Ronald
Bentley
*
Department of Biological Sciences, University of Pittsburgh, Pittsburgh, PA 15260, USA
First published on 22nd November 2007
Covering: 1945 to 2007
Tropones and tropolones have been of increasing interest in the past few years; approximately 200 such materials occur in nature. This article reviews the distribution of these materials, the biosynthetic pathways for their production and their biological activities. There are 154 references.
Ronald Bentley | Ronald Bentley worked on penicillin chemistry at Imperial College, London, obtaining his Ph.D. in 1945. After a Commonwealth Fund Fellowship for work on isotope tracers with David Rittenberg (College of Physicians and Surgeons, Columbia University, New York), he set up a mass spectrometer unit at the National Institute of Medical Research, London. Joining the University of Pittsburgh in 1953 he became Professor in 1960, studying the biosynthesis of various secondary metabolites (including pioneer studies on fungal tropolones), ubiquinone and vitamin K. An interest in stereochemistry led to a two-volume text, “Molecular Asymmetry in Biology”, published in 1969–1970. Retired at age 70 (1992) as Professor Emeritus he has continued writing and was an editor for The Oxford Dictionary of Biochemistry and Molecular Biology (1997, revised 2000). He received the Waksman Outstanding Educator Award from the Society for Industrial Microbiology in 2002. He takes pride in more than sixty years of membership of the RSC. |
In 1945, Michael Dewar made a brilliant leap of imagination and proposed a seven-membered ring structure for stipitatic acid based on 2-hydroxy-2,4,6-cycloheptatrien-1-one.4 This latter compound, 1 R1 = R2 = H, named tropolone, showed certain aromatic properties; Dewar had, in fact, unleashed the new field of non-benzenoid aromatic compounds. In addition, Dewar suggested a tropolone structure for colchicine (see later).5 The impact of Dewar's proposal was rapid and impressive. A review, “The Tropolones ” published in 1951, contained 80 citations;6 only 4 years later, Pauson's comprehensive “Tropones and Tropolones ” required 492 citations.7
Another theme in the tropolone story concerns the heartwood of Western red cedar. This wood contains materials that are antifungal and are responsible for its durability. Two materials, C10H12O2, had been isolated in 1933 and in 1946 were also found to be present in Swedish-grown Thuja plicata. Extensive investigations by Erdtman's group led to the isolation and structure determination of three, isomeric materials, termed thujaplicins.8 They were clearly shown to be isopropyl tropolones , e.g., α-thujaplicin 1 R1 = H, R2 = isopropyl.
Dewar and Erdtman were working towards the end of World War II, a time when normal scientific communications were often difficult, if not impossible. Such difficulties had impinged on the life of a Japanese chemist, Tetsuo Nozoe born 1902, who died in 1996 just before his 94th birthday.9 In 1926, he had accepted a position in Taipei, Formosa (Taiwan), at that time in Japanese hands and in 1929 he became an Assistant Professor in the Faculty of Science and Agriculture of the newly established Taihoku Imperial University. In early work he studied the components of saponins, wool wax, and essential oils from leaves of taiwanhinoki (Chamaecyparis taiwanensis) and other trees. A dark-red wood pigment, “hinokitin” had been obtained in 1926 from oil of the heartwood of taiwanhinoki. In 1936, Nozoe treated an ether solution of hinokitin with aqueous alkali, obtaining a precipitate of ferric hydroxide and an enolic compound, C10H12O2. He named the latter compound, hinokitiol, and showed that hinokitin was an iron complex of hinokitiol, (C10H11O2)3Fe.
Assigning a structure to hinokitiol proved difficult; one possibility was “the enol form of the seven-membered α-diketone”, 2. Since common knowledge at that time was that “no such compound could exist in nature in a stable form” this structure was temporarily abandoned.10 Further examination of hinokitiol and degradative studies led him to reconsider the seven-membered ring. A tentative tropolone structure, presented in 1940, was actually for a dihydro form, C10H14O23—there had been analytical problems. After the hinokitiol structure had been reconfirmed as C10H12O2, Nozoe obtained a copy of Pauling's, “The Nature of the Chemical Bond” and under its influence considered the seven-membered structure 4a, 4b, 4c as “a new type of aromatic compound stabilized by resonance”. Nozoe states that he presented this “tropolonoid” formula at the Formosa Branch of the Society of the Chemical Industry, in 1941, but that it was received with skepticism.10 Unhappily, there is apparently no published record of this presentation.
With the end of World War II, the Taihoku Imperial University became the National Taiwan University (Republic of China). Nozoe was able to remain there and in 1946 resumed his hinokitiol work, now with a large research group. Substitution reactions convinced him that hinokitiol was indeed an entirely new type of aromatic compound “quite different from the benzene series”. In 1947, his colleague, Katsura, described the isopropyl tropolone structure for hinokitiol at a meeting of the Formosan Medical Association and published it in medical journals. In 1948, Nozoe was repatriated to Japan to become a professor at Tohoku Imperial University in Sendai. There he finally learned of Erdtman's thujaplicin work and of Dewar's 1945 proposal. An account of his work in Formosa, given as lectures in Osaka and Tokyo, was published in Japanese in 1949, and in English translation in 1950.11 Erdtman had suspected the identity of hinokitiol and β-thujaplicin and this identity was directly confirmed by Nozoe in reports of 1949 and 1950, and by Erdtman.10,11
Is there good reason to acknowledge that the first tropolone structure was presented by Nozoe? His early work in Formosa was hindered by the unavailability after 1937 of European chemical journals and texts; by 1941 American materials could not be obtained. His publications in Japanese and in not well-known journals were similarly unavailable to European and American chemists during World War II. Even in 1951, neither his original papers nor abstracts were available to Cook and Loudon for their review.6 His later work from about 1950 slowly received wide attention and acclaim. However, the details of the early Formosa work remained unclear until the publication of his autobiography in 1991.10 This work was originally hand written in Japanese but was issued in translation. Another valuable source of information is the article, “In Memoriam”, written by former students.9 The account given here is largely based on these works.
There is no doubt that Nozoe independently deduced a correct tropolone structure for hinokitiol (β-thujaplicin) and that this was published by Katsura in 1947–1948, well before Nozoe was aware of the work of Dewar and Erdtman. As previously noted, a seven-membered ring structure 3 was published for hinokitiol at a time when the composition was believed to be C10H14O2. Although it is regrettable that there is apparently no written confirmation, Nozoe states that he first presented the correct tropolone structure for hinokitiol at a meeting in Formosa in 1941. Since there is no reason to doubt Nozoe's integrity, it appears that he should be credited as the first to describe a natural tropolone (not then so named) four years prior to Dewar's seminal paper. Nozoe was a highly dedicated chemist, publishing over his long lifetime 402 research papers and 46 other articles (book chapters, reviews, etc.). Had the war not interrupted his research (and note that 1941 was the year of the attack on Pearl Harbor) there is little doubt that he would have published the correct structure for hinokitiol in the usual literature, perhaps as early as 1943. This reappraisal of Nozoe's role in no way detracts from the achievement of Dewar; in fact, Nozoe was reluctant to claim priority in later years. Thus, in an important paper in Nature in 1951 he wrote of “compounds having a tropolone nucleus…as initially proposed by Dewar”.12
Although in 1955 tropolones were described as relatively scarce in nature,7 it can be estimated that more than 200 are now known. The structures of natural tropones and tropolones cover a wide range from tropolone itself to complex multicyclic products and some materials regarded as alkaloids . There is no generally agreed classification system. Tropolonoids are commonly found in fungi and plants; a few have been isolated from bacteria.
When the oak Quercus sessiliflora is infested with the gall-producing insect Dryophanta taschenbergii (Hymenoptera), it produces “galls”—abnormal swellings or growths. This phenomenon is observed in many other plants in response to the activities of various fungi, insects and mites. The Q. sessiliflora galls contain a red-pigmented diglucoside, of which the aglycone is a material named purpurogallin. This material is also readily obtained synthetically by oxidation of pyrogallol. The structure of purpurogallin remained enigmatic until Barltrop and Nicholson in 1948 proposed correctly that it was a benzotropolone.22 These authors noted the close structural relationships with stipitatic acid and colchicine. This deduction is yet one more instance of the influence of Dewar's proposal. Purpurogallin8 R1 = R2 = R3 = H, occurs widely. Purpurogallin carboxylic acid, 8 R1 = R2 = H, R3 = COOH, a biosynthetic intermediate (see later) is less well known but has recently been reported from “teapigment”. More complex structures that contain a purpurogallin unit are discussed later.
Fomentariol 8 R1 = R3 = –CHCH–CH2OH, R2 = H from Fomes fomentarius (a fungus) is a purpurogallin structure modified by two hydroxy-3-propenyl units, one each on the phenyl and tropolone moieties.23 Another structure related to that of purpurogallin is goupiolone A 9 isolated from the aerial parts of Goupia glabra, a plant of the Amazon region of Peru. It is accompanied by a more complex material, goupiolone B 10, apparently derived from an unusual Diels–Alder reaction between a tropolone unit and a benzyne intermediate formed from a naphthalenoid unit with four hydroxyl groups.24 Aurantricholone 11 from the fungus Tricholoma aurantium is another benzotropolone of the purpurogallin type, but having only two hydroxyl groups on the benzene ring.25
A number of fungi (e.g., P. aurantio-virens, P. cyclopium-viridicatum, P. johannioli, P. puberulum) produce puberulonic acid 16 R1 = COOH as anhydride with COOH at C-4, R2 = H, R3 = OH, and puberulic acid 16 R1 = R2 = H, R3 = OH, but no other tropolone compounds.2 The dicarboxylic acids, stipitatonic acid and puberulonic acid, are actually isolated as the anhydrides .
Another structural variation in this series is the presence of further OH groups on the cycloheptatriene ring.36 Examples are 7-hydroxy-3-isopropyltropolone (α-thujaplicinol) 22 R1 = CH(CH3)2, R2 = H from Cupressus pygmaea; 7-hydroxy-4-isopropyltropolone (β-thujaplicinol) 22 R1 = H, R2 = CH(CH3)2; 7-hydroxy-3-isopropenyltropolone (α-dolabrinol) 22 R1 = C(CH3)CH2, R2 = H from C. pygmaea. Related structures with methoxy groups are pygmaein23 R1 = CH3, R2 = H and isopygmaein23 R1 = H, R2 = CH3.38 Compounds with an oxygen atom not directly linked to the cycloheptatriene system are nootkatinol 24 from Chamaecyparis nootkatensis36 and 4-acetyltropolone, a minor component from Thujopsis dolabrata 18 R1 = R3 = H, R2 = COMe.39 β-Thujaplicin-producing cell suspension cultures of Cupressus lusitanica (see later) produce the methyl ether of β-thujaplicin (2-methoxy-6-[methylethyl]cyclohepta-2,4,6-trien-1-one) under certain conditions.40
Troposulfenin 26 is only poorly characterized and in any case is a simple prototropic tautomer of thiotropocin 25. It has been implied indirectly that these two materials are identical.41c,41d Moreover, Laatsch has cited unpublished observations of Liang indicating an identity between thiotropocin 25 and tropodithietic acid 27.44 This conclusion is based largely on an X-ray crystal structure for 27 and, also, on a reinterpretation of the 13C NMR data for these two compounds (A. Zeeck, personal communication to RB). A possible source of difference between 25 and 27 is that while the former gave a bromobenzyl derivative the latter did not. However, shortage of material may have prevented the development of appropriate experimental conditions (A. Zeeck, personal communication to RB).
A compound found in Burkholderia cenocepacia (a Gram-negative bacterium formerly Pseudomonas cepacia, ATCC 17759) contains two tropolone rings linked by a single sulfur atom at the 7 and 7′ positions—R–S–R, R = tropolone .45
Streptomyces echinoruber sp. nov. produces several pigments, of which the most abundant is rubrulone 44.59 The structure contains one ring containing five carbons and one nitrogen.
Colchicine, a tricyclic alkaloid , contains a trimethoxyphenyl ring linked to a seven-membered (cycloheptane) ring carrying an acetamide unit, that is itself linked to a tropolone moiety. The correct structure was finally proposed by Dewar in 1945.5 There are many possibilities for variation in this general structural type, and indeed, a large number of colchicine-related materials occur naturally.60 Only a selection of these minor alkaloids can be given here but there are several, more comprehensive treatments.61–63 The chemical synthesis of colchicine has attracted much interest and even today presents a challenge.64 This major topic will not be considered here.
Commonly found variations on the basic colchicine structure 45 R1 = R2 = Me, R3 = COMe, involve methylation and the amide function as listed below:
Several compounds containing a benzyl unit on the nitrogen atom are known. A typical structure is that of speciosamine46 R1 = R2 = R3 = Me, R4 = H.
Other examples are as follows:
There is a small number of related, non-nitrogenous compounds, for instance in Colchicum ritchii:65colchicone47 R1 = R2 = Me; demethylcolchicone47 R1 = H, R2 = Me; cornigerone47 R1(R2) = –CH2–. In addition to the above compounds, colchicine can occur as the glucoside, colchicoside 48. The common form is 3-O-demethylcolchicine-3-O-β-D-glucopyranoside, but the α anomer has also been reported.66
In the earliest experimental work on tropolone biosynthesis, stipitatic acid16 R1 = R2 = R3 = H was used since it was readily available by growth of Talaromyces stipitatus (then termed Penicillium stipitatum) on simple media. Following its identification, stipitatonic acid 16 R1 = R3 = H, R2 = COOH also became available from this organism.29,67 In 1958, radioactivity from [1-14C]-D-glucose was shown to be readily incorporated into stipitatic acid and chemical degradations of the labelled metabolite indicated that the observed distribution of 14C was not that expected from a shikimate pathway.68 Utilization of acetate was also shown (see later). In similar work, the utilization of [1-14C]-D-glucose was confirmed and the molar specific activity of isolated stipitatic acid was five times higher than that of phenylalanine and tyrosine.69
For chemical degradation, C-9 and C-8 were readily obtained by decarboxylation of stipitatonic and stipitatic acids (Scheme 1). The known conversion of stipitatic acid to malonic and aconitic acids was intensively examined to define optimum conditions, as was the alkaline isomerization, stipitatic acid → 5-hydroxyisophthalic acid.70 Although it was initially assumed that in this isomerization both C-1 and C-2 of the tropolone ring would be extruded, it was shown that, in fact, only C-1 was involved.70–72
Scheme 1 Chemical degradation of stipitatonic and stipitatic acids. For acetate, CH3 = † and COOH = *; for malonate, CH2 = ‡ and COOH = #; one carbon unit from S-adenosylmethionine = ⁁. |
Soon after the identification of stipitatic acid as a tropolone , a possible biosynthetic mechanism by a one-carbon addition to a polyhydroxy benzenoid structure, followed by a ring expansion, had been proposed by several authors; early citations are given in reference 73. In fact, substantial utilization of 14C labelled formate and formaldehyde was also observed in the early experiments. In addition, both [1-14C]- and [2-14C]-acetate were well utilized; [14C]-formate labelled only C-7 of stipitatic acid, and [1-14C]-acetate labelled predominantly C-4 and C-6. Utilization of [1-14C]-acetate was confirmed, but location of the isotope was not determined.69 The cyclization of an octulonic acid phosphate, CO2 fixation, as well as a C1 addition to a benzenoid structure were considered as possible biosynthetic mechanisms.69
Since the results pointed to a polyketide + C1 addition, followed by a ring expansion, malonic acid was also examined as a precursor; it, too, was well utilized, as expected for a polyketide process.73 In detailed studies, [1-14C]-acetate contributed four equally labelled carbons, C-2, -4, -6, -9, to stipitatonic acid and only three to stipitatic acid.74 On the other hand, [1,3-14C2]-malonate labelled C-2, -6 and -9 of stipitatonic acid to the extent of 30% each, but C-4 contained substantially less 14C, only 12%. In later work, where C-2 was unambiguously obtained, [1-14C]-acetate was incorporated to the extent of 25% at that position.72a From these results, it became clear that tropolone formation in T. stipitatus could be represented as follows: 1 acetyl-CoA + 3 malonyl-CoA + 1 C-1 unit → nine carbon tropolone + 3 CO2.
The early hypothesis for stipitatic acid formation had postulated the decarboxylation of a tropolone 3,4-dicarboxylic acid. When stipitatonic acid, a 4,5-dicarboxylic acid, was isolated (actually as an anhydride ) in 1959,29 this seemed a more likely possibility. In fact, an enzyme preparation from T. stipitatus, prepared by grinding mycelium with glass beads in phosphate buffer, was obtained75 and shown to decarboxylate stipitatonic acid to stipitatic acid (see later). These preparations also decarboxylated puberulonic acid suggesting that in both series, the dicarboxylic acids are the precursors to the monocarboxylic acids.
Important questions were the nature of the acceptor for the C1 unit and the identity of the transferred moiety. It was likely that the C1 unit was from the methyl group of methionine, actually in its biologically active form, S-adenosylmethionine. Evidence for this supposition was that tropolone formation in T. stipitatus was inhibited by ethionine, leading instead to the formation of the polyketides triacetic acid lactone 49 R1 = H, R2 = Me and tetraacetic acid lactone (6-[2-oxopropyl]-4-hydroxy-2-pyrone) 49 R1 = H, R2 = MeCOCH2, and a small amount of methyl triacetic acid lactone 49 R1 = R2 = Me.76 The latter had also been reported from cultures not treated with ethionine.77 The formation of these classical polyketide structures under ethionine inhibition considerably strengthened the polyketide + C1 pathway, and suggested that C1 addition, probably a methylation, took place prior to aromatization. The role of triacetic acid lactone in fungal metabolism has been discussed in detail.78
The ethionine-inhibited cultures also contained some orsellinic acid50 R = COOH and orcinol50 R = H, but both of these materials were apparently derived from tetraacetic acid lactone.76 Neither orsellinic acid nor orcylaldehyde were able to function as tropolone precursors in the fungal system. However, 3-methylorsellinic acid 51 was found to be a good precursor when [3-14CH3]- and [1-14COOH]-radiomers were studied as precursors in T. stipitatus. The [3-14CH3]-3-methylorsellinic acid labelled both stipitatonic and stipitatic acids but the carboxyl labelled material labelled only C-9 in stipitatonic acid.79 At this point, the biosynthetic pathway could be summarized as follows: 1 acetyl-CoA + 3 malonyl-CoA + AdoMet → 3-methylorsellinic acid → stipitatonic acid → stipitatic acid + CO2. In the initial condensation, 3 CO2 are lost.
The oxidative mechanism for the ring enlargement of 3-methylorsellinate was envisioned as either the action of a mono-oxygenase or a dioxygenase enzyme (Scheme 2).80 When T. stipitatus was grown in an 18O2 enriched atmosphere, stipitatonic acid and its derivatives gave no evidence for any M + 4 species in high resolution mass spectrometry. An M + 2 peak indicated the presence of a single atom of oxygen. This was the expected result for a mono-oxygenase mechanism both for the enlargement, benzenoid → tropolonoid, and the oxidation, CH3 → COOH (actually as an anhydride ). The arguments relating to possible oxygen exchange reactions are complex and the original text should be consulted. This work did not specifically locate 18O at C-1 of stipitatonic acid as indicated by 52.
Scheme 2 Ring enlargement of 3-methylorsellinic acid. Monooxygenase mechanism, top line. Dioxygenase mechanism, bottom line. |
Many years later, the specific incorporation of 18O2 into stipitatic acid was studied in shake cultures of T. stipitatus with the aid of 13C NMR; the 13C NMR spectrum of stipitatic acid having been completely assigned.81 There was an 18O-induced isotope shift for the C-1 carbonyl (0.04 ppm upfield of the 13C16O resonance) but no observable 18O shift associated with the signal at C-6. These results strongly supported the previously suggested monooxygenase enlargement mechanism.
The conversion of the C-6 methyl group, derived from 3-methylorsellinic acid, into the C-9 position of the anhydride ring has been investigated and two likely intermediates have been identified (Scheme 3). Stipitalide 17 R = H had been isolated from shake cultures of T. stipitatus.30 Moreover, [14CHO]-3-methylorcylaldehyde 53 was a more efficient precursor (ten fold) for stipitatonic acid than was [14COOH]-3-methylorsellinic acid 51.31 While no clear reason for this fact was available, it did make possible the identification of a further intermediate, stipitaldehydic acid 17 R = OH, as well as a trace of stipitalide 17 R = H. Labelled stipitaldehydic acid was shown to be further converted to stipitatonic acid. The late stages of stipitatonic acid biosynthesis appear to be as follows: 3-methylorsellinic acid (or 3-methylorcylaldehyde) → stipitalide → stipitaldehydic acid → stipitatonic acid.
Scheme 3 Formation of stipitalide, 17 R = H and stipitaldehydic acid, 17 R = OH. |
Environmental factors influencing tropolone formation in growing surface and replacement cultures of T. stipitatus 1006 were described in 195969 and work with fungi growing in continuous culture has also been reported.82
There are two confusing factors in considering the biosynthesis of the puberulonic/puberulic pair. In the first place, if the four contiguous oxygens are treated as OH groups (i. e., ignoring the single CO), both of these compounds have an axis of symmetry. For the dicarboxylic acid anhydride structure of puberulonic acid, the IR spectrum suggests that the solid material is 54a (anhydride of 6,7-dihydroxytropolone-3,4-dicarboxylic acid). However, by rotation of the molecular model about the axis shown in 54a, structure 54b is derived and, because of the tautomerism of CO → C–OH, can be written as the anhydride of 6,7-dihydroxytropolone-4,5-dicarboxylic acid 54c. This is the structure usually used in biosynthetic considerations because of the relationship with stipitatonic acid.
The second difficulty was that an initial study of tropolone biosynthesis in Penicillium aurantio-virens, using labelled acetates and formate, indicated that C-9 of puberulonic acid (numbering as in 54c) was not derived from [1-14C]acetate (as is the case for stipitatonic acid) but from an unidentified one carbon pool.83 However, in a later feeding of [1-14C]acetate to P. aurantio-virens, a distribution pattern of 14C at C-2, -4, -6, and -9, analogous to that found in stipitatonic acid, strongly implied that stipitatonic acid and puberulonic acid 54d are formed from the same set of precursors.84 Moreover, small amounts of stipitatonic acid and methyl triacetic acid lactone 49 R1 = R2 = Me were present in the P. aurantio-virens cultures; however, labelled stipitatonic acid was not converted to puberulonic acid (e.g., by hydroxylation). Surprisingly, neither [14COOH]-3-methylorsellinic acid nor the corresponding [14CHO]orcylaldehyde were utilized for puberulonic acid formation. Furthermore, [14COOH]-3-methylorsellinic acid did not contribute radioactivity to the stipitatonic acid formed by P. aurantio-virens as it does in T. stipitatus. Despite these difficulties, it seems likely that puberulonic/puberulic acid formation follows the pattern of the other fungal tropolones , requiring only the additional hydroxylation at C-7.
Sepedonin 6, formed by Sepedonium chrysospermum, contains two more carbon atoms than do stipitatonic and puberulonic acids. Using 13C-labelled substrates (acetates, formate) and determination of label positions by proton NMR spectroscopy, as well as experiments with 14C labelled precursors, indicated a labelling pattern 55 generally analogous to that of the other fungal tropolones .85 In particular, C-8 of sepedonin (sepedonin numbering as in 55) was labelled exclusively by formate. In this case, the conventional polyketide acceptor is CH3CO(CH2CO)3CH2CO–S–Enz rather than CH3CO(CH2CO)2CH2CO–S–Enz for the nine carbon fungal tropolones .
Although experimental validation is lacking, it seems likely that the malettinins 30, 31 and chaetospiron 32 are also typical polyketide metabolites.
The enzyme activity was specific for tropolone-4,5-dicarboxylic acids; tropolone 3,4-dicarboxylic acid, and its 5-, 6-, and 7-hydroxy derivatives, were not substrates. These materials, however, showed inhibitory properties, the highest inhibition, 60%, being obtained with tropolone-3,4-dicarboxylic acid and its 6-hydroxy derivative. The highest specific activity of the enzyme was found in the mycelia of 8-day old cultures, considerably before the maximum formation of the tropolone acids. The enzyme has been assigned the EC number, 4.1.1.60.
Exhaustively dialyzed enzyme preparations produced labelled stipitatic acid from [14CH3]-S-adenosylmethionine without the addition of any organic substrate; potential methyl group acceptors had little or no effect on the system. Radio-activity from labelled acetate was incorporated into the enzymatically active protein , suggesting the presence of an enzyme bound polyketide that could function as methyl group acceptor. The protein was estimated to have a molecular mass of 105000.
It is likely that more detailed information on the enzymology of stipitatonic and stipitatic acid biosynthesis will become available as a result of the determination of the genome sequence for T. stipitatus. While not itself a pathogen, this organism is a close relative of Penicillium marneffei, an organism causing invasive disease in immunocompromised humans. The work involves groups headed by W. C. Nierman at TIGR and A. Andrianopoulos of the University of Melbourne (http://www.tigr.org/faculty/Nierman_William/index.shtml; http://www.uninews.unimelb.edu.au/articleid_3206.html). The sequence was said to be complete by May 2007 (http://fungalgenomes.org/blog/). With the genome sequence available, it should be possible to locate genes for polyketide synthases, decarboxylases and methylases and hence to facilitate the identification of enzymes.
In recent work, 3-methylorcinaldehyde has been identified as the polyketide -derived intermediate for xenovulene biosynthesis via tropolonoid intermediates (see Section 2.10). For the role of this compound in stipitatonic acid biosynthesis see Section 3.
3-Methylorsellinic acid is also a well-known component of lichen depsides and depsidones.92 Thus, the phenolic group of the depside atranorin is the methyl ester of 3-methylorsellinic acid. Moreover, cells of the lichen Evernia prunastri, immobilized in calcium alginate, produced 3-methylorsellinic acid when a specific oxidase enzyme was inhibited by sodium azide.93
In further work, [10-14C]geraniol was incorporated into β-thujaplicin by C. lusitanica cell cultures, indicating a biosynthetic route via geranyl pyrophosphate.97 Moreover, this new work indicated a poor utilization of [2-14C]mevalonate; in earlier experiments the purification of the β-thujaplicin may have been incomplete. NMR analysis of β-thujaplicin samples labelled by growth in the presence of [1-13C]-, [2-13C]- and [U-13C]-glucose indicated that the geranyl pyrophosphate was mostly derived by the mevalonate-independent pathway; i.e., by reaction of glyceraldehyde 3-phosphate + pyruvate to form 1-deoxy-D-xylulose phosphate. It is probable that both a mevalonate pathway and a mevalonate-independent pathway may be involved. The acetate/mevalonate pathway is apparently localized in the cytoplasm whereas the mevalonate-independent route is located in plastids; the two pathways are possibly cross-linked.94
Furthermore a new monoterpene , (1S,2S,6S)-(+)-1,6-epoxy-4(8)-p-menthen-2-ol 57, has been isolated from C. lusitanica cultures (Scheme 4); it was accompanied by ten other known monoterpenes , including terpinolene 56 and sabinene 60.98 It seems likely that this new compound is an intermediate in β-thujaplicin biosynthesis.99 The mechanism by which a single methyl group (e.g., in (1S,2S,6S)-(+)-1,6-epoxy-4(8)-p-menthen-2-ol 57) is incorporated into the seven-membered tropolone ring is currently not known. A major competing pathway is believed to be as follows: geranyl pyrophosphate → 4-terpineol 58 → 4-hydroxyphellandric methyl ester 59. The situation is complex and other terpene synthases such as those for sabinene, β-ocimene, and myrcene may also be involved. In view of the large number of monoterpenes produced by the elicited cultures of C. lusitanica, assessing the precise intermediates involved in β-thujaplicin biosynthesis will be difficult and time consuming. It will be of interest to determine whether there is a common pathway for the biosynthesis of all three thujaplicins.
Scheme 4 Biosynthesis of β-thujaplicin and related compounds. |
One consideration is that there is a feedback regulation of β-thujaplicin production in the cell cultures. When phytoalexin levels reached about 40 mg L−1 in the culture medium, production apparently halted but could be restored by exchange of old medium with new. Under these conditions, a methyl ether of β-thujaplicin was formed; apparently, the excess accumulation of β-thujaplicin was relieved by this metabolic regulation.101
The effects of organic acids (e.g., those of the tricarboxylic acid cycle) on β-thujaplicin production were also investigated. Some small stimulatory effects were noted (e.g., with 1 mM fumaric acid) but there were also inhibitory effects (e.g., 2 mM sodium acetate, 2 mM sodium benzoate).94 However, addition of terpinyl acetate and 2-carene stimulated β-thujaplicin production by 2 to 2.5 fold; addition of other monoterpenes was inhibitory. The metabolism of terpinyl acetate was believed to be as follows: terpinyl acetate → 4-terpineol → thujane → 3-thujone → β-thujaplicin. The action of 2-carene was explained by a different pathway via 3-caren-5-one.
A very active research area concerns the signalling mechanisms indicating the presence of potential pathogens to plant cells and to the activation of appropriate defense mechanisms. In many cases, such responses are initiated by “elicitor” molecules that can be microbial or plant-derived (biotic elicitors) or result from chemical or mechanical stresses (abiotic inducers). This report cannot attempt a detailed account of this extensive topic but will briefly note some of the experimental work dealing with β-thujaplicin production. There is, however, an extensive and recent account of elicitor signal transduction leading to the formation of plant secondary metabolites.102 Also published recently is an account of the molecular regulation of induced terpenoid synthesis in conifers.103
In cell cultures, β-thujaplicin production can be stimulated by a yeast elicitor, methyl jasmonate, and other stresses. The commonly used yeast elicitor preparation contains a 70–80% ethanol-insoluble, polysaccharide component. Elicitor-treated cultures accumulate high amounts of β-thujaplicin at early stages (e.g., 4 days of incubation) and then accumulate other monoterpenes at later stages post-elicitation. Thujaplicin may be the primary defense agent with the other monoterpenes providing a secondary defense.99 The yeast elicitor generally produces an earlier and stronger response than does the octadecanoid plant growth regulator, methyl jasmonate. The content of articles considering many other factors involved in β-thujaplicin production by the C. lusitanica system (e.g., in signalling, biosynthesis and degradation), are noted here for the reader's convenience:
Hydrogen peroxide formation by superoxide anion synthases may mediate elicitor-induced biosynthesis; exogenously applied H2O2 (or use of a generation system) can stimulate biosynthetic activity.104
cAMP stimulates β-thujaplicin biosynthesis, possibly by an involvement of Ca2+ and K+ fluxes. There is cross talk between cAMP treatment and the ethylene signalling pathway.105
In addition to the just-noted role for ethylene signalling, jasmonate signalling is also involved and there is an interaction between these two modes; they are integral parts of the C. lusitanica signalling pathway.106
A further complexity is the role of oxidative stress; under conditions of Fe2+ stress C. lusitanica cells generate significant levels of reactive oxygen species, ROS. The ROS production inhibits cell growth but enhances ethylene and β-thujaplicin formation; as previously noted, H2O2 is a positive signal for β-thujaplicin formation.107
More recently, the interactions (cross talk) between ROS and nitric oxide have been delineated. Nitric oxide is involved in multiple physiological processes and plays a role in the elicitor-induced cell death.108
The metabolic flux of β-thujaplicin formation and that of other monoterpenes in the elicited cultures has been investigated.109
Yeast elicited cultures of T. plicata also show an increased stimulation of thujaplicin formation, but less detail is presently available than for the C. lusitanica system.95
Since the 16 carbon atoms of colchicine derived from a C6–C3 and a C6–C2 unit it was suggested that a dienone 61 with a good leaving group, X, and with the possibility for ring expansion might be involved in the biosynthetic mechanism.114 An important finding was the structure of androcymbine62 R = H, a cometabolite of colchicine in Androcymbium melathiodes. In fact, 3H labelled O-methylandrocymbine, 62 R = CH3, was an excellent precursor for colchicine. In turn, a possible precursor for O-methylandrocymbine was the diphenol 63, a compound termed autumnaline; this too was well incorporated into colchicine and demecolcine. A basic pathway, omitting various hydroxylation and methylation steps, could, therefore, be written to demecolcine 45 R1 = R2 = R3 = Me and hence colchicine 45 R1 = R2 = Me, R3 = COMe (Scheme 5).
Scheme 5 Biosynthesis of colchicine. |
It was established that the N-methyl group of autumnaline and O-methylandrocymbine is retained in demecolcine, but was lost in the formation of colchicine. With appropriately labelled precursor molecules, the following detail was obtained for the O-methylandrocymbine → colchicine portion (Scheme 6): O-methylandrocymbine 62 R = CH3 → N-formyldemecolcine64 R1 = Me, R2 = CHO → demecoline64 R1 = H, R2 = Me → N-formyl-N-deacetylcolchicine64 R1 = H, R2 = CHO → N-deacetylcolchicine64 R1 = R2 = H → colchicine64 R1 = H, R2 = Ac. The N-formyl group of N-formyldemecolcine derives from the C-3 carbon of autumnaline-marked as * in 63. (Of these intermediates, N-formyl-N-deacetylcolchicine, and N-deacetylcolchicine are known as minor alkaloids of C. autumnale and N-formyldemecolcine is present in C. cornigerum).
Scheme 6 Biosynthesis of colchicine: further details. AdoMet = S-adenosylmethionine. |
Further information has come from the use of a microsomal enzyme system present in immature seeds of C. autumnale.115 This system, dependent on NADPH, cytochrome P-450 and O2, catalyzed the intramolecular phenol-oxidative coupling of autumnaline 63 to a new material, termed isoandrocymbine, 65 (Scheme 6). On methylation with diazomethane, 65 was converted to O-methylandrocymbine 62 R = Me. Moreover, isoandrocymbine was converted to androcymbine by a soluble protein extract from C. autumnale seeds in the presence of [14CH3]-S-adenosylmethionine.116 Still more recently, isoandrocymbine 65 and O-methylandrocymbine 62 R = Me were isolated as colchicine cometabolites from C. stevenii, a Jordanian meadow saffron.
Scheme 7 Biosynthesis of sulfur-containing tropolonoids. * = 13C from [1,2-13C2]phenylacetate. |
As indicated earlier, purpurogallin, the simplest of these benzotropolones, is readily obtained from pyrogallol by either chemical or enzymatic oxidation. In vivo, gallic acid68 undergoes decarboxylation to pyrogallol69 followed by a complex enzymatic oxidation (Scheme 8). Purpurogallin is essentially a shikimate-derived product.
Scheme 8 Biosynthesis of purpurogallin. |
The leaves of Camellia sinensis are used to produce more than 300 different kinds of tea, with three main types: green tea (non-fermented), oolong tea (semi-fermented) and black tea (fermented). Fermentation is a rather inappropriate term since no new biological components (e.g., yeast) are added in the process. It is an oxidation, brought about by polyphenol oxidases present in the leaves. For green tea, this oxidation is prevented by steaming or drying the fresh leaves at elevated temperatures to inactivate oxidases. Black tea having experienced polyphenol oxidase activity contains various pigments including the orange-red coloured theaflavins and the red or dark brown thearubigins (heterogeneous polymers of flavan-3-ols and flavan-3-ol gallates). The theaflavins contain a benzotropolone nucleus.
Perhaps theaflavins should not be considered as true “natural products”; however, this distinction is not usually enforced. Green tea, closest to the natural leaves, is essentially free of theaflavin components. In a detailed HPLC analysis of 48 green teas, theaflavin components were reported as “not detected” with a single exception where the amount of theaflavin itself was 0.2 ± 0.02 mg g−1.120 Moreover, in this one case, theaflavin gallates were not detected. Typical structures are as follows: theaflavin72 R1 = R2 = H; theaflavin 3-gallate 72 R1 = galloyl, R2 = H; theaflavin 3′-gallate 72 R1 = H, R2 = galloyl; theaflavin 3,3′-digallate 72 R1 = R2 = galloyl (Scheme 9).
Scheme 9 Biosynthesis of theaflavin 72 R1 = R2 = H and epitheaflavic acid 73. |
Theaflavins are formed by co-oxidation of catechins during the fermentation process. Thus, for example, theaflavin itself 72 R1 = R2 = H, was formed by incubation of (−)-epicatechin 70 and (−)-epigallocatechin 71 with a crude polyphenol oxidase preparation from bananas (Scheme 9).121 From (−)-epicatechin 70 and gallic acid68, under influence of a polyphenol oxidase, there was obtained epitheaflavic acid 73.122 In other work, 18 benzotropolone derivatives were obtained by reaction of selected pairs of catechins and other materials (e.g., catechol, pyrogallol, gallic acid) with H2O2 in the presence of horseradish peroxidase. For structures, the reader is referred to the original paper.123
Theaflavins can be synthesized from epicatechin and epigallocatechin by many plants, even those not themselves containing catechins. Using homogenates from 62 plants (49 families), 46 were capable of theaflavin synthesis.124 Japanese pear (Pyrus pyrifolia) and unripe loquat (Eriobotrya japonica) homogenates were particularly active.
Polymeric compounds with higher molecular masses were obtained by reaction of theaflavins and tea catechins in the presence of horseradish peroxidase and have been identified in black tea itself. To give only one example, reaction of theaflavin 3-gallate 72 R1 = galloyl, R2 = H, with epicatechin gave theadibenzotropolone 74.125 Moreover, a theatribenzotropolone was also obtained. A different structural type of benzotropolone, theaflavate A 75 R = galloyl, was obtained by chemical oxidation of (−)-epicatechin-3-O-gallate with potassium ferricyanide and was also shown to be present in small amounts in black tea.123,126 Theaflavate B 75 R = H lacks the galloyl residue.
Compound | Activity | References |
---|---|---|
β-Thujaplicin, 18 R1 = R3 = H, R2 = isopropyl | Antimicrobial | 131,132 |
Antifungal | 130–132 | |
Insecticidal | 129 | |
Cytotoxic (human, murine cells) | 137 | |
Antitumor (due to metal chelation) | 134 | |
Chlamydia trachomatis inhibition | 135 | |
α-Thujaplicin, 18 R1 = isopropyl, R2 = R3 = H | Antibacterial | 136b |
Antifungal | 136b | |
Plant growth inhibitor | 136a | |
Insecticidal | 136b | |
Ascaricidal | 136b | |
Cytotoxic (murine cells) | 136a,b |
The related compounds, β-dolabrinol, γ-thujaplicin and 4-acetyltropolone (all components of Thujopsis dolabrata) have various antibacterial and antifungal activities, and are cytotoxic in vitro to the murine P 388 lymphocytic leukemia cell line.137 In a separate study of 4-acetyltropolone the following activities were observed: metalloprotease inhibition, plant growth inhibition, cytotoxicity, toxicity to mice.39
It has long been recognized that certain woods from trees of the Order of the Cupressales have a natural durability, and this durability is attributed to tropolone compounds. A present day USA catalogue devoted to products for “eco living with style” notes that hinoki wood and its oils are prized in Japan “for their ability to pamper and heal the skin, kill bacteria, and smooth frazzled nerves”. Moreover, hinoki wood is prized in areas exposed to moisture since “it resists mildew”. Among several hinoki wood items available for sale is a sauna stool for US $98 (Viva Terra catalogue, summer 2007).
As already indicated, the physiological activities of the thujaplicin group of tropolones do include fungicidal and insecticidal activity. Indeed, the commercial use of tropolones as substitutes for toxic wood preservatives (e.g., arsenic compounds) that are already in use, has been seriously considered.36 Since extraction of tropolones from wood, or chemical synthesis, is prohibitively expensive for such use, biotechnological methods have been considered. Cell cultures from Thuja plicata are one possibility95 and the extensive work on the C. lusitanica cell system has already been discussed.
There appears to be little or no use of β-thujaplicin in Europe and the USA. The USA FDA has no documents relating to hinokitiol/β-thujaplicin as of March 2007. Nevertheless, an Internet source indicates that there are 312 USA patents in some way involving hinokitiol. To take one example only, US Patent 7081258 describes its use in a hair growth-promoting composition. Visitors to Russia may be interested to know that, according to Internet sources, the Grand Hotel Europe in St Petersburg has available various cosmetology services including those of “Hinoki Clinical”, described as a “deluxe-category therapeutic cosmetics company” in Japan. The active component in this company's products is hinokitiol, the main qualities of which are said to be as follows: powerful antibacterial action, unique anti-inflammatory activity, deep penetration of skin, normalization and stimulation of cell activity, nourishing activity, treatment of pigmentation marks, improvement of skin tone. The “Spa Club” of the Orient Express also features Hinoki Clinical cosmetics in the “Japanese compartment”.
Several tropolone types, including purpurogallin, protect cultured cells from oxidative stress-mediated damage, specifically hydrogen peroxide-induced DNA damage and apoptosis. In cultured Jurkat cells (an immortalized line of T lymphocyte cells originally derived from a 14 year old boy with T cell leukemia), tropolone , β-thujaplicin, purpurogallin and trimethylcolchicinic acid were protective; however, colchicine and tetramethyl purpurogallin ester were not. Hydrogen peroxide-induced apoptosis was also inhibited by tropolone . These results were attributed to the formation of a redox-inactive iron complex. Under some conditions (e.g., absence of exogenous hydrogen peroxide) tropolone enhanced iron-mediated DNA damage, possibly by formation of a lipophilic redox-active complex.139
β-Thujaplicinol 22 R1 = H, R2 = CH(CH3)2 and manicol 7 are potent and selective inhibitors of the HIV-1 ribonuclease H (RNase H) activity of human immunodeficiency virus-type 1 reverse transcriptase. However, β-thujaplicin was inactive. This work emphasized the importance of the 2,7-dihydroxy function in both of these natural tropolones , possibly through a role in metal chelation at the RNase H active site.142
In addition, seven natural tropolones were investigated as inhibitors for the magnesium-dependent strand transfer activity of HIV-1 integrase. The materials were identified by NSC compound numbers as follows (NSC = Nomenclature Standards Committee of the USA FDA): 18806, β-thujaplicinol; 310618, manicol; 43339, nootkatin; 89303, tropolone ; 18804, β-thujaplicin; 18805, β-thujaplicin; 43338, β-thujaplicin. As with the inhibition of the RNase H activity, β-thujaplicinol, and manicol were inhibitors of HIV-1 integrase with Mn2+ as cofactor . The three measurable activities of HIV-1 integrase—3′-processing, strand transfer, disintegration—were studied in detail with β-thujaplicinol; the role of the 7 hydroxy group was again evident. The hydroxy tropolones probably chelate the divalent metal, Mg2+ or Mn2+, at the enzyme's active site. It was also observed that β-thujaplicinol had a weak cytoprotective activity against HIV-1IIIB on lymphoid MY-2 cells.143
In connection with the importance of the 7 hydroxy group, it may be noted that 7-hydroxytropolone 1 R1 = OH, R2 = H, a metabolite of Streptomyces neyagawaensis, is an inhibitor of aminoglycoside-2″-O-adenylyltransferase.15,144 Moreover, 3,7-dihydroxytropolone 1 R1 = R2 = OH, a metabolite of Streptomyces tropolofaciens sp. nov. K611-97, has weak antibacterial activity against a variety of Gram-positive and Gram-negative organisms, weak antifungal activity, but a strong cytotoxicity against murine B16 melanoma cells. It had some antitumor activity against B16 melanoma in mice but was inactive against P388 leukemia.16
Despite its toxicity, colchicine has become famous for the treatment of gout—an inherited metabolic disorder, usually in males, involving problems in uric acidmetabolism and progressive chronic arthritis. A classic gout attack is acute in onset and characterized by extreme pain; there can also be marked swelling, warmth, erythema and tenderness of a single joint. The metatarsophalangeal joint of the big toe is often the afflicted member. Colchicine is also used for treatment of Familial Mediterranean Fever, an inherited disorder, frequently but not exclusively found in persons of Mediterranean origin, and characterized by recurrent fever and peritonitis (inflammation of the peritoneal cavity). Colchicine analogues have been tested for anti-cancer activity but toxicity makes their use problematic.147 In medical use, a dose of 1 mg colchicine orally every 2 hours is used to treat an acute gout attack but no more than 7 mg should be taken in a 24 hour period. For chronic disease, the dose is one to three 0.6 mg tablets per day.
Colchicine does not cure gout and the precise mechanism for pain relief is unclear. Elevated blood levels of uric acid lead to crystallization of poorly soluble urates. These foreign materials are then attacked by macrophages and leukocytes, resulting in the release of cytokines and interleukins, thus leading to the inflammatory condition. Colchicine functions by inhibition of the action of the leukocytes. This probably involves the major pharmacological property of colchicine—the ability to bind to tubulin dimers and thus to prevent polymerization to microtubules.147 The microtubules are necessary for several processes including ameboid motility; inhibition of this activity prevents the migration of macrophages and leukocytes and, in turn, the gout-derived inflammation.
As well as gout and Familial Mediterranean Fever, colchicine is approved by the USA FDA for treatment of secondary amyloidosis and scleroderma. In Europe, the European League Against Rheumatism (EULAR) has issued comprehensive recommendations for both diagnosis and treatment of gout.148 Another recent literature review and consensus statement for German, Austrian and Turkish caregivers is available for colchicine use in children and adolescents with Familial Mediterranean Fever.149
In humans, colchicine halts the division and activity of white blood cells; similarly, in plants, the division of plant cells can be arrested or delayed. Injection of colchicine into a plant ovary blocks formation of the mitotic spindle in the dividing, fertilized egg. In consequence, the cell does not divide but the chromosome number in the egg is doubled. This phenomenon provides a useful tool for the study of polyploidy in plants.
Colchicine contains two elements of chirality so four stereoisomeric structures are possible. The carbon at position 7 (carrying the NHAc substituent) is a typical chiral centre. The other stereochemical element is axial chirality (atropisomerism) about the single bond between rings A and C. Specification of axial chirality as aR and aS is described in a standard text.150 In the natural stereoisomer, (−)-colchicine, that binds strongly to tubulin, C-7 has S configuration. Extensive NMR and ORD studies, and X-ray crystallography of an analogue of the thio series, indicate that the axial chirality in (−)-colchicine is aS.147 Hence natural colchicine is (−)-(aS,7S)-colchicine 76. The aR–aS equilibrium, 76 ↔ 77, studied by NMR, indicates an important role for the cycloheptane ring B. The acetamido group at C-7 energetically favours a pseudo-equatorial orientation, and this in turn favours the aS configuration 78. The activation energy for the atropisomerism is 22–24 kcal mol−1. The two remaining stereoisomers are the atropisomers with the C-7 acetamido group in the R configuration—(+)-(aR,7R) and (−)-(aS,7R).
There has been much study of various analogues of (−)-colchicine, including the role of the MeO groups, the C-7 acetamido group, contraction of the tropolone ring to phenyl, and in “iso” compounds reversal of the substituents at C-9 and C-10.62,63,147 One hope is that the design of colchicine analogues may produce compounds with high tubulin-binding activity but lower toxicity; so far however, no practicable anti-cancer agent based on colchicine has been developed. One example is a recent paper describing variations in the B ring of allocolchicinoids (these “allo” compounds have a phenyl ring in place of the usual tropolone ). Thus allocolchicine 79 and N-acetylcolchinol-O-methyl ether (NCME) 80 are known as potent anti-tubulin agents. Although some further, variant synthetic structures inhibited tubulin assembly as well as NCME, none of them revealed growth inhibition of MCF-7 breast cancer cells comparable to that of colchicine itself.151
The binding of colchicine to tubulin is slow and poorly reversible; the exact mechanism by which it stabilizes tubulin against spontaneous decay is unclear. A recent “footprinting” technique shows that colchicine affects mainly the α-subunit of tubulin, involving cysteine residues at positions 295, 305, 315 and 316. This domain is substantially distant from the binding site at the α/β interface. The B ring of colchicine is bound on the α-subunit and the A and C rings are on the β-subunit. An elegant model is available152 as is an X-ray structure showing the colchicine-binding site on tubulin.153
This journal is © The Royal Society of Chemistry 2008 |