The molecular process of membrane-protein integration is strictly regulated by so-called translocon, a channel composed of a protein complex. Recently, we found a novel integration-stimulating factor termed MPIase (Membrane Protein Integrase) in this integration process on Escherichia coli cells. Surprisingly, this factor was not a proteinous molecule but a sort of glycolipid. In this study, we succeeded in the extraction and purification of MPIase from E. coli strain MC4100 cells and determined its total structure. MPIase was purified from inner-membrane components of the cells, using MonoQ ion-exchange column chromatography and subsequent liquid-liquid partition column chromatography on sephadex LH-20 gel. The constituents of MPIase and its hydrolyzed products were elucidated by 1D and 2D NMR, MALDI-TOF MS, GC-MS and Q-TOF MS. These analyses demonstrated that MPIase contains trisaccharide repeating units composed of G1cNAc, ManNAcA and Fuc4NAc, a pyrophosphate diester linkage, and a diacylglycerol moiety. Furthermore, we compared 13C NMR chemical shift values of MPIase with those of synthetic disaccharide substructures to conclude the glycosidic linkages and sequence of the repeating unit as a-Fuc4NAc-(1→4)-P-ManNAcA-(1→4)-a-GlcNAc(1→3). Consequently, the total structure of MPIase was determined as shown in Figure 1.
The dinoflagellate Karenia brevisulcata (Fig. 2) caused a devastating algal bloom in Wellington Harbor, New Zealand, in early 1998 that killed marine life over a wide area. More than 150 people along the coastline were also reported to have adverse affects. In this presentation, the isolation, biological activity, and structural studies of the causative toxins will be reported. Bulk cultures of K brevisulcata were extracted using 5%#aq. acetone and absorbed on HP-20 resin followed by elution with acetone. The crude extract showed cytotoxicity and mouse lethality. The extract was partitioned between chloroform and aqueous methanol under neutral and acidic conditions. Each fraction contained different toxins. The neutral lipophilic fraction was separated on a diol column guided by cytotoxicity against mouse leukemia P388 cells. Final purification by HPLC on a C30 column led to isolation of four major toxins, tentatively named KBT-F, KBT-G, KBT-H, and KBT-I. (Fig. 3) Their MALDI MS spectra showed their very high molecular weights over 2000. The UV maxima at 227 nm suggested that these molecules have conjugated structures. The IC50 against P388 cells were estimated to be 0.3-5.5 ng/ml. The mouse lethality (i.p.) of KBT-F and KBT-G were 40 and 30 pg/kg, respectively. In the complex proton NMR spectrum of KBT-F, two doublet and eleven singlet methyls, an aldehyde conjugated to a double bond could be observed. In addition, widespread oxygen-bearing methine signals in the spectra suggested polycyclic ether structures (Fig. 5, 6). The planar structures of KBT-F (Cio7Hiso038) and KBT-H (Cio7Hi6o039), each of which have 24 ether rings, 13 hydroxyl groups, 2 doublet methyl, 11 singlet methyl, and an enal side chain (Fig. 8) were successfully assigned by using several 2D NMR measurements and high resolution MALDI MS.
Dinoflagellates are the rich sources of huge polyol compounds. It is intriguing why dinoflagellates produce such huge and complex compounds. However, their physiological functions or roles in ecosystem are scarcely known. To solve this problem, it is needed to find more huge compounds and to clarify the mechanisms of their biological and physiological activities. In our continuous research about the polyol compounds, a novel huge polyol compound, amdigenol A, was isolated from the Okinawan dinoflagellate Amphidinium sp. The structural elucidation of amdigenol A was performed by the combination of 2D-NMR spectroscopic analyses, the chemical degradation with Grubbs catalyst, and ESI-MS/MS analyses. Amdigenol A, whose molecular weight was 2169 mass unit, was consisted of a C98-linear long carbon backbone with two exo-methylene and four methyl group including a sulfate ester, thirty-six hydroxyl group, and three tetrahydropyrans. The structural elucidation of amdigenol A should contribute to identify the other species of huge polyol compounds produced by the same dinoflagellate.
Cyanobacteria have drawn attention for their ability to produce a large number and variety of bioactive compounds. In our ongoing efforts to explore cytotoxic cyanobacterial metabolites, we isolated novel macrolides leptolyngbyalides A-D (1-4) from marine cyanobacterium Leptolyngbya sp. We report here the isolation, structure determination, and bioactivities of leptolyngbyalides and related compounds. Leptolyngbya sp. collected in Okinawa, Japan, was extracted with Me0H. After partition of the extract, the organic layer was subjected to fractionation using column chromatography. Finally, bioassay-guided fractionation afforded leptolyngbyalide A (1) and its analogs (2-4) as colorless amorphous solids. The molecular formula of 1 was determined to be C_<46>H_<76>O_<15> on the basis of ESIMS and NMR data. The planer structures of 1 and its analogs were established by spectral analysis, including 2D NMR techniques. Sugar moiety was determined to be xylose based on the coupling constants of acylated derivative. We are now trying to prepare for co-crystal of actin-leptolyngbyalide A for the purpose of X-ray crystal structure determination. Leptolyngbyalides showed remarkable growth inhibition against HeLa S_3 cells with IC_<50> values of 0.099-0.64 μM. They were revealed to depolymerize F-acin, by monitoring the fluorescent intensity of pyrene-labeled actin. Trypan blue dye exclusion assay and DNA ladder analysis showed that 1-4 induced apoptosis in HeLa cells. In addition, 1 selectively inhibited the phosphorylation of ERK in PDGF-stimulated NRK cells. Based on the above result, we expected that other actin depolymerizing compounds also exhibit apoptosis inducing activity. Aplyronine A (5) and mycalolide B (6), known as actin depolymerizing compounds, were revealed to induce apoptosis against HeLa S_3 cells and HL60 cells. These results made us focus on the potential correlation between depolymerization of actin and apoptosis. Then, we employed synthetic analogs of aplyronine A 7-10 to evaluate the correlation. Growth inhibitory activities of 7-10 were positively correlated with their actin depolymerizing activities, and analog 10 induced apoptosis against HL60 cells. These results suggested the possibility that apoptosis induced by 1-6 were caused by depolymerization of actin, and it is the subject of further study.
Filamentous fungi are rich sources of bioactive compounds, and therefore a huge number of studies to explore bioactive natural products from fungal secondary metabolites have been done. Recent fungal genome sequencing programs, however, have revealed that they contain a greater number of gene clusters encoding uncharacterized secondary metabolites in addition to those of natural products previously isolated from the same fungi. The silent biosynthetic genes generally remain unexpressed under a variety of laboratory culture conditions. The transcriptionally suppressed biosynthetic pathways are anticipated to be good sources of wide range of bioactive metabolites. Recently, HDACs were shown to suppress the production of secondary metabolites in Aspergillus nidulans and secondary metabolism in fungi is strongly affected by epigenetic regulation by chromatin re-modeling through chemical modifications to DNA and histone. Further evidence is that treatment of HDAC or DNA methyltransferase inhibitors with fungi induced the production of secondary metabolites. In order to maximize the opportunity for finding new bioactive compounds from the silent biosynthetic genes in fungi, we introduced this chemical epigenetic manipulation into our natural products exploitation. During screening for the fungal secondary metabolites under various inhibitor conditions, we found the remarkable change of the secondary metabolite profile in culture medium of Torrubiella luteorostrata (SBHA-1 mM) and Cordyceps indigotica (SBHA-1 mM), C. annullata (SBHA-500 μM) and Gibellula formosana (SBHA-1 mM and RG-108-1 mM) by comparison with those negative controls. After incensement, we carried out the isolation and structure elucidation of those EtOAc extracts and succeeded to obtain various new natural products. Furthermore, we found Chaetomium indicum, C. globosum, C. mollipilium, Epicoccum purpurascens showed the significant change of their secondary metabolism under the condition of SBHA 1 mM. C. mollipilium also showed the change by nicotinamide and this is the first example that an inhibitor of NAD+ dependent HDACs affected the secondary metabolism in fungi.
Alstonia pneumatophora and A. angustiloba are members of the Apocynaceae family growing in Malaysia and Indonesia. Traditionally, their leaves have been used for the treatment of periodic fever and malaria. Alstonia species so far have been shown to produce various skeletal alkaloids, including scholaricine-type alkaloids such as echitamidine, uleine-type alkaloids such as undulifoline, and bisindole alkaloids such as villalstonine. In the search for structurally and biogenetically interesting alkaloids from tropical plants found in Malaysia, the alkaloid components of the leaves of A. pneumatophora and A. angustiloba were investigated. As a result, alsmaphorazines A and B, novel indole alkaloids with an unprecedented polycyclic skeleton possessing an oxyamine moiety, alsmaphorazines C - E, novel indole alkaloids with an unprecedented pentacyclic skeleton fused octahydropyrrolo[2,3-b]pyrrole and azabicyclo[3.3.1]nonane moiety, and alpneumines A-H were isolated from A. pneumatophore, and alstilobanines A - E were isolated from A. angustiloba. The structures of alsmaphorazines A - E were determined by 1D & 2D-NMR data (^1H-^1H COSY, HSQC, HMBC, NOESY) and the absolute configuration of alsmaphorazines B and C was assigned by CD spectral analysis and the modified Mosher's method. Biogenetically, alsmaphorazines A - E may be derived from preakuammicine and plausible biogenetic route of alsmaphorazines A - E from preakuammicine is proposed. Indole alkaloids isolated in this research were evaluated for inhibition of the NO production, vasorelaxant effect, and anti-melanogenesis activity. Alsmaphorazine A was found to dose-dependently inhibit the NO production in LPS-stimulated J774.1 cell line without affecting the cell viability, alstilobanine A showed a moderate vasorelaxant activity against phenylephrine-induced contraction of isolated rat aorta and alpneumine G showed anti-melanogenesis activity in B16 mouse melanoma cells.
Conjugation in Blepharisma japonicum is induced by interaction between complementary mating-types I and II, which excrete blepharmone (gamone 1) and blepharismone (gamone 2), respectively. Gamone 1 transforms type II cells such that they can unite, and gamone 2 similarly transforms type I cells. Moreover, each gamone promotes the production of the other gamone. Gamone 2 has been identified as calcium-3-(2-formylamino-5-hydroxybenzoyl) lactate and has been synthesized chemically. Gamone 1 was first isolated and characterized as a glycoprotein of 20-30 kDa containing 175 amino acids and 6 sugars. The 305 amino acid sequence of gamone 1 was determined by Harumoto et al from gamone 1 cDNA, the N-terminus for the protein, however, could not found suggesting the modification of N-terminal amino acid of gamone 1. This time we identified by MALDI-TOFF MS analyses the N-terminus for gamone 1 as glutamine^<28> modified to be pyroglutamine. And also, we determined the structure and the position for the glycan moiety of the gamone 1, using the purified glyopeptides obtained by the chymotrypsin or pronase digestion. Six types of structural heterogeneity was observed for the glycan of gamone 1, and the glycans being attached to N^<141>, N^<149>, N^<165. and N^<213> glutamines which have the consensus sequence of NX[ST].
Phytophthora is a genus of oomycetes and one of the best known agricultural pests. The most notorious species of this genus is Phytophthora infestans, which causes the destructive potato blight disease responsible for the "Great Irish Famine" in the mid 1840s. Sexual reproduction of the heterothallic members require two strains designated as Al and A2 mating types, which secrete unique mating hormones αl and α2, respectively, and regulate sexual reproduction in the counter mating type. Although the chemical characterization of a hormunes had been a long-term issue, al was isolated and characterized by us in 2005. Here we describe the absolute stereostructure of the second mating hormone α2 by spectroscopic analysis and total synthesis. The hormone α2 was found to be a linear diterpene, which is similar to al. The interspecies universality of both αl and α2 in Phytophthora was then demonstrated by hormone response experiments and hormone production analysis with 45 heterothallic strains. In addition, stable isotope-labeling experiments revealed the pathway, in which α2 is biosynthesized from phytol by A2 mating type strains and metabolized to αl by Al.
Pradimicin A (PRM-A) is an antibiotic isolated from actinomycetes. This compound is unique as a non-peptidic natural product with the lectin-like property of being able to recognize D-mannopyranoside (Man) in the presence of Ca' ion. Recently, PRM-A has been attracting attention as a conceptually novel drug candidate for human immunodeficiency virus (HIV). Although the anti-HIV effects are explained by its specific binding to Man residues of glycans on the viral envelope, the molecular basis of Man recognition by PRM-A remains unclear. Conventional interaction analysis of PRM-A with Ca^<2+> ion and Man in solution has been hampered by aggregation of the ternary PRM-A/Ca^<2+>/Man complexes and complicated three-component equilibrium. In this report, we describe the interaction analysis of PRM-A with Ca^<2+> ion and methyl a-D-mannopyranoside (Man-OMe) in the solid state. The use of the solid PRM-A aggregate composed of exclusively the [PRM-A2/Ca^<2+>/Man-OMe_2] complex facilitated the analysis by eliminating the problem associated with the complicated equilibrium in solution. Solid-state ^<113>Cd-NMR experiments using ^<113>Cd^<2+> ion as a surrogate probe for Ca^<2+> ion led us to propose an unprecedented binding model, in which PRM-A binds Man-OMe in a Ca^<2+>"-mediated manner through the carboxylate group of D-alanine moiety. This binding model was further supported by two-dimensional dipolar assisted rotational resonance (2D-DARR) experiments using biosynthetically ^<13>C-enriched PRM-As, which confirmed that the D-alanine moiety and A-C rings of PRM-A constitute the binding site for Man-OMe. The present study provides a clue toward the full elucidation of the molecular basis of Man recognition by PRM-A.
Determination of the molecular targets of natural products has had profound impacts on studies of complex cellular machinery. Natural products are often coevolved with proteins targets and thereby likely to exhibit high selectivity to the human counterparts of those targets. Here we report isolation of a selective protein target of aurilide, a potent cytotoxic marine natural product. The ability of aurilide at low concentrations to induce apoptosis in human cancer cells has encouraged a range of biological analyses. Nonetheless, its mechanism of action has remained unknown. To identify the target of this potent cytotoxic molecule, we biochemically isolated a selective aurilide -binding protein, prohibitin 1 (PHB1). Our mechanistic analyses indicate that the interaction of aurilide with PHB1 in mitochondria activates the proteolytic processing of optic atrophy 1 (OPA1), leading to mitochondrial fragmentation and apoptosis.
The allenic carotenoids, fucoxanthin and peridinin, have been paid much attention as the main light-harvesting pigments in photosynthesis in the sea, because they exhibit high energy transfer efficiencies to Chlorophyll a. This energy transfer efficiency is thought to be related to the intricate structures of these carotenoids. In order to examine their relationship between the unique structures and super ability in carrying out energy transfer, we called to our attention to peridinin, and synthesized peridinin and its derivatives. In addition, ultrafast time-resolved optical absorption and Stark spectroscopic measurements of these peridinin and its modified derivatives have revealed that an intramolecular charge transfer (ICT) state behaves independently from S, state and the S, lifetime of the ICT state converges to a value of 10ア1 ps in methanol for all the peridinin analogues regardless of the extent of a-electron conjugation, and that the allene and C37 carbon skeleton contribute to the generation of a large dipole moment in the excited state of the molecule. A next stage is then to demonstrate the generality of these results by investigating the characteristics of other carotenoids having the ability of high energy transfer efficiency. We focused on fucoxanthin, which possesses the similar ability to peridinin. Fucoxanthin forms the fucoxanthin-chlorophyll a/c-protein (FCP) complex and possesses an allene function and a β,γ-epoxy keto moiety, which is known to be extremely labile to alkali, in a π-electron conjugated system. Only one synthesis of fucoxanthin was reported. The stereochemical control of both the epoxidation and polyene chain formation, however, had not been achieved responsible for this labile moiety. We then started the stereocontrolled synthesis of fucoxanthin, that should be applicable for the syntheses of various kinds of fucoxanthin modified derivatives. Our synthetic strategy had two key points. First, we bisected an allenic segment and a β-ketoepoxy segment to make up the library of each half-segments to synthesize the variously designed fucoxanthin derivatives and coupled them by using the modified-Julia olefination. Second, we intended to realize the satisfactorily stereocontrolled introduction of the cis-epoxide to the C3-a-homoallylalcohol by utilizing the Sharpless epoxidation and then followed by inversion of the resulting unnatural stereochemistry at the C3 hydroxyl group to the corrected one by using the Mitsunobu reaction to control the stereochemistry in the terminal six-membered ring. Herein, we report the stereocontrolled total syntheses of fucoxanthin and polyene chain and allene modified derivatives based on the established method for the natural fucoxanthin synthesis. In addition, ultrafast time-resolved optical absorption measurements were performed on polyene chain modified-fucoxanthin derivatives, which possessed 5 to 8 conjugated olefins. The results showed that the 5, lifetime on fucoxanthin and its derivatives was a similar tendency to that in the case of peridinin. Although the shorter polyene chain derivatives became the longer lifetime in hexane, the S_1 lifetime in methanol converged to a value of 15-22 ps. These results strongly support the unique characteristics of ICT state that we observed in peridinin.
Amyloid β3 (Aβ) aggregation has been associated with the neurodegenerative pathology and a cascade of harmful event related to Alzheimer's disease (AD). Inhibition of Aβ aggregation and attenuation of the AP-induced cytotoxicity could be valuable therapeutics of patients with AD. Our previous study showed that 3,5-di-O-caffeoylquinic acid (3,5-di-CQA) and 3,4,5-tri-O-caffeoylquinic acid (3,4,5-tri-CQA) had a neuroprotective effect against amyloid-β_<1_42> (Aβ42)-induced apoptosis on cell death through the overexpression of glycolysis enzymes. We also found 3,5-di-CQA induced the improvement of spatial learning and memory on senescence accelerated prone mice (SAMP8). In this study, we report the inhibitory effect of CQA on Aβ42 aggregation and its cellular toxicity. Using thioflavin T assay and transmission electron microscope (TEM), 4,5-di-CQA and 3,4,5-tri-CQA inhibited the aggregation of Aβ42 in a dose-dependent manner. Structure-activity relationship of CQA and its derivatives suggest that caffeoyl group and phenolic hydroxyl group are essential for the inhibitory activity especially in a number-dependent manner of the caffeoyl groups. CQA inhibited the transformation into (β-sheet in Aβ42 and the cytotoxicity of Aβ42 on human neuroblastoma SH-SY5Y. On the other hand, CQA did not accelerate the generation of Aβ42 trimer. These results suggest that CQA could induce the formation of non-toxic conformer of Aβ42 characterized with turn at positions 25 and 26 and reduce neurotoxicity on cells by CQA treatment. CQA might be a promising agent for the prevention of AD.
Aplyronine A (ApA, 1) is an antitumor macrolide from the sea hare Aplysia kurodai. A PEG-linked biotin derivative 4 of ApA was prepared and was shown to exhibit potent cytotoxicity and cause actin disassembly in tumor cells. With the use of this probe, actin-related proteins 2/3 (Arp2 and Arp3) as well as actin were specifically purified from the whole lysate of HeLa S3 cells. Arp2 and Arp3 are the key proteins of the Arp2/3 complex, which binds to the sides of an existing actin filament and initiates growth of a new actin filament to form branched-actin-filament networks. Thus, ApA (1) may inhibit the ability of the Arp2/3 complex to bind to and branch F-actin, which may enhance actin filament disassembly. To the best of our knowledge, this is the first example that actin-depolymerizing molecules can also bind to actin-related proteins. A fluorescent derivative 7 possessing a tetramethylrhodamine moiety was also prepared from 1. With the use of this probe, time-lapse analyses in the living HEp2 cells that expressed GFP-actin as well as HeLa S3 cells were performed. Further structural and functional analyses of ApA-binding proteins as well as their roles in the antitumor activities of 1 will be presented.
The plant hormone auxin is a master regulator in plant growth and development. Indole 3-acetic acid (IAA), predominant naturally occurring auxin regulates the cell division and elongation leading to auxin-mediated developmental process. Auxin is perceived in F-box protein, TIR1/AFB auxin receptor, component of the E3 ubiquitin-ligase complex SCF^<TIR1>, and enhance the degradation of Aux/IAA repressor in the ubiquitin-proteasome pathway. Thus, auxin-dependant turnover of Aux/IAA regulate the gene expression. In this study, we describe a series of novel auxin antagonists for the modulation of the TIR1-Aux/IAA interaction. Based on the structure of TIR1-auixn complex, we identified the potent auxin antagonist, a-(phenylethyl-2-oxo)-IAA as a lead compound for TIR1/AFB receptors by in silico virtual screening, and designed auxinole as most potent auxin antagonist specific to TIR1. Auxinole bind into TIR1 receptor to block the formation of TIR1-IAA-Aux/IAA complex, and inhibited auxin-responsive gene expression. Molecular docking study demonstrated that auxinole would strongly interact with Phe82 residue of TIR1 that is crucial residue for Aux/IAA recognition. Auxinole also competitively inhibited various auxin responses in plant, such as lateral root and root hair formation, hypocotyl elongation and tropic response to gravity. Our works not only substantiates the useful chemical tools for plant biology, but also demonstrate a new class of inhibitor for the protein-protein interaction induced by small molecule that is common mechanism in other plant hormones perception, such as jasmonate, gibberellin, and abscisic acid.
Ciguatera fish poisoning (CFP) is a form of food poisoning caused by the ingestion of a variety of reef fish, and it affects more than 50,000 people each year. The extremely low level of the causative neurotoxins, ciguatoxins, in fish has hampered the preparation of antibodies for detecting the toxins. In this paper, we have developed a thiol strategy to synthesize the keyhole limpet hemocyanin (KLH)-conjugate (11) of the ABCDE-ring fragment of Pacific ciguatoxins, CTX IB (1) and 54-deoxyCTX1B (4), and succeeded in production of a monoclonal antibody (3G8) against their left wings by immunizing mice with the hapten-KLH conjugate (11) as the synthetic antigen. The mAb 3G8 lacks the cross-reactivity with other related marine toxins. Sandwich enzyme-linked immunosorbent assay (ELISA) utilizing 3G8 and the previously prepared monoclonal antibody (8H4) enabled us to detect 1 specifically at lower than 0.5 ng/mL (nM).
HsPKS1 from Huperzia serrata is a type III polyketide synthase (PKS) with remarkable substrate tolerance and catalytic potential. Here we present the synthesis of unnatural novel polyketide-alkaloid hybrid molecules by exploiting the enzyme reaction using precursor-directed and structure-based approaches. HsPKS1 produced novel pyridoisoindole (or benzopyridoisoindole) with the 6.5.6-fused (or 184.108.40.206-fused) ring system by the condensation of 2-carbamoylbenzoyl-CoA (or 3-carbamoyl-2-naphthoyl-CoA), a synthetic nitrogen-containing non-physiological starter substrate, with two molecules of malonyl-CoA. The structure-based S348G mutant not only extended the product chain length, but also altered the cyclization mechanism to produce a biologically active, ring-expanded 6.7.6-fused dibenzoazepine, by the condensation of 2-carbamoylbenzoyl-CoA with three malonyl-CoAs. Thus, the basic nitrogen atom and the structure-based mutagenesis enabled additional C-C and C-N bond formation to generate the novel polyketide-alkaloid scaffold. Benzalacetone synthase (BAS) from Rheum palmatum is a structurally simple, plant-specific type III PKS, which catalyzes the one-step decarboxylative condensation of malonyl-CoA with 4-coumaroyl-CoA. The type III PKS exhibits unusually broad substrate specificity and notable catalytic versatility. Here we report that R. palmatum BAS efficiently produces a series of unnatural, novel tetramic acid derivatives by the condensation of malonyl-CoA with aminoacyl-CoA thioesters, chemically synthesized from L- and D-amino acids. Remarkably, the novel tetramic acid dimer from D-phenylalanoyl-CoA, showed moderate antiproliferative activity against murine leukemia P388 cells.
In this study, mono- and pentacyclic C_<35> terpenes from Bacillus subtilis were biosynthesized via the cyclization of C_<35> isoprenoid by using purified enzymes, including the first identified new terpene cyclase, tetrapreny1-13-curcumene synthase, that shows no sequence homology to any of the known terpene cyclases. Based on these findings, we propose that these C_<35> terpenes are called the new family of "sesquarterpenes." This study demonstrated that a tetrapreny1-β-curcumene cyclase (TC) from B. subtilis, which was originally identified as a sesquarterpene cyclase converting a head-to-tail type of monocycle to a pentacycle, also cyclized a tail-to-tail type of linear squalene into a bicyclic triterpenol, 8α-Hydroxypolypoda-13,17,21-triene. The 8α-Hydroxypolypoda-13,17,21-triene was found to be a natural triterpene from B. megaterium, suggesting that the TC is bifunctional, cyclizing both tetrapreny1-α-curcumene and squalene in vivo. This is the first report describing the bifunctional terpene cyclase, which biosynthesizes 2 classes of cyclic terpenes with different numbers of carbons as natural products in the organism. Non-pathogenic Mycobacterium species also produce cyclic sesquarterpenes, which are biosynthesized via cyclization of Z-type C_<35> polyprenyl diphosphate. To provide deeper insight into the biosynthesis of sesquarterpenes, we carried out functional analyses of three Z-prenyltransferase homologues in M vanbaalenii identified by genomic analysis. Mvan_3822, a novel bi-functional Z-prenyltransferase, biosynthesizes C_<35>-heptaprenyl diphosphate as a main product from E,E-FPP and E,E,E-GGPP, but produces a C_<50>-decaprenyl diphosphate from GPP. Mvan_1705 is a novel Z,E,E-GGPP synthase. In addition, novel cyclic C_<35>-terpenes, 14E- and 14Z-dehydroheptaprenylcycline, were identified as minor metabolites in non-pathogenic Mycobacterium cells. Sesquarterpenes could be biosynthesized by two routes, in which E- and Z-geometric isomers of heptaprenyl diphosphate are produced from E,E-FPP and E,E,E-GGPP, and the prenylreductase responsible for the biosynthesis of sesquarterpenes may work to reduce both E- and Z-prenyl residues. The studies on sesquarterpenes promise to be an attractive field for expanding our understanding of the terpene world.
Vicenistatin is a 20-membered macrolactam antibiotic isolated from Streptomyces halstedii HC34. Its hybrid structure consisting of a unique amino acid (3-amino-2-methylpropionate, AmP) and polyketide is characteristic and thus prompted us to investigate the biosynthetic pathway, especially concerning the construction of the unique amino acid and its combination with the polyketide pathway. Our previous incorporation study indicated that the AmP moiety of vicenistatin is derived from L-glutamate (Glu) via 3-methylaspartate (MeAsp). Also, since the free form of AmP was not incorporated into the moiety, MeAsp seemed to be modified as CoA or ACP thioester before decarboxylation. Further, bioinfomatic analysis of the vicenistatin biosynthetic (yin) gene cluster revealed that eight unique proteins (VinH, I, J, K, L, M, N and 0) seemed to be responsible for the unique amino acid biosynthesis accompanying four PKSs and seven deoxysugar biosynthetic enzymes. In this report, we characterized the eight unique proteins, which were all involved in the AmP biosynthesis and the connection of the generated AmP with the polyketide pathway expectedly. At first, inactivation of the vinl gene for glutamate mutase E subunit resulted in the loss of vicenistain production, while the disruptant recovered it when MeAsp was added into the culture. This result suggested that glutamate mutase S subuint VinH and E subuint VinI generate MeAsp from Glu cooperatively. To investigate the other six proteins (VinJ, K, L, M, N and 0), we prepared the recombinant enzymes, which were successfully expressed in Escherichia coli. Two ATP-dependent ligases VinM and VinN were predicted to act as freestanding adenylation domains that catalyze two half reactions: The first is ATP-dependent activation of carboxyl group of amino acid and the second half-reaction is the transfer of the aminoacyl moiety of the adenylate intermediate to ACP (VinL) or its equivalent. Based on typical photometric assay by detection of released pyrophosphate, VinM was found to recognize small size of amino acids such as alanine, glycine and serine. On the other hands, VinN specifically recognized MeAsp. Futher, when the holo form of VinL was added into the VinN reaction mixture with ATP and MeAsp, the expected MeAsp-VinL formation was clearly detected by LC-ESI-MS analysis. Subsequently, the formed MeAsp-VinL was decarboxylated to be AmP-VinL by a PLP-dependent decarboxylase VinO. The generated AmP-VinL was then reacted with VinM, ATP and alanine to examine the additional attachment of aminoacyl group onto AmP-VinL to prevent a presumable thermodynamically favorable six-membered lactam formation during the chain elongation by PKS. As a result, a dipeptidyl ACP, Ala-AmP-VinL, was efficiently formed by VinM. The Ala-AmP-VinL was then used as a substrate for an acyltransferase VinK, which was speculated to catalyze the acyl transfer reaction between VinL and VinPl-ACP domain at the loading module of PKS. As we expected, the dipeptide transfer reaction occurred efficiently to give Ala-AmP-VinP 1 -ACP, which should be a starter molecule of vicenilactam PKS. Finally, to investigate the function of VinJ, a peptidase, we chemically synthesized the ethyl ester of N-alanyl-secovicenilactam as a substrate analog for the elongated polyketide intermediate. As we expected, VinJ catalyzed the hydrolysis of the analog to afford secovicenilactam ethyl ester, which was then macrocyclized by VinP4-TE to give vicnilactam to complete the biosynthesis. In this study, we clarified the unique starter biosynthesis of vicenistatin. It should be noteworthy that these amino acid carrying enzymes are highly conserved in the other macrolactam biosynthetic gene cluster, indicating that the clarified dipeptide strategy to carry amino acids appears to be a common way in the macrolactam biosynthesis. Thus, our finding shed a light on the rational design of engineered
Cytochrome P450s are ubiquitous monooxygenase enzymes that are responsible for the drug metabolism and biosynthesis of bioactive organic small molecules. For many bacterial species, P450 monooxygenases play substantial roles in the biosynthesis of useful secondary metabolites such as antibiotics. To date, thousands of P450 genes have been found in the genomes of most organisms, but substrates of many of their product enzymes have not yet been identified and thus need to be clarified. Moreover, characterizing the substrate specificity for natural and engineered P450 enzymes of interest is thought to be important for the production of useful organic small molecules and expansion of the library. Therefore, technology identifying substrates and/or characterizing the substrate specificity for P450 enzymes of interest would be very important not only for P450 chemistry and biology but also for the pharmaceutical discovery and production process. Toward this end, we have developed a droplet array format on an NADH-TEG-SC-slide, and an NAD^+ detection system using biotinylated acetophenone derivative 3. By using this platform, the substrate of P450cam was successfully identified on the array format.
Both fusicoccin A (FC) and structurally related cotylenin A (CN) are diterpene glucosides and show a phytohormone-like activity. However, only CN induces the differentiation of human myeloid leukemia cells. Since the CN producer lost its ability to proliferate during preservation, a study on the relationship between structure and activity was carried out and an elimination of hydroxyl group at 12-position of FC was essential to have the CN-like activity. Moreover, modified FC with hydroxyl group at 3-position was recently shown to be more effective. Therefore, we tried to identify a gene catalyzing 12-hydroxylation and to breed a mutant producing a modified FC without hydroxyl group at 12-position by a disruption of the gene. Previous identification of fusicocca-2,10(14)-diene synthase gene in Phomopsis amygdali, a FC producer, enabled us to identify a partial gene cluster for bisoynthesis of FC. However, other biosynthetic genes still remained unknown. In this study, we identified another gene cluster containing nine genes by draft genome sequencing. Of these, two cytochrome P450s catalyzing 813 and 9-hydrozylations, glycosylation, methylation, prenylation, and acetylation genes were confirmed to encode enzymes with the expected activities. We also identified a cytochrome P450 catalyzing 12-hydroxylations by a gene disruption experiment. The remaining one cytochrome P450 gene therefore probably catalyzes hydroxylation at the 19-position.
(-)-Gelsemoxonine, a relatively new member of gelsemium alkaloids, was isolated from the leaves of Gelsemium elegans in 1991. While the structure of gelsemoxonine was misassigned originally, it was revised in 2003 by Aimi on the basis of an X-ray crystallographic analysis. The correct structure features the unique azetidine moiety bearing a tetra-substituted carbon center, which is quite unique among indole alkaloids. Combined with the quaternary center of the spiro-N-methoxy indolinone and the intricate cyclic system, gelsemoxonine posed a formidable challenge to synthetic chemistry. The first total synthesis of gelsemoxonine has been accomplished by means of 1) divinylcyclopropane-cycloheptadiene rearrangement of the highly functionalized substrate, which was successfully applied to assemble the spiro-quaternary carbon center connected to the bicyclic seven-membered core structure, as well as 2) a one-pot isomerization reaction from the α,β-unsaturated aldehyde to the saturated ester via the TMSCN-DBU combination that allowed a facile diastereoselective introduction of the latent nitrogen functionality of the unique azetidine moiety.
Saxitoxin [STX], which was first isolated as a paralytic shellfish poison, is a potent and specific blocker of voltage-gated Na+ channels [VGSC]. This unique biological activity prompted us to use STX as biochemical tools for the study of VGSC and other ion channels in the field of neurophysiology. In order to make the specific inhibitors of VGSC, an efficient synthetic route for STX skeleton has been studied in our laboratory. Here, we describe a total synthesis of decarbamoyl-a-saxitoxinol [dcSTXol (1)], a non-toxic, naturally occurring analog of STX produced by the cyanobacterium Lyngyba wollei. Our synthetic strategy for 1 is shown in Scheme 5. We envisioned that tricycle G, which contains AC ring unit of 1 could be synthesized from simple homopropargyl guanidine D by double bromocyclization and intramolecular N-alkylation (D→E→F→G). After introducing hydroxy group and second guanidine (G→H→I), opening of the hemiaminal ether and cyclic guanidine formation under an acidic condition would furnish 1. A precursor 21 for the bromocyclization reaction was synthesized from aziridine 11 in five steps, as shown in Scheme 6. Compound 11 underwent ring opening with NaN3, and the TBS ether of the product was transformed to the corresponding mesylate 19 in 2 steps. Sequential deprotection of the Ac and Boc groups yielded 21, which served as the precursor for bromocyclization. As anticipated, upon treatment of 21 with PyHBr3 in biphasic medium of CH_2Cl_2 and aqueous K_2CO_3, the key bromocyclization reaction and subsequent intramolecular N-alkylation took place to afford 24 as a single product in 36% overall yield in 6 steps from 11. After obtaining key intermediate 24, we next examined transformation of the gem-dibromomethylene group of 24 to carbonyl functionality as shown in Scheme 7. Compound 24 was successfully transformed into alcohol 26 in two steps: (i) Ac_2O, Et_3N in CH_2Cl_2 and (ii) reduction of the resulting enol acetate with NaBH_4. The azide group of 26 was then reduced under Staudinger conditions followed by conventional guanidinylation to give 27. Deprotection of the Cbz groups under hydrogenolytic conditions yielded 28. Upon treatment with B(OCOCF_3)_3 in TFA, 28 underwent ring opening of the hemiaminal ether and subsequent cyclization of the second guanidine to furnish 1 in 73% yield. In summary, we have achieved the concise total syntheis of dcSTXol (1), featuring key bromocyclization for tricyclic intermediate 24 from simple homopropargyl guanidine 21. In addition, novel transformation of the gem-dibromomethylene group to enol acetate has been developed. Further studies toward the synthesis of other STX analogs employing these methodologies are currently under investigation in our laboratory.
Chirality transfer through [3,3]-sigmatropic rearrangements is one of the most reliable approaches for accessing enantiopure multifunctional molecules. To maximize the potential utility of chirality transfer, we developed two types of novel sequential sigmatropic rearrangements of allylic diols: 1) the sequential Claisen/Claisen rearrangement (Scheme 2), and 2) the sequential Claisen/Overman rearrangement (Table 1, Scheme 3). While the former rearrangement gave two identical functional groups without protecting group manipulation of homoallylic diols, the latter rearrangement installed two different functional groups. Especially, the latter Claisen/Overman rearrangement was more challenging because it required the suppression of the second competitive Claisen rearrangement after the first rearrangement (12→9). To overcome this inherent problem, the Claisen rearrangement via cyclic orthoamides was employed, furnishing the singly rearranged products 12, which underwent Overman rearrangement stereoselectively. To demonstrate the practical utility of the sequential sigmatropic rearrangements, we applied this methodology in the total synthesis of (-)-kainic acid (1, Scheme 5). A conspicuous feature of our approach is the use of chirality transfer in acyclic compounds made possible by taking advantage of the three secondary hydroxy groups embedded in D-arabinose (15). An S_N2' reaction (21→22) and the sequential Claisen/Overman rearrangements (23→24) could establish all of the chiral setereocenters (C2, C3, C4) in kainic acid (1). The resulting product 24 was successfully transformed to (-)-kainic acid (1).
(+)-Ophiobolin A (1) was first isolated as a metabolite from the culture broth of the pathogenic plant fungus Ophiobolus miyabeanus, and from its absolute structure proof, the first naturally occurring sesterterpene was identified. Recent studies have revealed that 1 is cytotoxic to various cancer cell lines. 1 possesses eight stereogenic centers on the unique ring system including a spirocyclic ether. The A-ring incorporates three successive stereogenic centers including a chiral tertiary alcohol and two contiguous stereogenic centers at the cis-fused AB-ring junction. The CD-ring is a spirocyclic ether, possessing a total of five stereogenic centers, four of which are successive, as well as a quaternary stereogenic center at the trans-fused BC-ring junction. The potent bioactivities as well as the complex structure make 1 an attractive synthetic target. The total synthesis of 1 commenced with the construction of the spirocyclic CD-ring moiety. The spirocyclic CD-ring moiety was successfully constructed via the Lewis acid-promoted intramolecular Hosomi-Sakurai reaction. The CD-ring moiety was efficiently assembled with the A-ring fragment by the reaction reported by Utimoto et al. To the best of our knowledge, this is the first application of this coupling reaction to natural product synthesis. The first construction of the ophiobolin carbocyclic skeleton has been achieved by RCM, but it was sensitive to the substrate structure and use of the substrate with a less bulky protective group, such as benzyl ether, was crucial. Finally, the total synthesis of (+)-ophiobolin A has been achieved for the first time, via a convergent approach.
Recently, many efforts have been made toward the efficiency in the natural product synthesis, and the more efficient stereocontrol as well as the shorter number of steps is required for that purpose. Herein, we report the detail of efficient total syntheses of three natural products using original methods for controlling the stereochemistries of ether rings. <Asymmetric synthesis of sniroacetal pheromone based on the substituent-indenendent stereocontrol at the spirocenter> Spiroacetal is one of the basic substructure in the bioactive natural products. Spiroacetalic natural products exhibit extensive biological function such as pheromonal, antibacterial and anticancer activities and so on. In the syntheses of these spiroacetals, their spirocenters are generally controlled by the stereochemical and/or electrochemical effect(s) of the substituent(s). However, stereocontrol of the spirocenter is sometimes difficult when effect(s) is(are) insufficient or when competitive effects exist. We developed a novel constructive method for the completely stereoselective synthesis of spiroacetals by using a stereochemical restriction of 4-membered-ring sulfide, and which provide the new synthetic approach to spiroacetalic compounds. Asymmetric dialkylation of SAMP-hydrazone (S)-4 prepared from 3-thietanone (3) gave compound 5. Optically pure (+)-olean was synthesized from 5 via acid treatment followed by desulfurization. Even the very unstable and non-anomeric compound 16 could also be synthesized by the same method from RANP-hydrazone (R)-4.<Asymmetric total synthesis of (+)-anthecularin> Anthecularin (19) was isolated from a lipophilic extract of the aerial parts of Greek Anthemis auriculata by Karioti and co-workers in 2006. This compound exhibits antitrypanosomal and antiplasmodial activity and inhibited two key enzymes of the plasmodial type II fatty acid biosynthesis pathway, PfFabI and PfFabG. Anthecularin has a characteristic structure with fused oxabicyclo[3.2.1]octene, cyclohexene and butyrolactone rings. Although four contiguous asymmetric centers exist in the molecule, only the relative stereochemistry has been clarified. Its unique structure prompted us to embark on the total synthesis. We achieved the total synthesis of anthecularin via the two ring closing metatheses as key reactions. The stereocontrolled addition of vinyllithium to 37a + 38a was achieved by the chelation effect of methoxyethoxy group. The overall yield was 7.9% over 18 steps, and the absolute configuration of anthecularin was determined to be 2S,3S,4R,8S by the comparison of the specific rotations between the natural and our synthetic sample.
Ellagitannins are a kind of natural hydrolyzable tannins. Despite of the wide-ranging structural diversity, the structure can basically be expressed as D-glucose possessing hexahydroxydiphenoyl (HHDP) bridge. Ellagitannins have been considered to be biosynthesized via β-pentagalloylglucose (PGG) by oxidative coupling of gallates. However, detailed mechanism of the coupling reaction and the regioselectivity in the HHDP-bridge formation has not been elucidated. We have developed a new method for synthesizing HHDP group via intramolecular coupling of 4-O-benzylgallates induced by CuC1_2-n-BuNH_2, and have applied this reaction to several total syntheses of ellagitannins. We here describe a chemical simulation of the oxidative coupling of β-PGG. The aim of this study was to examine the coupling-prone positions. As an analog of β-PGG, penta-0-(4-0-benzylgalloy1)-β-D-glucopyranose was designed for the chemical simulation. The 4-O-benzylgallates on the substrate were coupled by CuCl_2-BuNH_2. As a result, the couplings were mainly induced in two positions; one was of the 1- and 6-O-gallates, and the other was of the 4- and 6-O-gallates. The coupling between the 1- and 6-positions was the first chemical formation of the HHDP-bridge accompanying inversion of glucopyranose ring into the axial-rich conformation. Additionally, application of the results allowed total syntheses of davidiin and tellimagrandin II in four steps from D-glucose.
Penostatins A - I were isolated from Penicillium sp. OUPS-79 separated from a marine alga by Numata. They exhibited significant cytotoxicity against P388 cells. Because of the intriguing structural features, biological profiles and less availability, they represent attracting targets for total synthesis. Although a report on the synthesis of 5-deoxy-penostatin A has only been published, total synthesis has never been reported so far. Here, we report the first total synthesis of penostatin B (2). The key features of the synthesis are 1) Lewis acid-mediated highly diastereoselective introduction of an alkenyl side chain to the dihydropyran ring, 2) successful application of a relay ring-closing metathesis (RRCM) to the construction of the dihydropyranone ring, and 3) the use of a group and face selective Pauson-Khand reaction for the diastereoselective construction of the bicyclic skeleton.
Recently, natural biologically-active a-acyl-y-hydoroxylactams with decalin skeletons and their alkoxyaminal-type analogues have been reported. The series of compounds attract much attention for their various and interesting biological activity. Each of these compounds shows the different biological activity owing to the structural variation in the decalin skeleton and the substituents on the γ-position of lactam rings. However, their activity spectra and mode of actions have not yet been declared. Therefore, we started synthetic study for this type of compounds in order to elucidate the above mentioned undefined points. We would like to report here the first total synthesis and structure determination of H1V-integrase inhibitor Oteromycin (1) based on a novel synthetic strategy for the α-acyl-γ-hydroxylactams. In the course of this study, intramolecular Diels-Alder (IMDA) reaction of 1 was found to be applicable for the synthesis of CYP3A4 inhibitor Diaporthichalasin (2). Therefore, we also would like to describe the first total synthesis of 2 as an extension of our developed synthetic strategy. First, total synthesis of 1 was investigated. Chiral vinyl iodide and trienyl borane were respectively prepared from (+)-citronellal and trans-2-methyl-2-butenal. These compounds were then coupled by Suzuki-Miyaura reaction, and the following two step reactions gave the corresponding pentaenal-type cyclization precursor. IMDA reaction of this precursor proceeded stereoselectively to afford the desired decalin fragment. On the other hand, N-Teoc-protected y-benzyl-y-methoxylactam was also prepared as another important fragment. Aldol reaction between these fragments proceeded smoothly, and the following several transformations gave the desired precursor of 1. However, hydrolysis of the methoxyaminal moiety gave the undesired pentacyclic compound because of the exclusively occured IMDA reaction between diene-type side chain on the decalin skeleton and the α,β-unsaturated lactam moiety. Therefore, we tried to develop the second synthetic strategy which is not through the hydrolysis step. An interesting model synthesis of α-acyl-γ-hydroxylactams utilizing the epoxidation of the enol form of α-acyl-α,β-unsaturated lactam was reported by Snider's group in 2004. However, this method has not been applied to the total synthesis of natural products. It is probably due to the unsatisfactory stability of the intermediate α-acyl-α,β-unsaturated lactam under the basic condition. Therefore, we developed novel synthetic pathway via the in situ preparation of α-acyl-α,β-unsaturated lactam starting from the corresponding saturated-type one. That is, saturated-type α-acyllactam was synthesized and subjected to the phenylselenenylation to give the corresponding a-selenolactam. The obtained a-selenolactam was then treated with two equivalents of MMPP (magnesium monoperoxyphthalate hexahydrate). In this reaction condition, oxidative elimination of phenylselenenyl group, enolization of the resulting α,β-unsaturated lactam, epoxidation at β-γ position and ring opening to the desired γ-hydoroxylactam subsequently proceeded in one pot. Thus, the first total synthesis of 1 was successfully achieved. The absolute configuration of 1 was assigned by comparison of the optical rotation with reported data. However, the configuration of C24-position could not be determined from all of the spectral data. Therefore, IMDA reaction between diene side chain on the decalin skeleton and the a,13-unsaturated lactam moiety was carried out to observe the cross-peaks in the NOE spectrum of the resulting pentacyclic compound. As a result, the absolute configuration of C24-position was defined as 5, and the chemical structure of 1 was entirely determined. Surprisingly, the carbon skeleton and the relative configulation of the obtained pentacyclic compound were the same as that reported as 2. Therefore, we also studied the
One-pot processes, the sequential conjugation of plural bond forming reactions, are powerful tools in organic synthesis. Conjugate addition of lithium amide 4 to enoate 5 forms an N-C bond and generates lithium enolate 6, which is further applicable as a carbonucleophile2b) reacting in an SN2 manner with electrophile 8 (Scheme 1). We have recently succeeded in the application of this tandem conjugate lithioamination-alkylation to the asymmetric total synthesis of (-)-aspidospermidine. When an electrophile has two reaction sites like dihalide 12, the initially introduced nitrogen atom in 13 undergoes an intramolecular S_N2 reaction to form piperidine 11 (Scheme 2). We describe herein an asymmetric one-pot piperidine formation using our asymmetric conjugate lithioamination of a lithium amide with an enoate as the central machinery, and the total synthesis of (-)-kopsinine 11) (Fig. 1). An asymmetric conjugate addition of 4 was performed with 14 using chiral diether ent-3 in toluene at -78℃ for 1.5h and then at -40℃ for 2h to afford, after quenching with HC1 in Me0H, 3-aminoester 16 with 95% ee in 75% yield (Scheme 3). Then, alkylation of enolate 13 as the second step was examined. After lithioamination, alkylation was performed by the addition of 12a in THF/HMPA to give, after treatment with tetrabutylammonium fluoride to remove the TBS group, 17 as a sole diastereomer in 70% yield with disappointingly poorer 86% ee (entry 1). DMPU was found to be a good additive, and 17 was obtained with excellent enantio- and diastereoselectivity (entry 2). Finally, conducting the conjugate addition step at -65℃ for 15 h, 17 was produced in 89% yield with 98% ee (entry 3). A one-pot piperidine formation was realized by sequential asymmetric conjugate lithioamination, stereoselective alkylation with 1-chloro-3-iodopropane 12b in the presence of DMPU, giving cis-11 with 98% ee in 85% yield. N-Hydroxyethylpiperidine methyl ester 19 was obtained in 85% three-step yield from cis-10 by hydrogenolysis, N-hydroxyethylation, and simultaneous de-Boc and methyl esterification by heating in refluxing methanol with sulfuric acid (Scheme 4). Claisen condensation of 19 with lithium enolate 21b gave ketoester 23 in 72% yield (Scheme 6). Mesylation of 23 and subsequent treatment with potassium tert-butoxide gave pentacyclic 10, a known synthetic intermediate of (-)-kopsinine (1), in 61% two-step yield. (-)-kopsinine (1) was synthesized by reported sequences from 15, whose absolute configuration was then undoubtedly confirmed (Scheme 7).
Kaitocephalin (1), isolated from Eupenicillium shearii PF1191 by Seto and Shin-ya et al., is known to exhibit a potent inhibitory activity against neuronal cell death by the antagonistic action on AMPA as well as NMDA glutamate receptors. This natural product has, therefore, considerable potential as a promising lead compound for developing therapeutic agents against neuronal diseases such as stroke and epilepsy. However, its detailed neurobiological study becomes very difficult at present because the fungus has not produced kitocephalin. Due to such extremely low availability from natural sources as well as the intriguing biological activity and synthetic challenges, kaitocephalin (1) has attracted much attention in the chemical and biological communities. Thus, there have been a number of synthetic studies including total syntheses achieved by five groups. In this symposium, we will report a new approach to 1, which relies on a C-H amination reaction of a sulfamate followed by formation of a pyrrolidine skeleton by nucleophilic opening of the resulting cyclic sulfamate. Thus, Rh(II)-catalyzed C-H amination reaction of 2lusing Rh_2(esp)_2 gave 22, which, upon tert-butoxycarbonylation and base treatment, produced pyrrolidine 24. 24 was converted to advanced key intermediate 30 via another Rh(II)-catalyzed C-H amination reaction for the introduction of a nitrogen atom at the allylic position of the cyclopenten ring of 26 giving 27 and installation of the 3-5-dichloro-4-acetoxybenzoyl group. From 30, the key precursor 31 for the synthesis of 1 was successfully synthesized by ozonolysis followed by Ru04-catalyzed oxidative cleavage.
In 2002, mannopeptimycins 1-5 are isolated from Streptomyces hygroscopicus LL-AC98 as potent antibiotics by He and his co-workers. Mannnopeptimycins exhibited antimicrobial activity against Gram-positive bacteria, including methicillin-resistant streptococci and vancomycin-resistant enterococci. Mannopeptimycins consist of 6 amino acids including 3 unnatural residues. Aiha-A (α-amino-β[4'-(2' -imino-imida -zolidiny1)]-(β-hydroxypropionic acid-A) and Aiha-B are unnatural cyclic guanidine containing amino acids which are only found in mannopeptimycins. For further elucidation of mechanism of action and structure-activity relationships of mannnopeptimycins, the development of efficient synthetic methodology is important. However, there has been no synthetic report for mannopeptimycin aglycon. Only recently, syntheses of Aiha-A and Aiha-B were disclosed by two groups. Herein, we wish to report the total synthesis of proposed structure of mannopeptimycin aglycon 6. Aiha-A and Aiha-B were synthesized from the common synthetic intermediate 19 via asymmetric aldol reaction with aldehyde 17. Threo product 21 smoothly cyclized to afford the corresponding N,0-cyclic acetal 22 in the presence of silica-gel. On the other hand, erythro product 20 did not form the N,0-cyclic acetal due to steric repulsion. 20 and 22 were readily separated by using silica gel column chromatography. A cyclic guanidine moiety was introduced using Goodman reagent to afford Aiha A 26 and Aiha B 28, respectively. L-Ser 8 was coupled with Aiha-B 28 using HATU. The following condensation with Aiha-A 26 using HATU afforded the undesired product 31. We could overcome the problem by employing PyBOP as a condensation agent. The coupling between 25 and 30, and the crucial macrolactamization proceeded well using HATU. Global deprotection under hydrogenation condition afforded the aimed product 6 in good yield. The structure was confirmed by ^1H NMR, ^<13>C NMR, ^1H-^1H COSY, HMBC, HSQC, MS analyses.
Polytheonamide B was isolated from a marine sponge Theonella swinhoei as the largest non-ribosomal peptide, and was found to have a potent cytotoxicity caused by formation of 06.3-helix structure and a monovalent cation channel in the lipid bilayers (Figure 1). The length of 136.3-helical ion channel formed by the single molecule of polytheonamide is long enough to span the lipid bilayer. These unique features are intriguing as the structural platform for designing novel ion channel molecules. Here, we report the total synthesis and functional study of the artificial ion channel peptide, dansylated polytheonamide mimic (1, Figure 2). The original sequence of polytheonamide B contains various non-proteinogenic amino acids, which require the multistep syntheses. To preserve the ability of forming stable 06.3-helical structure and reduce the synthetic steps in the total synthesis of 1, these amino acids were substituted to commercially available or synthetically more accessible ones, which have the similar amide and hydroxyl groups. At the 44th residue, the alkyne-type side chain was introduced as a handle for functionalization via click chemistry. Non-proteinogenic amino acids were synthesized prior to peptide synthesis (Figure 3). These amino acids were used in the automated solid-phase peptide syntheses to give the two peptide segments (Scheme 1). Subsequently, thioesterification of the N-terminal segment and introduction of the dansyl moiety to the C-terminal segment led to segments 15 and 18, respectively. Ag^+ mediated coupling reaction between 15 and 18 gave the protected peptide 2. Then, the 15 protective groups (Tmb, Tr, t-Bu) were simultaneously removed under the acidic conditions to generate 1. To evaluate functions of the novel artificial peptide, we performed the biological studies. The cytotoxicity assay against p388 mouse leuckemia cells revealed that 1 had a strong cytotoxicity (IC50 = 12 nM). Furthermore, the membrane disruption and ion transport assays using fluorescent probe-encapsulated liposomes indicated that 1 transported alkaline metal cations across the membrane without the membrane disruption (Figure 4). In conclusion, design and total synthesis of 1 was successfully achieved. Synthesized 1 has a strong cytotoxicity and cation channel activity. The simplified structure and automated synthetic route are beneficial for practical syntheses of various derivatives. Our rationally designed peptide structure will be utilized as a novel platform of a channel structure.
Indole alkaloids composed of elaborated cyclic arrays have been one of the most valuable resources for discovering biologically active substances. To gain concise and flexible access to alkaloid analogs possessing natural and unexploited skeletons, we are developing a synthetic process based on divergent cyclizations inspired by the biogenic strategy generating structural diversity of natural products. With intentions to emulate the biosynthetic proposal of iboga and aspidosperma alkaloids, we designed a common dihydropyridine precursor 3 for divergent Diels-Alder reactions in either of two ways: (1) the dihydropyridine reacts as a diene leading to the iboga-type skeleton; (2) the dihydropyridine serves as a dienophile to form the aspidosperma-type skeleton. First of all, we developed an original protocol for constructing sensitive 1,6-dihydropyridine (DHP) rings employing a cationic Cu(I) catalyst. Next, the key precursor 21 for the Cu(I)-catalyzed DHP synthesis was successfully prepared through conjugate addition of the common intermediate 25 with methyl propiolate 20 and concomitant Hofmann elimination. The Cu(I)-catalyzed 1,6-DHP cyclization and subsequent Diels-Alder cycloaddition under mild conditions (45℃) allowed a cascade synthesis of iboga-type skeleton 28. Meanwhile, divergent cyclization proceeded to furnish ngouniensine-type skeleton 29 by the simple modification of reaction temperature (100℃). Furthermore, the reaction site of 25 with 20 was altered depending on the solvents to produce a distinct precursor for the Cu(I)-catalyzed cyclization, which allowed successful construction of a novel tetracyclic skeleton 31. We also achieved total synthesis of (±)-catharanthine 34 through protecting group-free 9-step sequence in order to demonstrate applicability and potential of the currently developed synthetic process efficiently generating complexity and diversity of natural product relevant scaffolds in a programmable fashion.
Marinostatin (MST) isolated from a marine organism is a serine protease inhibitor consisting of 12 amino acids with two internal ester linkages formed between the β-hydroxyl and (β-carboxyl groups, Thr^3-Asp^9 and Ser^8-Asp^<11>. MST was synthesized by regioselective intramolecular esterification employing two sets of orthogonally removable side chain protecting groups for Ser/Thr and Asp. SAR study revealed that the ester linkage with Thr^3-Asp^9, the cis-conformation at Pro^7 and the N-terminal Phe^1-Ala^2 are the structural requirements for expression of the inhibitory activity. These findings were also supported by analyzing the solution and enzyme-bound structures of MST. Of particular note is that cis-Pro^7 may promote the internal hydrogen bond between the NH proton of Are and the carbonyl oxygen atom of the ester linkage with Thr^3-Asp^9 to protect its scissile bond of Met^4-Arg^5. This could be responsible for enhancing the potency. To elucidate the importance of backbone conformation at position 7, 16 and cis/trans-olefin analogs 17/18, in which cis/trans-olefins are substituted for the amide bond of Tyr^6-Ala^7, were synthesized. Although Ala^7 in 16 takes a trans-conformation in the solution structure, it takes a cis-conformation in the enzyme-bound structure. This implies that Ala^7 would isomerize from a trans to cis conformation when it binds to an enzyme, resulting in a certain inhibitory potency. However, the trans-olefin analog 18 lost the potency while the cis-olefin analog 17 displayed almost the same potency as that of MST. These results clearly indicated that the cis-conformation at position 7 is indispensable for binding to an enzyme in a canonical manner. By applying the structural motif of MST, we were able to rationally design protease inhibitory specificities that differed from those of the natural product.
Recently, kainoids, such as kainic acid, have received significant attention due to their potent binding affinity for ionotropic glutamate receptors (iGluRs). iGluRs are involved in important neurophysiological processes, such as memory and learning. Although many synthetic investigations of kainoids have been reported to date, efficient synthetic methods are still strongly required. During the pioneering investigations on acromelic acid A, isolated by the Shirahama group, it was discovered that a synthetic derivative, methoxyphenyl kainic acid, possessed more potent activity than the natural compound. Inspired by this interesting structure-activity relationship, we launched an investigation into the development of efficient synthetic methods for achieving MFPA and phenylkainic acid. In our synthetic strategy, we envisioned that the three consecutive chiral centers were constructed based on the stereochemistry of the C4 position. We employed our asymmetric intermolecular C-H insertion reaction assisted by the chiral auxiliary using diazo ester and cyclohexadiene to afford the desired diene ester in high yield with good diastereoselectivity. After the conversion into the lactone by successive ozonolysis of cyclohexadiene, nitrogen was installed with Ns amide to give the corresponding hemiaminal. The reduction of the aminal and deprotection of the acetal induced the cyclization to construct a pyrrolidine ring with the correct stereochemistry. Introduction of two cyano groups were performed by diastereoselective Strecker-type reaction and Mitsunobu reaction. Finally, hydrolysis of the two cyano groups gave the synthetic kainoids. Furthermore, several investigations using the synthesized kainoids have elucidated that these compounds selectively bind to iGluRs and are equally effective for mice in vivo.
Pectenotoxin-2 (1), isolated from the Japanese scallop Patinopecten Yessoensis and the dinoflagellate Dinophysis fortii as a causative toxin of the Diarrheic Shellfish Poisoning, is a polyether macrolide characterized by a 34-membered macrolactone, a non-anomeric [6,5]-spirocyclic acetal, a bicyclic acetal, three oxolanes, and a 6-membered cyclic hemiacetal. It exhibits potent cytotoxicity against several cancer cell lines due to its actin depolymerizing activity. Because the structural complexity and strong bioactivity attracted our attention, we initiated the program toward its total synthesis. In this symposium, we will describe the details of the recently finished total synthesis of pectenotoxin-2 (1) and its congeners, pectenotoxin-2b (3) and -2c (2). The synthesis of pectenotoxin-2 was performed by a convergent route starting from the Cl-C7, C8-C20, C21-C30, and C31-C40 segments. The C1 -C7 and C31-C40 segments were connected by esterification to form the left-half segment of 1. The C8-C20 segment was united with the C21-C30 segments with forming the bicyclic acetal moiety by a process including the sulfone coupling between the segments, oxidation to a ketone, desulfonylation with SmI2, and acetalization. The connection of the resulting C8-C30 (the right-half) with the left half by sulfone coupling was followed by the formation of an anomeric [6,5]-spirocyclic acetal. After several steps of functional group interconversion, the ring-closing olefin metathesis at C30-C31 successfully constructed pectenotoxin-2b (3). Finally, anomeric [6,5]-spirocyclic acetal 3 was converted to non-anomeric [6,5]-spirocyclic acetal 1 by a short-period treatment with aq-TFA in acetonitrile and to double-anomeric [6,6]-spirocyclic acetal 2 by a long-period exposure to the same acidic conditions. Synthetic pectenotoxin-2 (1) exhibited significant inhibition of the viability of HepG2 and Caco2. Synthetic pectenotoxin-2b (3) and -2c (2) also inhibited the viability, but with weaker IC50 values.
C-1027 is a potent antitumor chromoprotein antibiotic. It is composed of an apoprotein and a labile 9-membered enediyne chromophore 1. The C-1027 chromophore 1 quickly aromatizes via Masamune-Bergman cyclization at ambient temperature, generating p-benzyne biradical 2 that exerts its biological activity by abstracting hydrogen from DNA. The chemical instability and the complex architecture distinguish 1 as a challenging target for the total synthesis. We report a total synthesis of the protected aglycon and the chromophore 1. A key feature of our syntheis is construction of the nine-membered enediyne from the nine-membered diyne via the reductive olefination of vicinal diol derivative. First, we examined the reductive olefination in the presence of a benzoxazine moiety and synthesis of the protected aglycon 15. As shown in Scheme 2, the adequately protected benzoxazine moiety 9 was condensed with readily prepared nine-membered diyne 4. The product 10 was converted to 13, which is the suitable precursor of the reductive olefination. The reduction of 13 with samarium iodide was successful to afford the protected aglycon 15 via dehydration of hemiacetal using a mesylation-elimination sequence. Then, we attempted the total synthesis of 1. Highly diastereoselective glycosylation under Schmidt conditions enabled preparation of 23 (Scheme 4). The reductive olefination of 23 and subsequent dehydration also worked well to give the protected chromophore 24. Finally, formation of 1 was confirmed by ESI-MS (HRMS) after global deprotection with TBAF.
Gangliosides are complex glycosphingolipids that are ubiquitous components of mammalian cell membranes. GM3, which is a simple ganglioside, was known to inhibit cell proliferation and self-phosphorylation of EGFR stimulated by EGF. On the other hand, plasma membrane-associated human sialidase NEU3 hydrolyzes the glycosidic linkage of sialic acid in GM3 to produce lactosylceramide, and is up-regulated in various cancer cells. Recent reports suggested that enzymatic metabolism of GM3 by NEU3 may be involved in cancer malignancy. To clarify the relationship between GM3 metabolism and cancer malignancy by NEU3, we designed sialidase-resistant CF_2- and CH_2-linked GM3 analogues, in which the oxygen atom of sialoside linkage is replaced by the CF_2 or CH_2 group, as a ganglioside probe and a substrate-based NEU3 inhibitor. This time, we succeeded in the synthesis of both GM3 analogues. During the course of the synthesis, we developed the novel methodologies for the construction of the key C-sialoside linkages utilizing Ireland-Claisen rearrangement reaction. Both CF_2- and CH_2-sialoside linkages could be formed in highly stereo-selective manner. Moreover, we revealed the unique temperature effect in this rearrangement reaction, in which stereo-selectivity was increased when the reaction temperature was raised. Detail of synthesis and preliminary results of biological activity of GM3 analogues will be presented.
Regioselective functionalization of one of the multiple hydroxy groups is a fundamental challenge in current organic synthesis. Synthesis of complex glycoconjugates has usually been achieved by multi-step protection/deprotection procedures, which results in long synthetic steps and low atom economy. Recently, we have developed a perfectly regioselective acylation of D-glucose derivative 2 with organocatalyst 1 (Scheme 1). We report here total synthesis and functionalization of natural glycosides via organocatalytic regioselective acylation. Toward the total synthesis of natural glycosides, we investigated functional group tolerance in the regioselective acylation catalyzed by 1. Introduction of functionalized acyl groups such as a-amino acyl, cinnamoyl, and galloyl groups to monosaccharide 2 and disaccharides 11-13 has been achieved in high regioselectivity (rs). (Figure 2, 4-10 and Table 2) These results indicate that this protocol for regioselective acylation is applicable to the introduction of functionalized acyl groups to a variety of saccharides with a glucopyranose terminus. Regioselective functionalization of lanatoside C (14) was then investigated. Compound 14 is a naturally occurring cardiac glycoside which has been used for the treatment of atrial fibrillation and supraventricular tachycardia. (Figure 3) As expected, acylation of 14 took place at C(4''')-OH in high regioselectivity in the presence of catalyst 1. On the other hand, the acyl group was introduced to C(3''')-OH selectively in DMAP catalyzed acylation. (Scheme 2, A) Because the biological activities of some glycosides have been known to depend on the acyl groups by their position and functionality, the direct regioselective introduction of various acyl groups seems to be a unique and useful method for exploring bioactive materials. Regioselective total syntheses of multifidoside A (15) and B (16) have been achieved. Compounds 15 and 16, 4-0-p-coumaroylglycosides isolated from pteris multifida in China, showed cytotoxicity against the HepG2 tumor cell line with IC_<50> < 10μM. (Figure 4) We have accomplished the first total syntheses of 15 and 16 via two types of organocatalytic reactions. The first key reaction is an intramolecular asymmetric aldol reaction of 19, which was synthesized from commercially available 20 by nine steps. (Scheme 3) After thorough screening of known and unknown catalysts, we found that catalyst 24 has reactivity sufficient to give indanone 25 in high stereoselectivity. The second key reaction is organocatalytic regioselective acylation of 26 and 28. As expected, acylation took place at C(4)-OH in high regioselectivity (up to 86% rs). Finally deprotection of the products gave 15 and 16. (Scheme 4 and 5) Further optimization of regio- and enantioselectivity of the key reactions is currently underway.
We have been developing the method for the total chemical synthesis of glycoprotein. Recently, we found a new synthetic method of peptide-alpha-thioester that is a key component of the modern chemical protein synthesis. In addition to this, we have succeeded the structural analysis of a chemically synthesized homogeneous glycosyl-chemokine (CCL1) by quasi-racemate X-ray crystallography. In this presentation, we would like to discuss the details of these results.  We have pursued a new synthetic method of peptide-alpha-thioester by using an un-protected peptide as a starting material. We designed a three-step conversion sequence consists of S-thiocarbonylation of the C-terminal Cys residue (Figure 2 (i)), treatment of the afforded S-thiocarbonyl peptide with N-acetylguanidine (Figure 2 (ii)) and thiolysis reaction (Figure 2 (iii)). S-Thiocarbonyl group could induce an activation of alpha-nitrogen on the Cys residue in the presence of N-acetylguanidine. This provided a cleavage of peptide bond at AA-Cys sequence and afforded a peptide having N-acetylguanidine at C-terminal. We found this compound was easily converted to a corresponding peptide-alpha-thioester under neutral buffer solution. We also successfully obtained a desired glycopeptide-thioester by this new methodology (Figure 3).  There are well known difficulties in X-ray diffraction structural analysis of glycoproteins, which can be hard to crystallize because of the heterogeneity of the oligosaccharide moiety. Recently, racemic protein crystallization has reported. This method is facile and straightforward crystallization method by using racemic protein which is a 1 : 1 mixture of native protein consist of L-amino acids (L-protein) and mirror image protein consist of D-amino acids (D-protein). We envisioned quasi-racemic protein consists of a 1 : 1 mixture of native glycoprotein and D-protein would also provide facile crystallization. Then, we carried out convergent chemical syntheses of CCL1 derivatives and successfully obtained three distinct protein molecules: glycosyl-L-CCL1, L-CCL1 and D-CCL1 made from D-amino acids. Crystallization of the homogeneous L-glycoprotein was still difficult. In contrast, we could get a single crystal by means of quasi-racemic crystallization, from a mixture of glycosyl-L-CCL1 and D-CCLI as well as racemic crystallization, from a mixture of L-CCL1 and D-CCL1. Diffraction data from the quasi-racemic crystal and the racemic crystal have been acquired to 2.08 Angstroms and 2.0 Angstroms, respectively. Structural analyses revealed there are no significant structural differences between glycosyl-CCL1 and CCL1, and we could confirm the completion of the chemical synthesis of the glycoprotein.
Thielocin B1 (1), isolated from the fermentation broth of Thielavia terricola RF-143 in 1995, consists of five multi-substituted benzene rings, which are connected with a 2,2', 6,6'-biaryl ether and four ester linkages. Recently, it was found that 1 strongly inhibits protein-protein interactions (PPIs) of PAC3 homodimer (IC_<50>= 0.02 μM) without inhibition of other PPIs such as PAC1/PAC2 or TCF/P-catenin. Since we are interested in the mode of action mechanisms of 1, we performed total synthesis of 1, and docking studies by NMR and in silico analyses. The key intermediate 2,2', 6,6'-biaryl ether 4 was synthesized from 7-membered lactone 6, which was prepared by oxidative lactonization of benzophenone 7, followed by chemoselective reduction of the lactone and removal of the resulting alcohol. The side wing 5 was synthesized from aldehyde 8 via formation of 3 and its coupling by esterification. Condensation of biaryl ether 4 and 5 was smoothly performed using trifluoroacetic anhydride to afford 21 in high yield. Formylation of 21 by treatment with dichloromethyl methyl ether and AgOTf, followed by Kraus oxidation provided acid 2. Coupling of the resulting acid 2 with phenol 3 afforded 22 in 70% yield. Finally, removal of the benzyl groups by-hydrogenation furnished thielocin B1 (1), whose spectral data were in good agreement with those of the natural product.
In the course of our synthetic and structural studies on highly oxidized and structurally diverse cytotoxic triterpene polyethers, which are thought to be biogenetically squalene-derived natural products (oxasqualenoids), isolated from both marine and terrestrial organisms, we have been interested in the biogenesis of polyTHF rings-containing oxasqualenoids which might be synthesized in a single event by the cascade oxacyclization from acyclic precursor squalene polyepoxides.3 In this symposium, we report that the chemical synthesis of some THE rings-containing oxacyclic terpenoids, (-)-neroplofurol (1), (+)-ekeberin D4 (2), and (±)-glabrescol (3), has been accomplished on the basis of the hypothetical biomimetic oxacyclization triggered by acidic hydrolysis of the terminal epoxide (Scheme 2). In conclusion, we demonstrated that the biomimetic epoxide-opening cascades from acyclic precursor polyepoxides 4-6 to oxacyclic terpenoids 1-3, respectively, can be reproduced in a single event by chemical reaction (Scheme 3-5). These results experimentally support the Cane-Celmer-Westley hypothesis' as a general biogenesis of natural polyethers and accumulate chemical basis for the cyclization mechanism relative to the epoxide hydrolase Lsd19^2 in the biosynthesis of lasalocid A (Scheme 1).