6)	Magnolia grandiflora	16	[28]
11βH-dihydroparthenolide (7)	Ambrosia artemisiifolia	128	[28]
1,10-epoxycostunolide (8)	synthetic analog	64	[28]
1,10-epoxydihydrocostunolide (9)	synthetic analog	128	[28]
1,10-epoxyparthenolide (10)	synthetic analog	128	[28]
1,10-epoxydihydroparthenolide (11)	synthetic analog	128	[28]
α-cyclocostunolide (12)	synthetic analog	64	[26]
santamarine (13)	Ambrosia confertiflora	64	[28]
11βH,13-dihydrosantamarine (14)	Sonchus hierrensis	>128	[28]
β-cyclocostunolide (15)	synthetic analog	64	[26]
reynosin (16)	Ambrosia confertiflora	64	[28]
11βH,13-dihydroreynosin (17)	synthetic analog	>128	[28]
α-santonin (18)	Artemisia ramosa	>128	[26]
reynosin triol derivative (19)	-	>128	[28]
11,13-dihydro triol of reynosin (20)	-	>128	[28]
santamarine triol derivative (21)	-	>128	[28]
dehydrocostuslactone (22)	Saussurea lappa	2	[30]
7-hydroxydehydrocostuslactone (23)	Podachenium eminens	>128	[30]
zaluzanin C (24)	Zaluzania triloba	>128	[30]
dehydrocostuslactone, 4α(15)-epoxide (25)	synthetic analog	32	[30]
dehydrocostuslactone, 10β(14)-epoxide (26)	synthetic analog	32	[30]
dehydrocostuslactone, 10α(14)-epoxide (27)	synthetic analog	64	[30]
dehydrocostuslactone, 4β(15),10α(14)-diepoxide (28)	synthetic analog	64	[30]
dehydrocostuslactone, 4α(15),10α(14)-diepoxide (29)	synthetic analog	128	[30]
dehydrocostuslactone, 4α(15),10β(14)-diepoxide (30)	synthetic analog	128	[30]
micheliolide (31)	-	50	[30]
pumilin (32)	Berlandiera texana	128	[30]
damsin (33)	Ambrosia maritima	32	[26]
deacetylconfertiflorin (34)	Ambrosia confertiflora	128	[41], [26]
confertiflorin (35)	Ambrosia confertiflora	128	[41], [26]
deacetylisoconfertiflorin (36)	synthetic analog	128	[26]
parthenin (37)	Parthenium hysterophorus	64	[26]
tenulin (38)	Helenium amarum	128	[26]
peruvin (39)	Ambrosia peruviana	128	[26]
peruvin acetate (40)	synthetic analog	>128	[26]
burrodin (41)	Ambrosia dumosa	128	[42]
aromaticin (42)	Helenium aromaticin	16	[26]
alantolactone (43)	Inula helenium	32	[31]
isoalantolactone (44)	Inula helenium	32	[31]
encelin (45)	Montanoa speciosa	16	[31], [43]
1,2-dehydro-3-epi-isotelekin (46)	Montanoa speciosa	32	[31], [43]
ivalin (47)	Iva imbricata	64	[26]
11αH,13-dihydroisoalantolactone (48)	Inula helenium	>128	[31]
isoalloalantolactone (49)	Rudbeckia mollis	128	[44], [31]
alloalantolactone (50)	Rudbeckia subtomentosa	32	[31]
3-oxoalloalantolactone (51)	Rudbeckia subtomentosa	128	[31]
4α-epoxyisoalantolactone (52)	synthetic analog	32	[31]
5α-epoxyalantolactone (53)	synthetic analog	8	[31]
11,13-dihydroxyalantolactone (54)	synthetic analog	>128	[31]
6-epi-deacetyllaurenobiolide (55)	Montanoa grandiflora	16	[45]
4,5-epoxy-6-epideacetyllaurenobiolide (56)	synthetic analog	16	[26]
curcuphenol (57)	Euthamia leptocephela	16	(Robbs, S.L., 1997 unpublished)
nerolidol (58)	Magnolia acuminata	32	[26]
farnesol (59)	Sigma-Aldrich Co.	8	[27]
Diterpenes
sandaracopimara-8(14)-15-diene-7α,18-diol (60)	Tetradenia riparia	25 - 100d	[33]
sandracopimaric acid (61)	Juniperus excelsa	15.0b	[34]
sclareol (62)	-	6.0b	[34]
12-demethylmulticauline (63)	Salvia multicaulis	0.46c	[32]
multicaulin (64)	Salvia multicaulis	5.6c	[32]
12-demethylmultiorthoquinone (65)	Salvia multicaulis	1.2c	[32]
multiorthoquinone (66)	Salvia multicaulis	2.0c	[32]
12-methyl-5-dehydrohorminone (67)	Salvia multicaulis	1.2c	[32]
12-methyl-5-dehydroacetylhorminone (68)	Salvia multicaulis	0.89c	[32]
salvipimarone (69)	Salvia multicaulis	7.3c	[32]
9,12-cyclomulin-13-ol (70)	Azorella madreporica	20	[35]
juniperexcelsic acid (71)	Juniperus excelsa	14.4b	[34]
Triterpenes
ergosterol-5,8-endoperoxide (72)	Ajuga remota	1	[36]
ergosterol-5,8-endoperoxide acetate (73)	synthetic analog	8	[36]
ergosterol (74)	Sigma-Aldrich Co.	>128	[36]
12β-hydroxykulactone (75)	Melia volkensii	16	[37]
6β-hydroxykulactone (76)	Melia volkensii	4	[37]
kulonate (77)	Melia volkensii	16	[37]
(24R)-24,25-epoxycycloartan-3-one (78)	Borrichia frutescens	8	[25]
(3β,24R)-24,25-epoxycycloartan-3-ol (79)	Borrichia frutescens	8	[25]
(3β,24R)-24,25-epoxycycloartan-3-ol acetate (80)	Borrichia frutescens	>128	[25]
(23R)-3-oxolanosta-8,24-dien-23-ol (81)	Borrichia frutescens	64 - 128	[25]
compound 82	synthetic analog	>128	[25]
fusidic acid (83)	Sigma-Aldrich Co.	4	[25], [46]
zeorin (84)	Sarmienta scandens	8e	[38]
7β-acetyl-22-hydroxyhopane (85)	Sarmienta scandens	>128e	[38]
7β,22-dihydroxyhopane (86)	Sarmienta scandens	>128e	[38]
oleanolic acid (87)	Baccharis patagonica	64e	[38]
erythodiol (88)	Baccharis patagonica	64e	[38]
3-epioleanolic acid (89)	Junellia tridens	16	[39]
oleanonic acid (90)	Junellia tridens	16	[39]
lupeol (91)	Chuquiraga ulicina	64e	[38]
betulinic acid (92)	-	32e	[38]
betulin (93)	-	32e	[38]
epi-betulinic acid (94)	Monttea aphylla	64e	[38]
lupeol acetate (95)	Chuquiraga ulicina	>128e	[38]
lupenone (96)	Chuquiraga ulicina	>128e	[38]
3-hydroxynorlupen-2-one (97)	Chuquiraga ulicina	>128e	[38]
3-acetoxynorlupen-2-one (98)	Chuquiraga ulicina	>128e	[38]
ursolic acid (99)	Aspidosperma quebracho-blanco	32e	[38]
uvaol (100)	-	32e	[38]
pomolic acid (101)	Acaena pinnatifida	64e	[38]
pomolic acid acetate (102)	Acaena pinnatifida	32e	[38]
tormentic acid (103)	Acaena pinnatifida	32e	[38]
2-epi-tormentic acid (104)	Acaena pinnatifida	>128e	[38]
euscaphic acid (105)	Acaena pinnatifida	128e	[38]
niga-ichigoside F1 aglycone (106)	Acaena pinnatifida	>128e	[38]
Phytol, derivatives, and analogs
(E)-phytol (107)	Lucas volkensii	2	[27]
(E)-phytyl acetate (108)	synthetic analog	16	[27]
(E)-phytol methyl ether (109)	synthetic analog	16	[26]
phytyl amine (110)	synthetic analog	32	[26]
phytyl diisopropylamine (111)	synthetic analog	64	[26]
(Z)-phytol (112)	Sigma-Aldrich Co.	2	[27]
(3R,S,7R,11R)-phytanol (113)	synthetic analog	2	[27]
(E)-phytol epoxidation epimers (114)	synthetic analog	8	[27]
(3R,S,7R,11R)-phytanic acid (115)	synthetic analog	>128	[27]
2-phytylphenol (116)	synthetic analog	32	[26]
phytantriol (117)	Sigma-Aldrich Co.	16	[26]
2-hexadecanol (118)	Sigma-Aldrich Co.	8	[26]
a Biological activity determined using BACTEC radiorespirometric bioassay except where noted otherwise.
b Biological activity determined using disc-diffusion method.
c Biological activity determined using broth microdilution method.
d Biological activity determined by conventional proportion method. Clinical isolates of M. tuberculosis were used in the bioassay.
e Concentration reported in μM.

Sesquiterpenes

Among the over fifty sesquiterpenes, mainly sesquiterpene lactones of the germacranolide, guaianolide, and eudesmanolide type, tests resulted in MICs ranging from 2 μg/ml to >128 μg/ml. A series of α-methylene-γ-lactone containing germacranolides showed MICs at or below 64 μg/mL [28]. The relatively lipophilic costunolide (4) gave an MIC of 32 μg/ml while its 4-epoxide derivative, parthenolide (6), showed an increased activity against M. tuberculosis (MIC 16 μg/ml). The 1(10)-epoxycostunolide (8) was less active than 4 and 6 with an MIC of 64 μg/ml, suggesting that the position of the epoxide within the medium ring has a distinct influence on the activity. With the exception of the non-active diepoxide 1,10-epoxyparthenolide (10), all the sesquiterpene lactones in this series with an α-methylene-γ-lactone moiety (4, 6, 8, 13, 16, 12 and 15) have activities with MICs at 64 μg/ml or below. In contrast, their 11βH,13-dihydro derivatives (5, 7, 9, 14, and 17) as well as the sesquiterpenes obtained by reductive opening of the lactone ring (19, 20, and 21) [29] showed no activity against M. tuberculosis at concentrations below 128 μg/ml, suggesting that the presence of the exocyclic α-methylene-γ-lactone moiety is essential, but not sufficient, for activity.

Parthenolide, with an MIC of 16 μg/ml, was the most active germacranolide tested and exhibited a higher MIC than costunolide (MIC = 32 μg/ml). Moreover, 1(10)-epoxycostunolide (8) and the diepoxide 10, although both with an α-methylene-γ-lactone moiety and an epoxide at C-1(10), neither compound showed activity comparable with parthenolide (6). It has therefore been suggested that the higher antimycobacterial activity of parthenolide (6) may be due to the presence of two sites of alkylation, one site being the α-methylene-γ-lactone moiety and the other site at C-10, the electron deficient center being generated by a transannular cyclization with the C-1(10) double bond being the donor and the C-5 epoxide the nucleophilic receptor [28].

A series of guaianolides screened against M. tuberculosis demonstrated MICs ranging from 2 to >128 μg/ml, with the lipophilic dehydrocostuslactone (22) (MIC of 2 μg/ml) being the most active [30]. Semi-synthetic mono- and diepoxide derivatives of 22 (25 - 30), as well as its hydroxy analogs showed very distinct activity trends. Monoepoxides 25, 26, and 27 were more active than the more polar diepoxides 28, 29, and 30, which in turn were more active than the more polar hydroxy analogs 23 and 24. It has been suggested that the low MIC of 22 may be due not only to its exocyclic methylene lactone moiety but also to its lipophilic nature [30]. This was supported by the decreased activity from 22 to the more polar monoepoxides via the diepoxides to the most polar hydroxy analogs 23 and 24. Micheliolide (31) (MIC 50 μg/mL) holds a unique position, possibly due to the ability to undergo an intramolecular substitution involving the 1(10)-double bond as a nucleophilic donor and the C-4 hydroxy as a leaving group, thus allowing a double alkylation as in parthenolide (6).

The pseudoguaianolides damsin (33), parthenin (37), and aromaticin (42) showed antimycobacterial activity with MICs of 32, 64, and 16 μg/ml, respectively. This suggests that two alkylating sites within the molecule such as 42, increase activity. Again, the more polar C-8 hydroxy (34) and acetoxy (35) analogs of 33 gave MICs of 128 μg/ml, that is, much lower activity than 33, which may be due to the more polar nature of 34 and 35. Also, tenulin (38) differs from 42 in its absence of an α-methylene-γ-lactone moiety. Consequently, its MIC of 128 μg/ml further supports the necessity of this lactone group for significant activity.

Bioassay-directed investigations of Inula helenium and Rudbeckia subtomentosa resulted in the isolation of eudesmanolides, with MICs ranging from 8 to >128 μg/ml. Moderate activities were observed for alantolactone (43), isoalantolactone (44), 11αH,13-dihydroisoalantolactone (48), alloalantolactone (50), and 3-oxoalloalantolactone (51) [31]. Compounds 43, 44, and 50 all gave MICs of 32 μg/ml, while 48 and 51 were not active (MICs >128 and 128 μg/ml, respectively). Stereoselective epoxidation of 43 gave the eudesmanolide 5α-epoxyalantolactone (53) with an MIC of 8 μg/ml while oxidation of 43 with OsO₄ resulted in the inactive dihydroxy derivative 54. Analogs of 44 include the monoepoxide 52, the C-2 hydroxy derivative 47, the conjugated ketone 45, and alcohol 46 with MICs of 32, 64, 16, and 32 μg/ml, respectively.

As seen above with the various structural types of sesquiterpene lactones, the α-methylene-γ-lactone moiety appears to be an essential, but not sufficient, structural requirement for significant activity. The necessity of the presence of an α-methylene-γ-lactone group is supported by the moderate to high activity of α-methylene-γ-lactone bearing sesquiterpenes, when compared with the inactive 11αH,13-dihydro derivatives with values of 128 μg/ml or higher. The presence of a second alkylating site, such as an α,β-unsaturated carbonyl group and/or an epoxide function together with moderate to high lipophilicity, seems to enhance the in vitro antimycobacterial activity. For instance, encelin (45) is more active than 44 and 5α-epoxyalantolactone (53), with an MIC of 8 μg/ml, is significantly more active than its precursor 43 with an MIC of 32 μg/ml.

The sesquiterpenes curcuphenol (57), nerolidol (58), and farnesol (59) gave MICs of 16, 32, and 8 μg/ml, respectively. Compounds 58 and 59 are constitutional isomers differing in the positioning of the allylic alcohol. Due to the structural similarity of 57, 58, and 59 with phytol and its analogs, a more detailed discussion of their biological activities will follow later.[] [*] [*] []

Diterpenes

A limited number of diterpenes have been tested for antituberculosis activity and some have demonstrated remarkable biological activities against M. tuberculosis with MICs below 1 μg/ml. The most active among the diterpenes were recently reported by Ulubelen et al. from Salvia multicaulis, the norditerpenoid 12-demethylmulticauline (63) showing a remarkable MIC of 0.46 μg/ml and its C-12 methoxy analog, 64, with an MIC of 5.6 μg/ml [32]. Similarly, the phenolic o-quinone (65) gave an MIC of 1.2 μg/ml while its C-2 methoxy analog had an MIC of 2.0 μg/ml. Interestingly, the abietane diterpenoid 67 had an MIC of 1.2 while its C-12 acetate 68 was more active with an MIC of 0.89 μg/ml [32]. Compounds 60 [33], 61 [34], 62 [34], 69 [32], 70 [35], and 71 [34] gave MICs of 25 - 100, 15, 6, 7.3, 20, and 14.4 μg/ml, respectively. Biological activities of the diterpene phytol, its derivatives and structural analogs will be discussed later.[]

Triterpenes and Sterols

Numerous triterpenoids and sterols have been tested with MICs ranging from 1 to >128 μg/ml. Bioassay-guided investigations of active fractions from Ajuga remota led to the isolation of the most active triterpenoid, ergosterol-5,8-endoperoxide (72), with an MIC of 1 μg/ml [36]. Acetylation of 72 provided the acetate, 73, with an MIC of 8 μg/ml. In contrast, the parent compound, ergosterol (74), gave an MIC of >128 μg/ml. Compound 72 was prepared in a one step synthesis from commercially available ergosterol (74) [36] making it an ideal candidate for future structure-activity relationship studies. Within this set of ergosterol derivatives, it appears that the presence of a C-3-OH combined with the endoperoxide group is necessary for high antimycobacterial activity.

Bioactivity-guided investigations of Melia volkensii resulted in the isolation of 12β-hydroxykulactone (75), 6β-hydroxykulactone (76), and kulonate (77). Compounds 75 and 77 both had MICs of 16 μg/ml while compound 76 was more active with an MIC of 4 μg/ml [37]. From this set of limited data, it appears that a hydroxy group at C-6 gives rise to higher activity than a C-12 hydroxy group. Based on the same activity of 75 and 77, there appears to be little or no effect on biological activity upon methanolysis of the lactone moiety.

In another bioassay-guided investigation, Borrichia frutescens afforded a number of antimycobacterial cycloartanes [25]. The most active of these compounds were (24R)-24,25-epoxycycloartan-3-one (78) and (3β,24R)-24,25-epoxycycloartan-3-ol (79) with MICs of 8 μg/ml, whereas the acetate 80 was inactive (MIC >128 μg/ml). Unsuccessful attempts to oxidize 79 to 78 resulted in the synthesis of the inactive degradation product 82 (MIC >128 μg/ml), and the isolate 81 gave an MIC of 64 - 128 μg/ml. Correlations of structural features and the MICs of these five triterpenes suggest that the presence of the C-3 keto and/or β-hydroxy group, the cyclopropane ring and the epoxide moieties, as in 78 and 79, seem to play a major role in the in vitro antituberculosis activity. Both, the cyclopropane and epoxide functions are absent in 81, resulting in the loss of activity. Also, the loss of activity by acetylation of the C-3 hydroxy group (MIC >128 μg/mL) strongly suggests that either a free hydroxy or a keto group at C-3 is required for significant activity.

Several pentacyclic triterpenoids have been isolated with activities ranging from 8 to >128 μM. The most active of these compounds was zeorin (84), isolated by bioassay-guided fractionation of Sarmienta scandens and shown to have an MIC of 8 μg/ml [38]. 7β,22-Dihydroxyhopane (86) contains a β-hydroxy group at C-7 rather than an α-hydroxy at C-6 in 84, resulting in a loss of activity (MIC >128 μM). In addition, the C-7 acetoxy derivative of 86, 7β-acetyl-22-hydroxyhopane (85), was inactive [38]. 3-Epioleanolic acid (89) and oleanonic acid (90), which differ in the functional group present at C-3, were both isolated from Junellia tridens and shown to have MICs of 16 μg/ml [39]. Compound 89 contains an α-hydroxy group at C-3 while 90 bears a ketone. Oleanolic acid 87 differs from 89 by the presence of a C-3 β rather than an α-hydroxy group, resulting in a less active derivative with an MIC of 64 μM. The dihydroxy analog of 87, erythrodiol (88), gave a similar MIC value of 64 μg/ml. Similarly, ursolic acid (99) and betulinic acid (92) gave MIC values of 32 μM, identical to those of uvaol (100) and betulin (93) [38]. In contrast to the results obtained when comparing the MICs of compounds 87 and 89, betulinic acid (92) with its β-hydroxy group at C-3 is more active than its C-3 epimer, epi-betulinic acid (94) (MIC 64 μM). Also, replacement of the 3-hydroxy group of lupeol (91) with either an acetate or a ketone leads to complete loss of activity in lupeol acetate (95) and lupenone (96). A similar result was observed for the C-3 acetoxy derivative 80 when compared to 79. However, the 3-oxo derivative 78 gave the same MIC as 79 which is in contrast to the correlations mentioned above. Further complicating the situation was the observation that pomolic acid 3-acetate (102) was more active than pomolic acid (101) itself. Comparison of ursolic acid (99) with pomolic acid (101), tormentic acid (103), 2-epi-tormentic acid (104), euscaphic acid (105) and the aglycone of niga-ichigoside (106) showed that additional hydroxy groups in rings A and E can, in some cases, reduce activity [38]. These conflicting results observed among the various groups of triterpenoids makes it difficult to predict structural requirements for antimycobacterial activity.[] [*]

Phytol, Derivatives and Analogs

This group of open chain diterpenes covers phytol, its derivatives and structural analogs. Also included are the previously mentioned linear monoterpenes 1, 2, and 3 and diterpenes 57, 58, and 59. Within this set of structural analogs, all but two compounds showed MICs at or below 64 μg/ml. One of the most active compounds in this series, (E)-phytol (107) (MIC 2 μg/ml) was isolated from Lucas volkensii, using a bioassay-guided fractionation [27]. In addition, the analogs (Z)-phytol (112) and (3R,S,7R,11R)-phytanol (113) demonstrated MICs of 2 μg/ml, suggesting that the 2,3-double bond may not be essential for bioactivity. However, (E)-phytyl acetate (108) and (E)-phytol methyl ether (109) showed MICs of 16 μg/ml implying that a free hydroxy group, as present in 107, is required for significant activity. Due to their structural and biosynthetic similarities to (E)-phytol, geraniol (3) and farnesol (59) were also tested against M. tuberculosis. The monoterpene alcohol geraniol had an MIC of 64 μg/ml while the more lipophilic farnesol with a 15-carbon chain gave a significant increase in activity with an MIC of 8 μg/ml. Citronellol (1) and nerol (2) were tested due to their structural similarity to geraniol and gave MICs of 64 and 128 μg/ml, respectively. The identical MICs of 64 μg/ml between geraniol and citronellol further support the observation above for phytol and its reduced isomer (3R,S,7R,11R)-phytanol that the 2-double bond is not essential for biological activity. Additional derivatives and analogs tested include curcuphenol (57), nerolidol (58), phytylamine (110), phytyldiisopropylamine (111), (E)-phytol epoxidation epimers (114), (3R,S,7R,11R)-phytanic acid (115), 2-phytylphenol (116), phytantriol (117), and 2-hexadecanol (118) giving MICs of 16, 32, 32, 64, 8, >128, 32, 16, and 8 μg/ml, respectively [27], [26]. It is significant to note that a complete loss of activity is found when the C-1 hydroxy of (3R,S,7R,11R)-phytanol (113) is oxidized to the carboxylic acid, (3R,S,7R,11R)-phytanic acid (115), again suggesting that the free hydroxy is essential for high activity.

The most interesting observation among the above group of compounds is the relationship between the calculated log P values and the MIC. Higher lipophilicity generally results in higher antimycobacterial activity, within a series of structurally similar compounds. For example, the more lipophilic 20-carbon phytol (log P = 8.66) is more active than the 15-carbon farnesol (log P = 5.31) which is more active than the 10-carbon geraniol (3) (log P = 3.28). However, there appears to be a point at which further increasing the lipophilicity results in lower antimycobacterial activity (Fig. [1]). Replacing the hydroxy of phytol with nitrogen containing functional groups strongly reduces activity in spite of an increase in lipophilicity. For example, replacing the hydroxy of phytol with a diisopropylamine as in 111 (log P = 11.14) results in a more lipophylic molecule that is less active. Figure [1] is a plot of calculated log P values versus their respective MIC’s, for a selected group of phytol derivatives and analogs.

Fig. 1 The relationship between calculated Log P values and antimycobacterial activity of phytol, its derivatives, and structural analogs. Plotted compounds include: 1, 3, 57, 58, 59, 107, 108, 109, 111, 112, 113, 114, 116, 117, 118. Curve fitting was performed using a polynomial fourth order equation. *Log P values were estimated using ACD Labs Log P module [47].

Conclusions

A broad range of plant terpenoids from various classes have been evaluated for their in vitro antimycobacterial activity. The most active terpenoid presented in this review is the norditerpenoid 12-demethylmulticauline (63), isolated from the roots of Salvia multicaulis, with an MIC of 0.46 μg/ml. It is more active than the first line tuberculosis drug ethambutol and nearly as active as rifampin. Further terpenoids with high activities are the sesquiterpene dehydrocostuslactone (22) with an MIC of 2 μg/ml, the sterol ergosterol-5,8-endoperoxide (72) (MIC 1 μg/ml), and (E)-phytol (107) with an MIC of 2 μg/ml.

Distinct similarities between different classes of terpenes and the structural features important for high activity within each class have been observed. In each series it was shown that more lipophilic compounds are significantly more active than their more polar analogs. This was observed for the sesquiterpene lactone dehydrocostuslactone (22) which gave an MIC of 2 μg/ml, when compared to its C-3 and C-7 hydroxy analogs, both with MICs of >128 μg/ml. Furthermore, higher MICs were found for the monoepoxide derivatives of 22, followed by even higher MICs of the diepoxides. Additional support for a correlation between lipophilicity and antimycobacterial activity was found in the (E)-phytol series in which the highest activity was observed for the least polar 20-carbon (E)-phytol (107) followed by the 15-carbon analog farnesol (59), which was more active than the more polar 10-carbon geraniol (3).

Among phytol (107), its acetate (108), and methyl ether analog 109, both derivatives gave distinct decreases in activity. Also, a decrease of activity was found for the more lipophilic nitrogen containing phytol analogs 110 and 111. Similar relationships were observed among the triterpenes and sterols where a free hydroxy or carbonyl at C-3 appears to be essential for high activity. For instance, compounds 72 and 79 were both more active than their corresponding C-3 acetoxy derivatives, 73 and 80, respectively.

The resurgence of infectious diseases such as drug-resistant tuberculosis necessitates more intense future efforts in the discovery of new specific drugs from natural and synthetic sources. The recent development of efficient and reproducible bioassays plays a significant role in the present and future discovery and development of new anti-TB leads. A number of compounds covered in this review possess in vitro antimycobacterial activities comparable to standard anti-TB drugs. They certainly warrant further investigations on the long path from the initial activity findings to the development of new antituberculosis drugs.

Acknowledgements

The authors are grateful to Dr. Mohamed S. Rajab from the Department of Chemistry, Moi University, Eldoret, Kenya for helpful discussions.

References

• 1 Dye C, Scheele S, Dolin P, Pathania V, Raviglione M C. Consensus statement. Global burden of tuberculosis: estimated incidence, prevalence, and mortality by country. WHO Global Surveillance and Monitoring Project.  Journal of the American Medical Association. 1999;  282 677-86
• 2 Raviglione M C, Snider D E, Jr., Kochi A. Global epidemiology of tuberculosis. Morbidity and mortality of a worldwide epidemic.  Journal of the American Medical Association. 1995;  273 220-6
• 3 Grassi C, Peona V. New drugs for tuberculosis. European Respiratory Journal.  Supplement. 1995;  20 714s-8s
• 4 Sepkowitz K A, Raffalli J, Riley L, Kiehn T E, Armstrong D. Tuberculosis in the AIDS era.  Clinical Microbiology Reviews. 1995;  8 180-99
• 5 Cohn D L, Bustreo F, Raviglione M C. Drug-resistant tuberculosis: review of the worldwide situation and the WHO/IUATLD Global Surveillance Project. International Union Against Tuberculosis and Lung Disease.  Clinical Infectious Diseases. 1997;  24 S121-30
• 6 Espinal M A, Laszlo A, Simonsen L, Boulahbal F, Kim S J, Reniero A, Hoffner S, Rieder H L, Binkin N, Dye C, Williams R, Raviglione M C. Global trends in resistance to antituberculosis drugs.  New England Journal of Medicine. 2001;  344 1294-303
• 7 Chien H P, Yu M C, Ong T F, Lin T P, Luh K T. In vitro activity of rifabutin and rifampin against clinical isolates of Mycobacterium tuberculosis in Taiwan.  Journal of the Formosan Medical Association. 2000;  99 408-11
• 8 Yang B, Koga H, Ohno H, Ogawa K, Fukuda M, Hirakata Y, Maesaki S, Tomono K, Tashiro T, Kohno S. Relationship between antimycobacterial activities of rifampicin, rifabutin and KRM-1648 and rpoB mutations of Mycobacterium tuberculosis .  Journal of Antimicrobial Chemotherapy. 1998;  42 621-8
• 9 Matteelli A, Olliaro P, Signorini L, Cadeo G, Scalzini A, Bonazzi L, Caligaris S, Tomasoni L, Tebaldi A, Carosi G. Tolerability of twice-weekly rifabutin-isoniazid combinations versus daily isoniazid for latent tuberculosis in HIV-infected subjects: a pilot study.  International Journal of Tuberculosis and Lung Disease. 1999;  3 1043-6
• 10 Narita M, Stambaugh J J, Hollender E S, Jones D, Pitchenik A E, Ashkin D. Use of rifabutin with protease inhibitors for human immunodeficiency virus-infected patients with tuberculosis.  Clinical Infectious Diseases. 2000;  30 779-83
• 11 Parola D, Dell’Orso D, Giovagnoli S, Bianconi S, Terzano C. The tolerability and therapeutic efficacy of rifabutin in the treatment of pulmonary tuberculosis.  Recenti Progressi in Medicina. 1999;  90 254-7
• 12 Sirgel F A, Botha F J, Parkin D P, Van De Wal B W, Donald P R, Clark P K, Mitchison D A. The early bactericidal activity of rifabutin in patients with pulmonary tuberculosis measured by sputum viable counts: a new method of drug assessment.  Journal of Antimicrobial Chemotherapy. 1993;  32 867-75
• 13 Lenaerts A M, Chase S E, Chmielewski A J, Cynamon M H. Evaluation of rifapentine in long-term treatment regimens for tuberculosis in mice.  Antimicrobial Agents and Chemotherapy. 1999;  43 2356-60
• 14 Tomioka H. Prospects for development of new antimycobacterial drugs, with special reference to a new benzoxazinorifamycin, KRM-1648.  Archivum Immunologiae et Therapiae Experimentalis. 2000;  48 183-8
• 15 Maranetra K N. Quinolones and multidrug-resistant tuberculosis.  Chemotherapy. 1999;  45 12-8
• 16 Sirgel F A, Botha F J, Parkin D P, Van de Wal B W, Schall R, Donald P R, Mitchison D A. The early bactericidal activity of ciprofloxacin in patients with pulmonary tuberculosis.  American Journal of Respiratory and Critical Care Medicine. 1997;  156 901-5
• 17 Sirgel F A, Donald P R, Odhiambo J, Githui W, Umapathy K C, Paramasivan C N, Tam C M, Kam K M, Lam C W, Sole K M, Mitchison D A. A multicentre study of the early bactericidal activity of anti-tuberculosis drugs.  Journal of Antimicrobial Chemotherapy. 2000;  45 859-70
• 18 Yew W W, Chan C K, Chau C H, Tam C M, Leung C C, Wong P C, Lee J. Outcomes of patients with multidrug-resistant pulmonary tuberculosis treated with ofloxacin/levofloxacin-containing regimens.  Chest. 2000;  117 744-51
• 19 Gillespie S H, Billington O. Activity of moxifloxacin against mycobacteria.  Journal of Antimicrobial Chemotherapy. 1999;  44 393-5
• 20 Ji B, Lounis N, Maslo C, Truffot-Pernot C, Bonnafous P, Grosset J. In vitro and in vivo activities of moxifloxacin and clinafloxacin against Mycobacterium tuberculosis .  Antimicrobial Agents and Chemotherapy. 1998;  42 2066-9
• 21 Miyazaki E, Miyazaki M, Chen J M, Chaisson R E, Bishai W R. Moxifloxacin (BAY12 - 8039), a new 8-methoxyquinolone, is active in a mouse model of tuberculosis.  Antimicrobial Agents and Chemotherapy. 1999;  43 85-9
• 22 Cynamon M H, Klemens S P, Sharpe C A, Chase S. Activities of several novel oxazolidinones against Mycobacterium tuberculosis in a murine model.  Antimicrobial Agents and Chemotherapy. 1999;  43 1189-91
• 23 Stover C K, Warrener P, VanDevanter D R, Sherman D R, Arain T M, Langhorne M H, Anderson S W, Towell J A, Yuan Y, McMurray D N, Kreiswirth B N, Barry C E, Baker W R. A small-molecule nitroimidazopyran drug candidate for the treatment of tuberculosis.  Nature. 2000;  405 962-6
• 24 Collins L, Franzblau S G. Microplate alamar blue assay versus BACTEC 460 system for high-throughput screening of compounds against Mycobacterium tuberculosis and Mycobacterium avium .  Antimicrobial Agents and Chemotherapy. 1997;  41 1004-9
• 25 Cantrell C L, Lu T, Fronczek F R, Fischer N H, Adams L B, Franzblau S G. Antimycobacterial cycloartanes from Borrichia frutescens .  Journal of Natural Products. 1996;  59 1131-6
• 26 Cantrell C L. Antimycobacterial natural products from higher plants. Department of Chemistry Baton Rouge, LA; Louisiana State University 1998: 173 pp
• 27 Rajab M S, Cantrell C L, Franzblau S G, Fischer N H. Antimycobacterial activity of (E)-phytol and derivatives. A preliminary structure-activity study.  Planta Medica. 1998;  64 2-4
• 28 Fischer N H, Lu T, Cantrell C L, Castaneda-Acosta J, Quijano L, Franzblau S G. Antimycobacterial evaluation of germacranolides.  Phytochemistry. 1998;  49 559-64
• 29 Lu T, Fischer N H. Spectral data of chemical modification products of costunolide.  Spectroscopy Letters. 1996;  29 437-48
• 30 Cantrell C L, Nunez I S, Castaneda-Acosta J, Foroozesh M, Fronczek F R, Fischer N H, Franzblau S G. Antimycobacterial activities of dehydrocostus lactone and its oxidation products.  Journal of Natural Products. 1998;  61 1181-6
• 31 Cantrell C L, Abate L, Fronczek F R, Franzblau S G, Quijano L, Fischer N H. Antimycobacterial eudesmanolides from Inula helenium and Rudbeckia subtomentosa .  Planta Medica. 1999;  65 351-5
• 32 Ulubelen A, Topcu G, Johansson C B. Norditerpenoids and diterpenoids from Salvia multicaulis with antituberculous activity.  Journal of Natural Products. 1997;  60 1275-80
• 33 Puyvelde L V, Ntawukiliyayo J D, Portaels F. In vitro inhibition of mycobacteria by Rwandese medicinal plants.  Phytotherapy Research. 1994;  8 65-9
• 34 Topcu G, Erenler R, Cakmak O, Johansson C B, Celik C, Chai H -B, Pezzuto J M. Diterpenes from the berries of Juniperus excelsa .  Phytochemistry. 1999;  50 1195-9
• 35 Wachter G A, Franzblau S G, Montenegro G, Suarez E, Fortunato R H, Saavedra E, Timmermann B N. A new antitubercular mulinane diterpenoid from Azorella madreporica Clos.  Journal of Natural Products. 1998;  61 965-8
• 36 Cantrell C L, Rajab M S, Franzblau S G, Fronczek F R, Fischer N H. Antimycobacterial ergosterol-5,8-endoperoxide from Ajuga remota .  Planta Medica. 1999;  65 732-4
• 37 Cantrell C L, Rajab M S, Franzblau S G, Fischer N H. Antimycobacterial triterpenes from Melia volkensii .  Journal of Natural Products. 1999;  62 546-8
• 38 Wachter G A, Valcic S, Flagg M L, Franzblau S G, Montenegro G, Suarez E, Timmermann B N. Antitubercular activity of pentacyclic triterpenoids from plants of Argentina and Chile.  Phytomedicine. 1999;  6 341-5
• 39 Caldwell C G, Franzblau S G, Suarez E, Timmermann B N. Oleanane triterpenes from Junellia tridens .  Journal of Natural Products. 2000;  63 1611-4
• 40 Heifets L B, Cynamon M H. Drug susceptibility in the chemotherapy of mycobacterial infections. Boca Raton, Fla; CRC Press 1991
• 41 Fischer N H, Mabry T J. New pseudoguianolides from Ambrosis confertiflora .  Tetrahedron. 1967;  23 2529-38
• 42 Geissman T A, Matsueda S. Sesquiterpene lactones. Constituents of diploid and polyploid Ambrosia dumosa .  Phytochemistry. 1968;  7 1613-21
• 43 Quijano L, Gomez-Garibay F, Trejo-B R I, Rios T. Hydroxy-bis-dihydroencelin, a dimeric eudesmanolide and other eudesmanolides from Montanoa speciosa .  Phytochemistry. 1991;  30 3293-5
• 44 Vasquez M, Quijano L, Urbatsch L E, Fischer N H. Sesquiterpene lactones and other constituents from Rudbeckia mollis .  Phytochemistry. 1992;  31 2051-4
• 45 Quijano L, Calderon J S, Federico Gomez G, Jesus Lopez P, Rios T, Fronczek F R. The crystal structure of 6-epi-deacetyllaurenobiolide, a germacra-1(10),4-diene-12,8-alpha-olide from Montanoa grandiflora .  Phytochemistry. 1984;  23 1971-4
• 46 von Daehne W, Godtfredsen W O, Rasmussen P R. Structure-activity relationships in fusidic acid-type antibiotics.  Advances in Applied Microbiology. 1979;  25 95-146
• 47 Osterberg T, Norinder U. Prediction of drug transport processes using simple parameters and PLS statistics. The use of ACD/logP and ACD/ChemSketch descriptors.  European Journal of Pharmaceutical Science. 2001;  12 327-37