WO2008125989A2 - Polyketide based proteasome inhibitor - Google Patents

Abstract

A method of inhibiting the proteasome comprising contacting the proteasome with a polyketide.

Description

Novel proteasome inhibitor

The invention relates to a novel proteasome inhibitor, to its manufacture, and to the use of the proteasome inhibitor in the treatment and prophylaxis of disease Background

The proteasome is a laige, multi-enzyme complex which iegulates intracellular protein concentrations, and plays an essential role in cell signalling, sun ival and apoptosis (D Nandi et al , J Biosci 2006, 31, 137-155, G N DeMartino and C A Slaughter, J Biol Chem 1999, 274, 22123-22126) The oveiall 26S proteasome has a baπel-hke structure, formed by a catalytic coie particle (20S) and two regulatory caps ( 19S) The core particle consists of foui stacked heptameπc nng structuies which are composed of α (structural) and β (catalytic) subunits Of the se\ en β subunits, three possess catalytic activity - βl (caspase-hke), β2 (tiypsin-hke) and β5 (chymotiypsin-hke) (M Gi oil et al , Stntctiii e 2006, 14, 451-456) The 19S cap comprises 19 individual proteins organized into two subassembhes, and has both ATP-ase activity and ubiquitin-bindmg sites Mammalian cells also contain a pioteasome activatoi called PA28 which is composed ot two 28 kDa subunits (σ and β) (Z Zhang et al , J Biol Chem 1998, 273, 30660-30668) Togethei, the α and β subunits form a football-shaped structure (C W Giay et al , J MoI Biol 1994, 236, 7- 15) which maikedly stimulates the activity ot the 2OS pioteasome in uti o (M Unno et al , Sn uctw e 2002, 10, 609-618) A thud membei ot the family, PA28γ, forms a homomultimei, which activates the proteasome^" s hydiolysis ot short peptides (G N DeMartino and C A Slaughter J Biol Chem 1999, 274, 22123-22126, S WiIk et al , 4ι ch Biochem Biopln s 2000, 383, 265-271 ) Together, the PA28 iegulatory components help to mediate the iole of the proteasome in immune function (A Si)ts et al , MoI Immunol 2002, 39, 165-169)

Intracellulai pioteins are targeted toi degiadation by the conjugation of polyubiquitin chains to lysine iesidues of the protein Thiee enzymes are involved in the ubiquitination piocess a ubiquitin-actn ating enzyme (E l ), a ubiquitin-corπugating enzyme (E2) and a ubiquitin hgase (E3) The proteasome recognises ubiquitin-tagged proteins via its regulatory 19S subunit and catalyses their proteolytic degradation in an ATP-dependent fashion Unwound proteins are fed down the central channel in the 2OS core and are subject to piogiessive degradation The ubiquitin-proteasome pathway represents the main protein degradation pathway in eukaryotic cells The pathway is also involved in regulating apoptosis, intracellular signalling, responses to cellular stress (i e DNA damage, hypoxis), the modulation of many regulatory proteins that affect inflammatory piocesses, cell adhesion and chemotaxis, viral shedding and diffeientiation A further role is in the generation of antigenic epitopes for presentation on human leukocyte antigen (HLA) molecules Thus, the proteasome plays an essential and ubiquitous role in normal physiological functions

In addition to its role in noπnal physiological functions, the deiegulation of the proteasome has been found to play a majoi iole in the development and progression of cancel For example, regulation via ubiquitination has been found to be essential to many pioteins involv ed in tumongenesis, such as cychns, cychn-dependent kinases and cychn-dependent kinase inhibitors Further, tumour suppiessoi genes and oncogenes aie also involved in the ubiquitin-proteasome pathway, often as examples of E3 hgases Thus, it the proteasome becomes deregulated, this in turn leads to the deiegulation ot the expiession of cell cycle stimulatory and inhibitory proteins It has been tound that proteasome inhibitois can limit tumour development by stabilizing cell cycle inhibitory pioteins and causing cell cycle aπest and apoptosis (Hershko A Roles of ubiquitin- mediated proteolysis in cell cycle contiol Curr Opin Cell Biol 1997,9 788-799) Some types of cancel have been found to be particulaily prone to undergo apoptosis in response to inhibition ot the ubiquitin-proteasome pathway (Mam, A and Gelmann, E P The ubiquitin-proteasome pathway and its iole in cancer J Clin Oncol 2005, 23, 4776- 4789, Adams, J The proteasome a suitable antineoplastic target Nat Rev Cancel, 2004, 4 349-360 and Adams J The dev elopment of proteasome inhibitois as anticancer drugs Cancer Cell 2004, 5 417-421 ) Howevei, it is unclear why blockade of the ubiquitin- proteasome pathw ay is more toxic to cancel cells than it is to noπnal cells The pioteasome has also been found to be inv olv ed in a number ot other physiological and pathological piocesses including diseases w hich inv olv e angiogenesis (foi example metastatic cancel, diabetic ietinopathy and iheumatoid arthntis), cancel such as myeloid myeloma, prostate cancer, pancreatic cancer, breast cancer, lung cancer, and ovaπan cancer, solid tumours and lymphomas such as Non-Hodgkin's lymphoma, mantle cell lymphoma and follicular lymphoma, retroviral diseases such as HIV, chronic inflammatory conditions such as asthma, ischemia and reperfusion injury, multiple sclerosis, rheumatoid arthritis, psoπasis, inflammatory and degenerative conditions such as Alzheimer's disease, amyotrophic lateral sclerosis, autoimmune thyroid disease, cachexia, Crohn's disease, hepatitis B, inflammatory bowel disease, sepsis, systemic lupus erythematosus, acute stroke, myocardial infections and transplantation rejection such as giaft-versus-host disease Thus, proteasome inhibitors are useful in the treatment and prophylaxis of these diseases

Furthermore, proteasome inhibition interferes with the processing of endogenous pioteins for presentation on HLA molecules and so favouis the presentation of exogenously added peptide epitopes by dendritic cells (DCs) (Chromik, J et al Proteasome-inhibited dendπtic cells demonstrate impioved presentation of exogenous synthetic and natuial HLA-class I peptide epitopes J Immunol Methods, 2006, 308 77- 89, Wong, C et al Induction of primary, human antigen-specific cytotoxic T lymphocytes in vitro using dendπtic cells pulsed with peptides, J Immunother , 1998, 21( 1 ) 32-40) Thus, DC treatment with proteasome inhibitors has been proposed to improve DC-based vaccinations Recently it has emerged that the proteasome plays additional iegulatoiy roles in DNA tiansciiption, including possible ioles tor the 19S subunit independently ot its role in the 26S proteasome (Collins, G A and Tansey, W P "The pioteasome a utility tool for transcription⁷'^" Cuπ Opin Genet Dev , 2006, Apr 16(2) 197-202) These roles appear to involve iecognition of ubiquitin adducts and include turnover ot transcription factor occupancy at promoters as well as regulation of transcriptional elongation via proteasome recognition ot histone 2B mono-ubiquitinylation (Hegde, A N and Upadhya, S C "Pioteasome and transcription a destioyei goes into construction", Bioessays, 2006, 28(3) 235-239, Ezhkova, E and Tansey, W P , "Pioteasomal ATPases link ubiquintylation ot histone H2B to methylation of histone H3^"', MoI Cell 2004, 13, 13(3) 4^5-442) In this case, iegulation by the pioteasome i elates to pi omotei clearance of RNA Polymerase II In othei cases, the proteasome has been ieported to inteiact with chiomatin modifieis, ptesumably also \ ιa ubiquitin adducts Although some evidence for the specificity of gene regulation by the proteasome in yeast has been observed, recent data from mammalian systems indicates that chromatin based regulatory mechanisms can co-ordinate gene expression programs. It is therefore likely that the proteasome will be involved in co-ordinate regulation of specific gene expression programs For example, mouse embryonic stem cells are maintained in perpetual self renewal due to a specific gene expression program This piogram is maintained, at least in part, by a chromatin mechanism that includes the ubiquitination of histone 2B Inhibition of this pathway leads to exit from self renewal that can be monitored by loss of the chaiacteπstic markers, such as Oct4 or Nanog, or the acquisition of differentiation markers such as Fgf4 or Brachury, amongst many others (AuId, K L et al , "Genomic association of the proteasome demonstrates overlapping gene regulatory activity with transcπption factor substrates'', MoI Cell , 2006, Mar 17; 21(6) 861 -871 )

This repiesents one example of gene regulation through a chromatin based mechanism Theie are a vaπety of these mechanisms so specificity can be achieved Selective inhibition of one mechanism can therefore specifically alter gene expression profiles, which can lead to a therapeutic outcome

A number of naturally occurring and synthetic proteasome inhibitois are known in the art Examples of naturally occurring pioteasome inhibitors are Lactacystin (Fenteany G et al Inhibition of pioteasome activities and subunit-specific amino-terminal threonine modification by lactacystin, Science, 1995,268 726-731 ), Eponemycin and Epoxomycin (Sin N et al Eponemycin analogues syntheses and use as probes of angiogenesis, Biooig Med Chem , 1998,6 1209-1217) and Aclacinomycin (Figueiredo-Perena ME et al , The antitumor drug aclacinomycin A which inhibits the degradation of ubiquitinated proteins, shows selectivity for the chymotrypsin-like activity of the bovine pituitary 2OS proteasome, J Biol Chem , 1996,271 16455-16459)

Examples of synthetic pioteasome inhibitors are

• Calpain inhibitor I (WiIk S et al Probing the specificity of the bovine pituitaiy multicatalytic proteinase complex by inhibitots, activatois, and by chemical modifications Biomed Biochim Acta 1991 ,50 471 -478, Oilowski M et al Evidence foi the piesence of five distinct pioteolytic components in the pituitaiy multicatalytic proteinase complex pioperties of two components cleaving bonds on the caiboxyl side of branched chain and small neutral amino acids. Biochemistry. 1993 ;32: 1563-

1572);

• PS-519 (synthetic agent similar to lactacystin) (Soucy J. et al. A novel and efficient synthesis of a highly active analogue of clastro-lactacystin beta-lactone. J Am Chem Soc. 1999; 121 :9967-9976);

• Aldehyde inhibitors (CEP-1612, MGl 32) An B, et al., Novel dipeptidyl proteasome inhibitors overcome Bcl-2 protective function and selectively accumulate the cychn- dependent kinase inhibitor p27 and induce apoptosis in transformed, but not normal, human fibroblasts. Cell Death Differ. 199S;5: 1062-1075. Adams J. et al. Potent and selective inhibitors of the proteasome: dipeptidyl boronic acids. Bioorg Med Chem

Lett. 1998;8:333-338; Lee DH, Goldberg AL. Proteasome inhibitors, valuable new tools for cell biologists. Trends Cell Biol. 1998;8:397-403);

• Benzamide (CVT-634) (Lum RT et al., Selective inhibition of the chymotrypsin-like activity of the 2OS proteasome by 5-methoxy- l -indanone dipeptide benzamides. Bioorg Med Chem Lett. 1998;8:209-214);

• Boronic acid inhibitors (PS-341 ) (Adams J. et al. Potent and selective inhibitors of the proteasome: dipeptidyl boronic acids. Bioorg Med Chem Lett. 1998;8:333-338);

• Vinyl sulfone tπpeptides (Bogyo M et al., Covalent modification of the active site threonine of proteasomal beta subunits and the Escherichia coli homologue HsI V by a new class of inhibitors. Proc Natl Acad Sci U S A. 1997;94'6629-6634); and

• HIV-I protease inhibitor (ritonavir) (Andre P et al. An inhibitor of HIV-I protease modulates proteasome activity, antigen presentation, and T cell responses. Proc Natl Acad Sci U S A. 199S;95.13120-13124)

Bortezomib (formerly PS-341 ) (Velcade™) has been approved by the Food and Drug Administration for the treatment of multiple myeloma patients who have received at least one prior therapy.

The object of the invention is the identification of further proteasome inhibitors. Summary of the invention

The inventors have surprisingly found that the macrocychc polyketide kendomycin is a proteasome inhibitor. Accordingly, the invention provides a method of inhibiting the proteasome compπsing contacting the proteasome with a polyketide.

The invention also provides the use of a polyketide as a proteasome inhibitor.

Also provided is a polyketide for use in treating a disease or disorder in which the proteasome is involved. The use of a polyketide in the manufacture of a medicament for the treating a disease or disorder in which the proteasome is involved is also encompassed Likewise, there is provided a method of treating a patient having a disease or disorder in which the proteasome is involved, wherein said treating comprises inhibiting the proteasome using a polyketide. In a further embodiment, the invention provides the use of a polyketide as an adjuvant in a peptide-based vaccine

Also provided is a method for screening for analogues or deπvatives of kendomycin that have activity against a disease or disorder in which the proteasome is involved, wherein the method involves determining the ability of the analogue or deπvative to inhibit a proteasome.

A further aspect of the invention relates to a method of producing kendomycin by generating it in a recombinant system

Detailed description

The invention provides a method of inhibiting the proteasome compπsing contacting the proteasome with a polyketide The polyketide may inhibit the 2OS subunit, the 19S subunit or the 26Ss subunitof the proteasome In one preferred embodiment, the polyketide inhibits the 26S form of the proteasome In a preferred embodiment, the polyketide forms a complex with the pioteasome chymotryptic site. More preferably, the polyketide forms a covalent adduct with the active site threonine of the proteasome chymotryptic site Preferably, the covalent adduct is formed through the C20 position of the polyketide Complexes of polyketides and these proteasome subunits form one aspect of the present invention

The polyketide for use in the invention may be any polyketide, for example, kendomycin erythromycin, rapamycin, epothilone, πfamycin oi aveπnectin Deπvatives of known polyketides are also encompassed within the scope of the invention Preferably, the deπvative differs fiom the known polyketide in such a way that the pharmacological function of the molecule is not affected Preferred polyketides are macrocyclic polyketide compounds Preferred macrocyclic polyketide compounds are those having at least one of an E-tπsubstituted olefin (more preferably, an E-13,14-tπsubstituted olefin), an aliphatic ansa chain, a tetiahydropyran πng and a quinone-methide chromophore (more preferably a quinone-methide-lactol chromophore) Preferably, the macrocylic polyketide compound has two, three or all of these structures Foi example, an ansa system in which a tetrahydropyran πng is diiectly attached to the quinone- methide chromophore within a macrocyclic scaffold is preferred More preferably, the polyketide is of the structure shown in Formula I Formula I

wherein R is selected trom the gioup consisting ol alkyl, alkenyl, alkynyl, and

X is a heteioatom

Piereiably, R is alkyl and, more pieteiably, R is methyl Preferably, X is an oxygen atom

In a moie pieterred embodiment, the polyketide has the sti uctuie shown in Formula II Formula II:

wherein

R is selected from the group consisting of alkyl, alkenyl, alkynyl, and X is a heteroatom.

Preferably, R is alkyl and, more preferably, R is methyl Preferably, X is an oxygen atom

Even more preferably, the polyketide is kendomycin, as shown in Formula III Pharmaceutically acceptable derivatives or salts of the aforementioned polyketides are also envisaged

Foπnula III:

Where reference is made in the present application, or the context so requires, it will be understood that the valency of functional group is completed with hydrogen atoms; for example, -O- becomes -OH.

It will be appreciated that the compounds of the present invention can exist in a vaπety of stereochemical forms, which will be apparent to one skilled in the art. Positions of variable stereochemistry include those indicated with wavy lines Except where specifically indicated, the present invention extends to all such stereochemical forms.

Chemical Definitions

Where the compounds according to this invention have at least one chiral centre, they may accordingly exist as enantiomers. Where the compounds possess two or more chiral centres, they may additionally exist as diastereomers. Where the processes for the preparation of the compounds according to the invention give πse to mixture of stereoisomers, these isomers may be separated by conventional techniques such as preparative chromatography. The compounds may be prepared in racemic form or individual enantiomers may be prepared by standard techniques known to those skilled in the art, for example, by enantiospecific synthesis or resolution, formation of diastereomeπc pairs by salt formation with an optically active acid, followed by fractional crystallization and regeneration of the free base. The compounds may also be resolved by formation of diastereomeric esters or amides, followed by chromatographic separation and removal of the chiral auxiliary. Alternatively, the compounds may be resolved using a chiral HPLC column. It is to be understood that all such isomers and mixtures thereof are encompassed within the scope of the present invention.

Where a group comprises two or more moieties defined by a single carbon atom number, for example, C₂-₄ alkyl, the carbon atom number indicates the total number of carbon atoms in the group.

As used herein, the teπn "heteroatonT includes N, O, S and P. As used herein, the teπn "alkyl^"' refers to a straight or branched saturated monovalent hydrocarbon radical, having the number of carbon atoms as indicated. For example, the term "C|.₄-alkyl" includes Ci, C₂, C₃ and C₄ alkyl groups. By way of non-limiting example, suitable alkyl groups include methyl, ethyl, propyl, /sopropyl, butyl, /so-butyl and te/7-butyl. Preferred ranges of alkyl groups of the present invention are: Cι-₄-alkyl, C|.3-alkyl and C|.2-alkyl.

As used herein, the term "alkenyl" refers to a straight or branched unsaturated monovalent hydrocarbon radical, having the number of carbon atoms as indicated, and the distinguishing feature of a carbon-carbon double bond. For example, the teπn "C:-₄-alkenyl" includes C₂, C ₃ and C₄ alkenyl groups. By way of non-limiting example, suitable alkenyl groups include ethenyl, propenyl and butenyl, wherein the double bond may be located anywhere in the carbon chain. Preferred ranges of alkenyl groups of the present invention are: C₂-₄-alkenyl and C₂-₃-alkenyl.

As used herein, the teπn "alkynyl^" refers to a straight or branched unsaturated monovalent hydrocarbon radical, having the number of carbon atoms as indicated, and the distinguishing feature of a carbon-carbon triple bond. For example, the teπn "C₂-₄ alkynyl" includes C₂, C₃ and C₄ alkynyl groups. By way of non-limiting example, suitable alkynyl groups include ethynyl, propynyl and butynyl, wherein the triple bond may be located anywhere in the carbon chain. Preferred ranges of alkynyl groups of the present invention are: C₂-₄-alkynyl and C₂-₃-alkynyl. Certain compounds of the invention exist in various regioisomeric, enantiomeric, tautomeric and diastereomeric forms. It will be understood that the invention comprehends the different regioisomers, enantiomers, tautomers and diastereomers in isolation from each other as well as mixtures.

Kendomycin

Kendomycin is a macrocyclic polyketide that can be isolated from the mycelium of Streptomyces violaceoruber (Bode, H. B. and Zeeck, A. J. Chem. Soc. Perk. T 1 2000, 323-328). The structure of kendomycin is shown in Figure 9 and in Formula III. It has the molecular formula C₂₉H₄₂O₆ and a molecular weight of 486.6 kDa. The structure of kendomycin features an aliphatic ansa chain with a highly substituted tetrahydropyran ring attached to an unique quinone methide chromophore. Initially, kendomycin [(-)-TAN 2162] was shown to be a potent endothelin receptor antagonist (Funahashi, N. et al., Jap. Pat. 08,231 ,551 , [A2 960 910] 1996, Chem. Abstr. 1997, 126, 6553; Funahashi, Y. et al. Jap. Pat. 08, 231 ,552 [A2 960910], 1996, Chem. Abstr. 1996, 125, 326518). In 1998, studies performed in the United States revealed its anti-osteoporotic activity (Su, M.H. et al., US 5,728,727 [A 9803 17], 1998 (Chem. Abstr. 1998, 128, 239489). Kendomycin has also been found to display antibacterial activity against multi-drug resistant Staphylococcus aureus (MRSA) strains and other Gram-positive and Gram-negative organisms. In vitro cytotoxicity tests performed against three cancer cell lines: stomach adenocarcinoma (HMO2), hepatocellular carcinoma (HEPG2) and breast adenocarcinoma (MCF7) cell lines, revealed that kendomycin displays potent cytotoxicity against these cell lines (GI₅0 < 0.1 μM) (Bode, H. B. and Zeeck, A. Structure and biosynthesis of kendomycin, a carbocyclic ansa- compound from Streptomyces, J. Chen. Soc, Perkin Trans. 1 , 2000, 323-328). Remarkable, kendomycin was 20 times more active than cisplatin against breast adenocarcinoma (MCF-7) cells (TGI = 0.5 uM, compared to TGI = 10 uM for cisplatin) and 25 times more active against hepatocellular carcinoma (HEP G2) cells (TGI = 0.2 uM, compared to TGI = 5 uM for cisplatin). Activity was also comparable to the clinically-used anti-cancer drug doxorubicin. The reported data on the antitumor activity of kendomycin were obtained at a detection limit of 0.1 uM (i.e. the most diluted solution of kendomycin had a concentration of 0.1 uM) and the GIso was below this value for the three cell lines tested. However, although cytotoxicity against these three specific cell lines was demonstrated, the experiments earned out by Bode and Zeeck did not examine the cytotoxicity of kendomycin for noπnal cells, neither did they text the cytotoxicity of kendomycin against other cancer cell lines.

The cellular target(s) of kendomycin were not known prior to the present invention. The inventors have surprisingly found that kendomycin inhibits the proteasome. The invention therefore provides a method of inhibiting the proteasome comprising contacting the proteasome with a polyketide. Likewise, the use of a polyketide as a proteasome inhibitor is provided. The proteasome inhibition may be earned out in vitro or m vivo

As regards inhibition in vitro, the invention provides a method of inhibiting a proteasome comprising contacting the proteasome //; vitro with a polyketide An example of an industnal application of the in vitro method is the use in test methods for determining the function of the proteasome in normal physiological processes and in pathological processes An in vitro method is also of use in a screen for identifying further proteasome inhibitors Preferably, in the w vitro screening method, a control experiment is also earned out in which the proteasome is contacted in \ ιtro with a known proteasome inhibitor and the results of the method using the polyketide are compared with the control to detennine whether inhibition has occurred The proteasome is preferably in a cellular extract or in cell culture.

As regards the in vn o use, the invention provides a method of treating a patient having a disease or disorder in which the proteasome is involved, wherein said treating compnses inhibiting the proteasome using a polyketide Likewise, there is provided a polyketide for use in treating a disease or disorder in which the proteasome is involved, wherein said treating involves inhibiting the proteasome Preferably, such diseases or disorders are those diseases in which abnonnal proteasome activity plays a role. By "'abnormal ^" proteasome activity is meant a proteasome activity which differs between a representative sample of the population that have a particular disease or disorder and a representative sample of the population which do not have that particular disease or disorder Proteasome activity can be tested by methods known in the art. For example,

CHEMICON^'s 2OS Proteasome Activity Assay Kit provides a quick, convenient, and sensitive method for the detection of intracellular proteasome activity It is useful for measuring proteasome activity in cell lysates, /;; \ ιtro inhibitor screening, and testing purified proteasome enzyme A further example of a suitable method is described in Verdoes, M. et al., Chemistry and Biology, 13(1 1 ), 1217-1226, 2006. In the method of Verdoes et al., the activity of the proteasome is determined using fluorogenic substrates to determine chymotrypsin-like activity, trypsin-hke activity and peptidylglutamyl peptide hydrolytic activity. Diseases or disorders in which the proteasome is involved include diseases which involve angiogenesis (for example metastatic cancer, diabetic retinopathy and rheumatoid arthritis), cancer such as myeloid myeloma, prostate cancer, pancreatic cancer, breast cancer, lung cancer and ovarian cancer, solid tumours and lymphomas such as Non-Hodgkin^"s lymphoma, mantle cell lymphoma and follicular lymphoma; retroviral diseases such as HIV, chronic inflammatory conditions such as asthma; ischemia and reperfusion injury; multiple sclerosis; rheumatoid arthritis; psoπasis; inflammatory and degenerative conditions such as Alzheimer's disease; amyotrophic lateral sclerosis; autoimmune thyroid disease; cachexia, Crohn's disease, hepatitis B; inflammatory bowel disease, sepsis, systemic lupus erythematosus; acute stroke, myocardial infections and transplantation rejection such as graft-versus-host disease Thus, proteasome inhibitors are useful in the treatment of these diseases. Preferably, the cancer to be treated is not selected from stomach adenomcarcinoma HM02, hepatocellular carcinoma HEP G2 and breast adenocarcinoma MCF7 More preferably, the cancer to be treated is not selected from stomach adenomcarcinoma, hepatocellular carcinoma and breast adenocarcinoma

Preferred patient groups aie those that have been found to have abnormal pioteasome activity For example, wheie the disease to be treated is cancer, the patient is preferably a patient that has been found to have an abnormal proteasome activity Whether a patient has normal or abnormal proteasome activity can be determined by standard methods known in the art, such as those descπbed above

In some embodiments, during treatment, the polyketide forms a complex with the proteasome chymotryptic site and/or caspase is activated Preferably, the caspase is caspase-3 and/or caspase-8

The term "ti eating," as used herein, refers to retaiding or reversing the progress of, or alleviating or preventing either the disease or disorder to which the term "treating" applies, or one or more symptoms of such disease or disoider The term "tieatment," as used herein, refers to the act of treating a disease or disorder, as the term "treating" is defined above

In order to use a polyketide or a pharmaceutically acceptable salt thereof for the treatment of a patient, it is normally formulated in accordance with standard pharmaceutical practice as a pharmaceutical composition. Therefore in another aspect the polyketide is in the form of a pharmaceutical composition which compπses the polyketide (or a pharmaceutically acceptable deπvative or salt thereof) and one or more pharmaceutically acceptable earners and/or diluents. Suitable earners and/or diluents are well known in the art and include phannaceutical grade starch, mannitol, lactose, magnesium stearate, sodium saccharin, talcum, cellulose, glucose, sucrose, (or other sugar), magnesium carbonate, gelatin, oil, alcohol, detergents, emulsifiers or water (preferably stenle).

The polyketide may be administered by any suitable route For example, the administration may be by the oral, lingual, sublingual, buccal, rectal, topical, intravenous, intraarterial, intracardiac, subcutaneous, intranasal, transdennal, intramuscular, intraperitoneal, parenteral, intravaginal or intra-rectal routes. For these purposes the compounds of this invention may be fonnulated by means known in the art into the fonn of, for example, tablets, capsules, aqueous or oily solutions, suspensions, emulsions, creams, ointments, gels, nasal sprays, suppositories, finely divided powders or aerosols or nebulisers foi inhalation, and for parenteral use (including intravenous, intramuscular or infusion) stenle aqueous or oily solutions or suspensions or stenle emulsions

Phannaceutical compositions will typically compnse a therapeutically effective amount of the polyketide The term "therapeutically effective amount" as used herein, refers to an amount sufficient to detectably treat, ameliorate, prevent or detectably retard the progression of an unwanted disease or disorder or symptom associated with a particular disease or disorder The theiapeutically effective amount of the compound of the present invention depends on the administration route, the age, weight and sex of the patient, and the conditions of the disease or disorder to be treated Preferred therapeutically effective amounts aie in the range of 0 01 mg to 1000 mg of polyketide per dose, for example, 0 01 mg to 10 mg, 10 mg to 200 mg, 0 1 mg to 500 mg, 5 mg to 700 mg, 300 mg to 1000 mg or 400 mg to 650 mg per dose The administration may be earned out in single or multiple doses, for example, two, three, four, five or more doses Suitably, the polyketide may be administered for a period of continuous therapy, for example for a week or more, a month or more, three months or more, six months oi more or a year or more Duration of treatment will vary depending upon the particular disease or disorder being treated, the patient being treated, the treatment setting, etc In some cases, treatment might consist of a single administration of the compound In some embodiments, the compound will be administeied multiple times over a peπod of time sufficient to meet the needs of the particular situation Foi example, the compound may be administered at least twice, at least three, at least four, at least five, at least ten, at least fifteen oi at least twenty times over a penod of one day, one week, one month, two oi more months, six or more months or one or more years

The invention also extends to the use of a prodrug of a polyketide, such as an ester or amide thereol A prodrug is any compound that may be converted under physiological conditions or by soK olysis to a polyketide tor use in the invention or to a pharmaceutically acceptable salt ot a polyketide toi use in the invention

In addition to the treatment ot human subjects, the therapeutic methods ot the invention also will have significant veteπnaiy applications, e g toi treatment of livestock such as cattle, sheep, goats, cows, swine and the like, poultry such as chickens, ducks, geese, tuikeys and the like, horses, and pets such as dogs and cats

A polyketide may be administered simultaneously, subsequently or sequentially with one or more other actn e agents For example, the pharmaceutical composition may optionally be a combined prepaiation tor simultaneous, sepaiate or sequential use (including admimstiation) For example, when the polyketide is for use in treating cancer, the polyketide may be administered in combination with a know n chemotherapeutic compound, such as cisplatin or doxorubicin

Proteasome inhibition is known to interfere with the processing ot endogenous pioteins for presentation on HLA molecules and so tavours the presentation ot exogenously added peptide epitopes by dendritic cells (DCs) In a further embodiment, the invention theietoie provides the use ot a polyketide as an ad|uvant in a peptide-based vaccine Preteiably, the polyketide is used as an ad)uvant in a dendiitic cell-based \ accine in which the peptides are displa_\ed on the dendiitic cells The v accine itself may be formulated as a pharmaceutical composition, which may include one or more pharmaceutically acceptable earners and/or diluents as descnbed above. The vaccine is of particular use for treating or preventing a disease in which the presentation of exogenously added peptides is a particularly advantageous treatment method. Such diseases include cancer

Also encompassed by the invention is the use of a polyketide to alter a gene expression profile in a cell For example, when applied to a cancer stem cell, an altered profile caused by exposure of the cell to the polyketide may dπve the stem cell into exit from self-renewal, so depleting the stem cell pool and reducing the tumour burden Accordingly, a polyketide compound according to the invention, such as kendomycin, may be used to alter gene expression profiles in cells, either in vivo, in vitro or ex vivo, to effect a therapeutically beneficial outcome Preferably cells treated according to the invention are stem cells, particularly cancer stem cells

Also provided is a method for screening for polyketides that have activity against a disease or disorder in which the proteasome is involved, wherein the method involves determining the ability of the polyketide to inhibit a proteasome

Piefeiably, the polyketide is an analogue or derivative of kendomycin More preferably, the analogue or derivative of kendomycin is an analogue oi derivative that falls within the structural formula piovided as Formula 1 herein An "inhibitoi " of a pioteasome includes any molecule which deci eases the activity of the pioteasome Thus, a proteasome inhibitor can be a molecule which decreases activity of the proteasome, e g by interfering with interaction of the proteasome with another molecule, e g , its substiate In pieferred embodiments, inhibition may be effected by binding of the polyketide itself to the proteasome, foi example by allosteπc or steπc regulation or by charge-charge interactions

The inhibitor preferably affects one or more of the 20S proteasome, the 19S subunit, the 26S proteasome or a different part of the pioteasome

In one embodiment, the inhibitoi should exhibit an ICso value for inhibition of the pioteasome of about 100 μM oi less, 50 μM or less, 15 μM or less, 5 μM or less, piefeiably 500 nM oi less, more pieferably 100 nM or less ICs₀ values aie calculated using the concentration of inhibitoi that causes a 50% deciease in activity as compared to a control. This evaluation can, for example, be accomplished through conventional in vitro assays, such as by using one or more of the assays described below.

Preferably, the polyketide inhibits the proteasome by at least 10%, 25%, 50%, 75%, 85%, 90%, 95% or 100% relative to when the inhibitor is not present These IC50 values and the relative % levels of inhibition are preferably measured using the commercially-available Calbiochem assay for determining proteasome activity, which uses SDS-activated 2OS proteasomes from rabbit reticulocytes Using this kit, the chymotryptic activity of the proteasome is measured by monitoring the release of free 7- amino-4-methylcoumaπn (AMC) from the fluorogenic peptide Suc-Leu-Leu-Val-Tyr- AMC (L. J Crawford et al , Cancer Res 2006, 66, 6379-6386; E S. Lightcap et al Proteasome inhibition measurements: Clinical application, Clin. Chem., 46, 673-683, 2000) Proteasome activity can alternatively be determined by any suitable method known in the art.

In general, screening procedures for determining whether a polyketide is an inhibitor of the proteasome may involve using appropπate cells that express the proteasome that are contacted with a test polyketide to observe binding, or inhibition of a functional response The functional response of the cells contacted with the test polyketide is then compared with control cells that were not contacted with the test polyketide. Inhibitors of activation are generally assayed in the presence of a known agonist and the effect on activation by the agonist in the presence of the test compound is observed

A binding assay is a fairly inexpensive and simple in vitro assay to run Binding of a polyketide to a proteasome, in and of itself, can be inhibitory, due to steπc, allosteπc or charge-charge interactions By using this as an initial screen, one can evaluate libraries of polyketides for potential proteasome inhibitory activity The target proteasome in a binding assay can be either free in solution, fixed to a support, or expressed in a cell. A label (e g. radioactive, fluorescent, quenching, etc ) can be placed on the proteasome, the polyketide, or both to determine the presence or absence of binding Adherence of a polyketide to a surface bearing the proteasome may be detected by means of a label directly or indirectly associated with the polyketide or in an assay involving competition with a labelled competitor

For purposes of ui Mtro cellular assays, the polyketides that represent potential inhibitois of proteasome function can be administered to a cell in any numbei of ways Preferably, the polyketides can be added to the medium in which the cell is growing, such as tissue culture medium for cells grown in culture. The polyketide is preferably provided in standard serial dilutions or in an amount determined by analogy to known modulators. This approach can also be used to conduct a competitive binding assay to assess the inhibition of binding of the proteasome to a natural or artificial substrate or binding partner. In any case, one can measure, either directly or indirectly, the amount of free label versus bound label to determine binding. There are many known vaπations and adaptations of this approach to minimize interference with binding activity and optimize signal. For example, a competitive drug screening assay may be used, in which neutralising antibodies that are capable of binding the proteasome specifically compete with a test polyketide for binding. In this manner, the antibodies can be used to detect the presence of any test polyketide that possesses specific binding affinity for the proteasome. Another technique for polyketide screening which may be used provides for high throughput screening of polyketides having suitable binding affinity to proteasomes (see International patent application WO84/03564) In this method, large numbers of different polyketides are immobilised on a solid support, which may then be reacted with the proteasome and washed Bound proteasomes may then be detected using methods that are well known in the art

Persons skilled in the art will be able to devise alternative assays for identifying inhibitors of a proteasome. Of interest in this regard is Lokker NA et al , J Biol Chem , 1997, Dec 26,272(52):33037-44, which reports an example of an assay to identify antagonists (in this case neutralizing antibodies) Further, in vitro assays can also assess the ability of the polyketide to block an identified downstream effect of the proteasome, for example trypsin-hke activity, chymotrypsin-like activity and/or peptidylglutamyl peptide hydrolytic activity (Verdoes, M et al , Chemistry and Biology, 13( 1 1 ), 1217- 1226, 2006). An in vitro assay may alternatively monitor whether a putative proteasome inhibitor has the same effect as a known proteasome inhibitor, e g the ability to induce apoptosis would suggest that a polyketide is a proteasome inhibitor

Preferably, the method for screening for analogues or derivatives of kendomycin that have activity against a disease or disorder in which the pioteasome is involved, comprises the steps of: i) selecting as a test compound a polyketide; ii) incubating the test compound with SDS-activated 2OS proteasomes from rabbit reticulocytes to produce a test combination; iii) incubating a known proteasome inhibitor with SDS-activated 2OS proteasomes from rabbit reticulocytes to produce a control combination; iv) measuring the release of free 7-amino-4-methylcoumarin (AMC) from the fluorogenic peptide Suc-Leu-Leu-Val-Tyr-AMC in the test and control combinations; v) comparing the measurements made in step iv) between test and control combinations, wherein if the amount of release of free AMC is comparable between the test and control combinations, the test compound is a proteasome inhibitor.

Preferably, the known proteasome inhibitor used in step iii) is kendomycin. Alternatively, the screening method may comprise the steps of: contacting the proteasome with a fluorescently labelled putative proteasome detector and using in gel detection to determine whether the proteasome is labelled , as described in Groll, M. and Huber, R., Biochim. Biophys. Acta,1695( l -3):33-44, 2004 and in Verdoes, M. et al., Chemistry and Biology, 13( 1 1 ), 1217-1226, 2006. A library of analogues or derivatives may be tested by using a multiwell plate format.

Also encompassed within the scope of the invention are polyketides that are obtained using a screening method as described above, and the use of such analogues or derivatives in the methods described herein. Preferably, these polyketides are new analogues or derivatives of kendomycin. The invention provides a nucleic acid encoding the enzymes of the biosynthetic pathway for kendomycin, wherein the nucleic acid comprises or consists of the sequence provided in SEQ ID NO: 1. Also provided is a variant of a nucleic acid of SEQ ID NO: 1 , wherein the variant has a level of sequence identity of 80% or more, preferably, 90% or more, 95% or more, 98% or more, 99% or more or 99.5% or more to the sequence shown in SEQ ID NO: 1 . Preferably, the level of sequence identity is determined across the entire length of SEQ ID NO: 1 . Nucleic acids comprising or consisting of fragments of a nucleic acid of SEQ ID NO 1 or of a vaπant thereof are also provided Preferably, the fragment compnses at least one gene encoding an enzyme of the biosynthetic pathway for kendomycin Preferably, the fragment is at least 200 nucleotides in length More preferably, the fragment is at least 500, at least 1000, at least 5000, at least 10000, at least 20000, at least 30000, at least 35000, at least 40000, at least 45000, at least 50000, at least 60000, at least 65000, at least 66000, at least 66550, at least 66590 nucleotides in length

The invention further provides a method for the heterologous production of kendomycin by generating it in a recombinant system At present, it is only possible to produce kendomycin by isolating it from its natural host - Streptoim ces \ ιolaceoι uber Streptomvces vwlaceorubet (strain 3844-33C) initially attracted attention by its ability to degrade natural rubber (Jendrossek, D et al , FEMS Microbiol Lett 1997, 150, 179) By using complex media, this stiain was stnking when its extracts were applied to a chemical selection program (Giabley, S et al in Drug Discovery fiom Nature, ed S Grabley and R Thieπcke, Spπngei, Beilin-Heidelberg, 1999, 124-128) Kendomycin was isolated from the mycelium (70 mg/1) due to its intensive yellow colour and was identified as (-)-TAN 2162 by comparison of the physicochemical data Kendomycin is soluble in DMSO or methanol

Synthetic studies toward kendomycin have been initiated by a number of researchers (Bode, HB and Zeeck, A Structure and biosynthesis of kendomycin, a carbocyclic ansa- compound from Streptomyces J Chem Soc Pei k T l 2000,323-328, Smith, A B , III, Mesaros, E F , Meyei, E A , J Am Chem Soc 2005, J27( 19), 6948-6949) Howevei, kendomycin has a complex structuie and so its synthesis by standard synthetic routes is highly laborious The present invention provides a highly convergent, steieo-contiolled total synthesis of kendomycin via a strategy that allows toi an inexpensive large-scale pioduction of kendomycin The method provided by the present invention may also, with minoi modifications, be used to make a library of analogues The method of producing kendomycin preferably compnses expressing one or moie ot the proteins encoded by a nucleic acid of the invention in a recombinant system Preferably, the one oi moie pioteins aie encoded horn said nucleic acid in the iecombinant system Preferably, the method compnses expiessing all ot the pioteins encoded by a nucleic acid ot the invention in a recombinant system Preferably, the one or more proteins are encoded from said nucleic acid in the recombinant system Methods for the heterologous expression of polyketides are known in the art For example, a detailed protocol is descπbed in WO2006/046152 (Gene Bπdges GmbH), the content of which is incorporated herein by reference in its entirety

Preferably, the component genes of the biosynthetic pathway foi kendomycin have the sequence shown in SEQ ID NO 1 The sequence encoding the component genes of the biosynthetic pathway for kendomycin may alternatively have a level of sequence identity of 90% or more, preferably, 95% or more, 98% or more, 99% or more or 99 5% or more to the sequence shown in SEQ ID NO 1

Preferably, all of the genes encoding the biosynthetic pathway for kendomycin are comprised on a single vector However, the genes may alternatively be comprised on two, three, four, five or more vectois For example, if not all of the genes of the biosynthetic pathway are contained on one vector, the other genes lequired for activity of the pathway may be provided either on one or more additional vectois or may be integrated onto the chromosome of the second host cell, either naturally, or through directed chromosomal integration Further, the older of the genes in the biosynthetic pathway may be changed so that it differs from the ordei of genes descπbed in SEQ ID NO 1 If a host cell is used to expiess kendomycin, which already comprises part of the biosynthetic pathway for kendomycin, the conesponding portion of the gene cluster may be omitted from the one oi more vectors used in the methods of the invention Preferably, all the genes of the biosynthetic pathway that aie not aheady present in the host cell are encoded on one or more vectois

The method may additionally comprise the step of transforming the host cell with genes encoding enzymes required for making substrates that are required to synthesise kendomycin, but which aie not endogenously expressed in the host cell

Examples of suitable vectors will be known to those of skill in the art, and may be selected iationally to suit the requπements of any particular system, taking into account information known about the length of sequence to be cloned, the type of second host system to be used, and so on Of particular suitability will be episomal and virus-denved systems dein ed fiom bacteiial plasmids, bactenophage, cosmids and phagemids, and bacterial artificial chromosomes (BACs). BACs in particular may also be employed to deliver larger fragments of DNA than can be contained and expressed in a plasmid.

The one or more vectors are preferably constructed using the principles of recombineering. Recombineering (also known as "Red/ET recombination technology") is an ideal tool for large size DNA engineering and is well known in the art (see International patent applications WO99/29837 and WO02/062988; European patent applications 01 1 17529.6 and 0103276.2; United States patents 6,509, 156 and 6,355,412; and also Muyrers, J. P. P. et at., 2000, ET-Cloning: Think Recombination First, Genetic Eng., vol. 22, 77-98; Muyrers, J. P et at., 2001 , Techniques: Recombinogenic engineering-new options for cloning and manipulating DNA, Trends in Biochem. Sci., 26, 325-31 ; Zhang, Y et at., 2000, DNA cloning by homologous recombination in Escherichia coti., Nature Biotech., 18, 1314-1317; Muyrers J. P et at., 2000, Point mutation of bacterial artificial chromosomes by ET recombination, EMBO Reports, 1 , 239-243; Muyrers J. P et at., 2000, RecE/RecT and Redαa/Redβa initiate double-stranded break repair by specifically interacting with their respective partners, Genes Dev., 14, 1971-1982; Muyrers et at., 1999; Rapid modification of bacterial artificial chromosomes by ET-recombination, Nucleic Acid Res., 27, 1555- 1557; Zhang Y. et at., 1998, A new logic for DNA engineering using recombination in Escherichia coti, Nat. Genet., 20, 123-128; Narayanan K. et al., Efficient and precise engineering of a 200 kb β-globin human/bacterial artificial chromosome in E. coli DHl OB using an inducible homologous recombination system, Gene Therapy, 6, 442-447 and Zhang, Y. et at., 2003, BMC MoI Biol. 2003 Jan 16; 4 (1 ): 1 ; the contents of which are all incoiporated herein by reference).

A host cell is chosen that is capable of expressing the enzymes of the kendomycin biosynthetic pathway and which can, under suitable culture conditions, generate kendomycin. Preferably the host cell is not a natural producer of kendomycin. The host cell is preferably a Pseudomonas, a Mycobacterium or a Streptomycete (preferably

Streptomyces coelicolor). When the host cell is a Streptomycete, it is preferably not

Strcptomvces violaceoriiber strain 3844-33C, more preferably not Streptomyces violaceoriiber. The host cell should be cultured under conditions which are suitable for synthesis of kendomycin. Suitable conditions for growth of the host cell will be known to those of skill in the art. The genes of the biosynthetic pathway comprised in SEQ ID NO: 1 may be under the control of a single promoter. Alternatively, the genes of the biosynthetic pathway comprised in SEQ ID NO: 1 may be under the control of more than one promoter. For example, each gene may be under the control of a separate promoter. These promoters may all be the same promoter, or may be selected from two or more different promoters, or may each be a different promoter.

According to the methodology of the invention, the genes of the biosynthetic pathway are transcribed under the control of promoters that are functional in the host cell. Preferably, the promoters are found naturally in the host cell. This is an important element of the methodology of the invention, for it allows the transcription machinery of the host cell to recognise the promoters and thus transcribe the genes implicated in the biosynthetic pathway. In preferred systems according to the invention, an inducible promoter is used in one or more of the genes that form part of the biosynthetic pathway; in these systems, the inducing agent will preferably be added once the host cells have attained a high cell density. This will minimise cell death during earlier stages of growth as a result of potential toxicity of the kendomycin produced.

In a preferred embodiment, a method for the heterologous expression of kendomycin is provided, comprising: i) generating in a first host cell, a single vector comprising the component genes of the biosynthetic pathway for kendomycin; ii) transforming a second host cell with the vector; iii) culturing the second host cell under conditions which are suitable for synthesis of kendomycin; and wherein the genes of the biosynthetic pathway are transcribed under the control of promoters that are functional in the second host cell.

The first host cell is also preferably a host cell that is able to conjugate efficiently with the second host cell. For example, E. coli is a preferred first host cell. However, other suitable first host cells include other gram negative bacteria, particularly those that are well studied such as Pseudomomids and Salmonella species. Methods for genetic engineering of E. coli and Salmonella are described in full in known laboratory manuals such as that by Sambrook ct ai, Molecular Cloning; A Laboratory Manual, Third Edition (2001 ). The second host cell is a host cell which is capable of generating kendomycin under suitable culture conditions.

In one embodiment, only part of the kendomycin molecule is synthesised in the host cell, and this part is then used to synthesise the complete kendomycin molecule by way of standard chemical synthetic techniques. It is also envisaged that more than one part may be synthesised in more than one host, and that these parts may be used to synthesise the complete kendomycin molecule.

The methodology of the invention may be performed iteratively, with successive rounds of screening and selection in order to allow the molecular evolution of one or more of the genes that participates in the pathway toward a desired function Indeed, an entire pathway can be evolved in this fashion

For example, the genes encoding the enzymes of the biosynthetic pathway may optionally be further genetically engineered Mutagenesis of the genes encoding the enzymes is an advantageous way to alter the chemical product because the structure of kendomycin is directed by the specificity of the enzymes of the biosynthetic pathway. For example, genetic engineering may enable an increase in the half-life of kendomycin or may increase its specific activity Genetic manipulation of this type may be earned out by shuttling a vector selected in the second host cell back into the first host cell, or may be canned out directly in the host cell which is capable of generating kendomycin under suitable culture conditions, or in a further host cell

Because of the lelative ease of genetic manipulation in the cloning host, it is likely that in most circumstances, genetic manipulation will be effected in a first cloning host cell and then the vector transfonned back into a second host cell for screening and selection The use of a first host in which genetic engineering techniques are well established enables genetic engineering to be earned out with a high degree of accuracy and in particular enables site-directed mutagenesis to be earned out in order to alter kendomycin specifically Random and/or combinatonal mutagenic approaches may alternatively or additionally be used for the creation of hbranes of mutations, including appioaches such as DNA shuffling, STEP and sloppy PCR, and molecular evolution. A random and/oi combinatorial approach enables libraries of different analogues and derivatives of kendomycin to be created The genetic engineering of one or more genes in the biosynthetic pathway may involve any suitable type of mutagenesis, for example, substitution, deletion or insertion mutagenesis. If the sequence encoding the one or more genes contains redundant, irrelevant and potentially undesirable sequences, genetic engineering can be carried out to remove these sequences from the vector. Mutagenesis may be carried out by any suitable technique known in the art, for example, by site-directed mutagenesis or by transposon-mediated mutagenesis, as the skilled reader will appreciate. Site-directed mutagenesis may be used to insert new restriction sites, alter glycosylation patterns, change codon preference, produce splice variants, introduce mutations and so forth. Recombineering may also be used where appropriate.

Analogues and derivatives of kendomycin that are produced using the aforementioned methods may be screened for proteasome inhibition activity using a method according to the invention.

Methods for obtaining the kendomycin from the cell culture are well known in the art. An example of a suitable method is described in Bode, H. B. and Zeeck, A. (J. Chem. Soc, Perkin Trans 1 , 2000, 323-328). Briefly, the method of Bode and Zeeck involves centrifuging the fermentation broth and discarding the supernatant. The myocelial pellet is then treated repeatedly with acetone until extraction of kendomycin is complete. After evaporation of acetone, the remaining aqueous residue is lyophilised. The crude product of kendomycin is extracted with CH₂CI₂ in a Soxhlet extractor. The residue is altered using a short silica gel column and the filtrate is discarded. Kendomycin is then eluted with ethyl acetate. After evaporation of the solvent, the crude extract is crystallised.

The invention will now be described further, by way of example only, by reference to the following figures in which: Figure 1 shows the results of a cell propagation assay of U-937, KB-3-1 cell lines upon exposure to kendomycin. U-937 (o) and KB-3-1 cells (•) were incubated with a range of kendomycin concentrations for 5 days. The metabolic activity of the cells was then measured using MTT [http://www.atcc.org/common/documents/pdf/30-1010k.pdf]. The percentage growth was calculated by dividing the absorbance of the treated cells by the absorbance of control cells.

Figure 2 shows the results of a two-color JC- I analysis of mitochondrial membrane potential (Δψ) and light-scattering. Cells were treated with 2 μM kendomycin for 0, 20, 40 and 120 min (panels A, B, C and D, respectively). Cells were washed and labeled with JC-I and the fluorescence was measured on a flow cytometer using FLl (green) and FL2 (red) channels. The data reveal a time-dependent increase in green fluorescence with a concomitant decrease in red fluorescence, consistent with an overall decrease in mitochondrial membrane potential. Light scattering detection further shows an increase in the population of apoptotic cells 120 min after kendomycin treatment (F), relative to control cells (E).

Figure 3 provides evidence for apoptosis in U-937 cells. A) Cells were treated with 0, 1 and 2 μM kendomycin for 24 h, and then analyzed for DNA content after propidium iodine staining using flow cytometry. Kendomycin induces a dramatic increase in the sub-Gl population ( HD ), in a concentration-dependent manner. The relative distribution of cells within the phases of the cell cycle (GP ,SD , and G₂/MB ) does not appear to be affected. B) U-937 cells were treated with 2 μM kendomycin for different periods of time. Activation of caspase-3 in cell extracts was then determined by measuring the release of fluorogenic AMC (7-amino-4-methylcoumarin) from the substrate Z-DEVD- AMC. This analysis shows that kendomycin induces a time-dependent increase in caspase 3 activation. C) U-937 cells were treated with 1 or 2 μM kendomycin for 0 (lane 2) and 4 h (lane 1 ). DNA was then extracted and fragments separated by agarose gel electrophoresis. Lanes 3 and 4 are 100 and 500 bp DNA molecular weight standards, respectively. This analysis shows that exposure to kendomycin results in DNA laddering.

Figure 4 shows DIGE images of U-937 cells treated with kendomycin. Image A shows an overlay of 3 individual images (B, C, and D). 50 μg of protein extracted from control cells, treated cells and an internal standard, were individually stained with different Cy Dyes (Cy2, Cy3 and Cy5). The three samples were then pooled and separated on a pH 4- 7-gradient IPG-stripe, and subsequently on a 12.5% homogenous polyacrylamide gel.

Figure 5 shows the effect of kendomycin on rabbit reticulocyte proteasome. (A) Proteasome activity was assayed by detecting the release of the fluorophore AMC (7- amino-4-methylcoumarin) from substrate Suc-Leu-Leu-Val-Try-AMC in the presence of methanol O ), 2 μM kendomycin (O ), and 2 μM known proteasome inhibitor MG- 132 ( Λ (Goldbaum, O. et al. Glia, 2006, 53, 891 -901 ). Released AMC was quantified using a Victor 1420 multilabel counter (excitation 380 nm, emission 460 nm). (B) Percentage of proteasome activity following kendomycin treatment.

Figure 6 shows light microscopic investigation of kendomycin-treated PtK₂ cells. Light microscopic images of PtK₂ cells stained with Gimsa show vacuolization after treatment with 2 μM kendomycin (B), in comparison to control cells (A) (scale bar = 20 μra).

Figure 7 shows electron microscopy of PtK₂ cells following kendomycin treatment. PtK₂ cells were imaged after treatment with 2 μM kendomycin for 0 (A and F), 4 (B and C) and 8 h (D, E, G and H). The cytoplasm vacuolization appears in the periphery of the nucleus (B) and spreads throughout the cytoplasm (C). The size of the vacuoles increased with longer incubation times (D and E). The arrows indicate the Golgi apparatus, (scale bar = 1 μm for A, B and C; 2 μm for D and E). The mitochondria (M) in treated PtK₂ cells were swollen (G) in comparison to those of untreated controls (F). The nucleus shows a condensed morphology (H), which indicates the induction of apoptosis. The arrows in F-H show the mitochondria. Figure 8 shows the changes to ER and mitochondrial morphology in PtK₂ cells upon treatment with kendomycin. The ER of PtK₂ cells was visualized with anti-GRP94 antibodies and Alexa Fluor 488, while the nuclei were stained with DAPI. A fine network of ER is evident in control cells; it is concentrated around the nucleus and extends throughout the cytoplasm (4A). In contrast, PtK₂ cells treated with 2 μM kendomycin for 6 h show a much lower ER density in the vicinity of the nucleus, and contain vacuoles. The vacuoles extend from the perinuclear region to the plasma membrane (4B) (scale bar = 20 μm). Mitochondria in PtK₂ cells were stained with Mitotracker green FM and nuclei were stained with Hoechst 33258 (blue). The mitochondria of the control cells exhibit an elongated morphology (4C). Following 2 h of kendomycin treatment (4D), the mitochondria become swollen and ring shaped.

Figure 9 shows the structure of the polyketide kendomycin (Bode HB, Zeeck A: Structure and biosynthesis of kendomycin, a carbocyclic ansa-compound from Streptomyces. J Chem Soc Perk T 1 2000,323-328);

Figure 10 shows the organization of cosmids and plasmids containing fragments from the kendomycin biosynthetic gene cluster; Figure 11 gives a detailed descπption on the plasmids containing fragments from the kendomycin PKS cosmids,

Figure 12 shows the nucleotide sequence of the kendomycin biosynthetic gene cluster from 5 violaceo) ubei 3844-33C, Figure 13 gives a detailed descπption of genes from the kendomycin biosynthetic gene cluster and the deduced function of the encoded proteins

Examples

Example 1 Discovery of mode of action via proteasome inhibition

Identifying specific targets of natural products in cells remains a challenging task, due to the multiplicity of effects which are often elicited (S L Schreiber, Chem Eng News 2003, 51-61 , S J Leuenroth, C M Crews, Chem Biol 2005, 12, 1259-1268) To addiess this issue, we adopted a multidisciplinary approach, including microscopy-based morphological studies, immunofluorescence, and fluorescence-activated cell sorting (FACS) Oui genome-wide proteomics analysis of the iesponse ol human leukemic monocyte lymphoma (U-937) cells to kendomycin was significantly enabled by the recently-developed softwaie GeneTrail [http //genetiail bioinf uni-sb de] Collectively, oui data demonstiate that U-937 cells undergo apoptosis in response to kendomycin exposuie, and identified the proteasome as a piobable primary target We further show that kendomycin inhibits the chymotiypsin-hke activity of rabbit reticulocyte-deπved 20S proteasomes in \ ιtι o Morphological studies of a second mammalian cell line PtK.₂ by light and election micioscopy, reveal changes characteiistic of proteasome inhibition We theretore propose that the cytotoxicity of kendomycin aπses, at least in part, from proteasome inhibition Thus, kendomycin not only has potential as an anti-tumor agent, but is also ot use in elucidating the many biological roles of the proteasome in mammalian cells Results

Identification of a model cell line for mode-of-action studies

We evaluated several mammalian cell lines as potential model systems for studying the mode of action of kendomycin. Kangaroo rat epithelial (PtK.₂) cells were selected in order to enable ready visualization by microscopy, while human cervix carcinoma (KB- 3-1 ) cells and human leukemic monocyte lymphoma (U-937) cells are standard cell lines for testing anti-tumor compounds in vitro. In addition, in our hands, cell lines KB-3-1 and U-937 reproducibly yield protein which is amenable to proteomics analysis. The cells were incubated with serial dilutions of kendomycin (313 pg mL '-55.5 μg mL ') in 96-well plates for 5 days, and then tested for growth inhibition using MTT (3-(4, 5- dimethylthiazolyl-2)-2, 5-diphenyltetrazolium bromide) (T Mosmann, J Immunol Methods 1983, 65, 55-63) This analysis showed that kendomycin inhibits the growth of both KB-3-1 and U-937 at low micromolar concentrations (Figure 1 ), with ICsυ values of 1.2 and 0 8 μM, respectively (Figure 1 ). We were not able to determine an ICso value for the PtK? cells due to growth characteristics, but they nonetheless showed clear signs of growth inhibition (data not shown). Based on these results, we selected the U-937 cells for further study, as they exhibited the higher susceptibility to kendomycin tieatment between the two cancer cell lines Kendomycin induces apoptosis in U-937 tumor cells

Initial light microscopy studies revealed several indicators for apoptosis in kendomycin- treated U-937 cells, including surface blebbing and cell shrinkage (data not shown). As depolaπzation of the mitochondrial membrane is one of the key events in apoptosis (S Matsuyama, J C Reed, Cell Death Differ 2000, 7, 1 155-1 165), we monitored for changes to the mitochondrial membrane potential (Δψ) using the cationic dye JC-I (S Salvioh, et al , FEBS Lett 1997, 411, 77-82). In normal cells, JC- I passively diffuses across the plasma membrane and forms aggregates within active mitochondria, which then emit red flourescence In apoptotic cells in which the mitochondrial membrane potential is low, JC- I is instead monomeπc, distributed throughout the cytoplasm, and exhibits green fluorescence Therefore, we could monitor for a collapse of Δψ in kendomycin-treated cells through an increase in green fluorescence relative to red Figure 2A shows the data obtained on control (untreated) U-937 cells by fluorometric detection on a FACS-Calibur flow cytometer, in which only a small percentage of cells exhibit any depolarization of mitochondrial membranes. Although U-937 cells incubated with kendomycin showed a similar profile after 20 min incubation, an increase in green fluorescence (Figure 2B) was observed at 40 min, and a further increase at 60 min (Figure 2C). Concomitantly, we detected a loss of red fluorescence (for example, after 2 h incubation (Figure 2D)). In parallel, by light scattering detection (F. Mentz et al., Cytometry 1998, 32, 95-101), we observed a significant increase in the number of apoptotic cells (Figure 2F), in comparison with the controls (Figure 2E). To further confirm apoptotic induction, U-937 cells were incubated with varying concentrations of kendomycin ( 1-3 μM) for 24 h, and stained with the DNA-binding dye propidium iodide (I. Nicoletti et al., J. Immunol. Methods 1991 , 139, 271-279) prior to analysis by flow cytometry. Treatment with kendomycin yielded a concentration- dependent increase in the number of apoptotic cells; at 2 μM kendomycin, approximately 50% of the cell population was sub-Gl (apoptotic), significantly higher than in the untreated control (Figure 3A). Cell cycle progression did not appear to be arrested at any particular phase in the remaining cells. Direct measurement of caspase 3- like activity revealed a 3-fold higher level in kendomycin-treated cells following 3 h incubation, and the activity increased with longer incubation times (Figure 3B). By Western blot analysis, we also detected activation of caspase 8, 40-60 minutes after exposure to kendomycin (Supplementary Information). DNA fragmentation was also evident after 4 hours (Figure 3C). As mitochondrial membrane depolarization, caspase activation and DNA laddering are characteristic features of apoptosis, these data together demonstrate that kendomycin induces programmed cell death in U-937 cells. Global proteoniic response of U-937 cells to kendomycin

We aimed next to try to identify candidate primary targets for kendomycin in U-937 cells by analyzing the global proteoniic response to toxin treatment. Proteins were therefore extracted from kendomycin-treated (2 μM) and control cells after 3.5 h, labeled with three different fluorescent Cy Dyes (Cy2, Cy3 and Cy5; Table 1 ), and analyzed by fluorescent 2D-difference gel electrophoresis (DIGE), as described in the Experimental Section. This experiment reproducibly identified more than 1300 spots corresponding to individual proteins. Some 70 proteins whose expression levels were altered by more than ±2 fold (P < 0.001), were excised from the gels. 30% showed a decrease in expression, while the remaining 70% were up-regulated (Figures 4B and 4C). After tryptic digestion of the excised proteins, the peptides were analyzed by matrix assisted laser desorption/ionization-time of flight (MALDI-TOF) mass spectrometry. The resulting peptide mass fingerprints were then used to screen the SwissProt database using MASCOT software [http://www.matrixscience.com/search intro. html]. Of the 70 proteins analyzed, 50 were present at sufficient levels conclusively to be identified (Table 2). Among these proteins, 72% were up-regulated and 28% down-regulated, and so the analyzed proteins are likely to be representative of the original sample. The proteins affected by kendomycin are involved in diverse cellular functions, including translation, intercellular transport, stress and defense, protein folding, cellular organization and the ubiquitin-proteasome system (Table 2). One of the most pronounced changes was an increase in the level of heat shock proteins (Hsps), including Hsp27, DnaJ (an Hsp40 homologue), Hsp70, Hsp90 and HsplO5. We also observed up- regulation of some cytoskeletal proteins, including actin and tubulin. The expression of many proteins which interact with RNA including the heterogeneous nuclear ribonucleoproteins (hnRNPs) hnRNP-K and hnRNP-C, as well as the synaptotagmin binding cytoplasmic RNA interacting protein (SYNKRIP), was also significantly altered. The hnRNPs play important roles in transcription, processing of pre-mRNA, alternative splicing and nucleocytoplasmic shuttling, (B. Carpenter et al., Biochim. et Biophys. Λcta- Reviews on Cancer 2006, 1765, 85-100; A. S. Ma et al., J. Biol. Chem. 2002, 277, 18010-18020; Y. He et al., MoI. Biol. Cell 2004, 75, 252A-253A) and are among the most abundant proteins in the eukaryotic nucleus. We also detected a significant up- regulation of eukaryotic initiation factor 5A (eIF-5A; 25-fold) in treated cells, suggesting that kendomycin may cause this protein to accumulate.

Changes to the expression of four proteins associated with the proteasome were also observed. Our proteomics analysis showed that in U-937 cells subjected to kendomycin, the 19S ATPase subunit 3 was up-regulated, as was subunit α5, while expression of PA28α and PA28γ was reduced. Given the many effects elicited by kendomycin in U-937 cells, we took advantage of a recently-developed web-based interface called GeneTrail [http://genetrail.bioinf.uni- sb.de] to aid in our identification of candidate primary targets. We performed a so-called "over-representation analysis" to determine whether sets of genes whose expression levels had changed significantly after kendomycin treatment (at least 2-fold up- or down- regulation) belonged to any specific functional biological category. Our computational analysis showed that genes belonging to the Gene Ontology (GO) category "response to protein stimulus' (p = 0.0001 1 ) were statistically over-represented among the up- regulated genes; this category contains Hsp family members Hsp70, Hsp90 and Hspl05. However, Hsp proteins can be up-regulated as a non-specific response to many cell stressors, including proteasome inhibition. Analysis of the down-regulated genes, highlighted the GO category 'proteasome activator complex' (p = 0.00024) which includes regulatory components PA28α and PA28γ, suggesting a direct interaction between kendomycin and the proteasome. We therefore we focused our attention on the ability of kendomycin to act as a proteasome inhibitor.

Kendomycin inhibits the chymotrypsin-like activity of the rabbit reticulocyte proteasome To evaluate directly the inhibitory effect of kendomycin, we used a commercially- available kit (Calbiochem) containing SDS-activated 20S proteasomes from rabbit reticulocytes. The chymotryptic activity of the proteasomes was measured by monitoring the release of free 7-amino-4-methylcoumarin (AMC) from the fluorogenic peptide Suc-Leu-Leu-Val-Tyr-AMC (L. J. Crawford et al., Cancer Res. 2006, 66, 6379- 6386). Incubation of the proteasomes with 2 μM kendomycin decreased the chymotryptic activity by approximately 75% (Figure 5). This level of inhibition was achieved after 40 min of incubation, and is comparable to that observed with the known proteasome inhibitor MG-132 (O. Goldbaum et al., GHa 2006, 53, 891-901 ) under the same conditions (Figure 5). However, inhibition with kendomycin did not increase above 75%, while that of MG-132 was essentially complete following further incubation. Nonetheless, our data strongly suggest that kendomycin is a proteasome inhibitor.

Effects of kendomycin on the morphology of PtK: cells

Proteasome inhibition in multiple cell lines has been reported to activate mitochondria- dependent cell death (O. Goldbaum et al., GHa 2006, 53, 891-901 ; N. Mitsiades et al.,

Proc. Natl. Acad. Sci. USA 2002, 99, 14374-14379; B. Wagenknecht et al., J.

Neiirochem. 2000, 75, 2288-2297; N. R. Jana et al., J. Biol. Chcm. 2004, 279, 1 1680- 1 1685), as well to cause distinctive morphological changes to the endoplasmic reticulum (B. Wagenknecht et al., J. Neurochem. 2000, 75, 2288-2297; E.G. Mimnaugh et al., MoI. Cancer Res. 2006, 4, 667-681 ; C. Wojcik, Folia Histochem. Cytobiol. 1997, 35, 21 1-214.). We therefore looked for further evidence of proteasome inhibition in PtK₂ cells by light microscopy, following exposure to 1-2 μM kendomycin. This technique revealed that the predominant effect on the PtK₂ cells was to induce the foπnation of cytoplasmic vacuoles (Figures 6A and 6B). After 3—4 hours, the vacuoles were small, perinuclear and heterogeneous in morphology. However, vacuolar size increased over the following 4 hours, consistent with fusion of the vacuoles into larger structures. Analysis by electron microscopy (EM) after 2—4 hours of treatment (Figure 7A), showed that the cytoplasmic vacuoles appeared to be membrane-bound, and to contain electron semi-dense material, although the translucency increased with longer incubation times. A subset of organelles, including the rough endoplasmic reticulum and the Golgi apparatus, remained intact after 4 hours of exposure. However, following 8-10 hours of incubation, the cytoplasmic space was entirely occupied by large, translucent vacuoles, and most of the organelles (with the exception of the mitochondria) had disappeared. This pattern of vacuole formation and growth closely mirror that observed with MCF-7 human breast tumor cells exposed to the highly-specific proteasome inhibitor Bortezomib (PS-341 ) (E. G. Mimnaugh et al., MoI. Cancer. Ther. 2004, 3, 551-566). To demonstrate that the vacuoles were derived from the ER (E. G. Mimnaugh et al., MoI. Cancer Res. 2006, 4, 667-681 ; C. Wojcik, Folia Histochem. Cytobiol. 1997, 35, 21 1- 214; E.G. Mimnaugh et al., MoI. Cancer. Ther. 2004, 3, 551 -566) we carried out immunofluorescence labeling of PtKi cells with an antibody against GRP-94 (glucose- regulated protein 94), a specific ER marker protein. The fluorescence pattern of control cells showed a typical ER network concentrated around the nucleus, but also located throughout the cytoplasm (Figure 8A). In contrast, the ER of kendomycin-treated cells (4 h), showed reduced nuclear localization, and the presence of small vacuoles. Vacuolization increased with incubation time, and then expanded throughout the cytoplasm (Figure SB). To further probe effects on the mitochondria, cells were exposed to kendomycin for varying periods (see Figure 8), and then incubated with the mitochondria-specific green fluorescent stain MitoTracker-FM (Molecular Probes). In control cells, mitochondria had a typically elongated form, and were distributed evenly throughout the cytosol (Figure 8C). However, in kendomycin-treated cells, the organelles were swollen and ring-shaped (Figure 8D). Such swelling characteristically occurs following depolarization of the mitochondrial membrane during apoptosis (V. Gogvadze et al., Biochem. J. 2004, 378, 213-217). Thus, our microscopy studies revealed that kendomycin produces effects on both the ER and mitochondria in PtK.₂ cells which are fully consistent with proteasome inhibition.

Discussion Elucidating the precise mode of action of natural products can greatly facilitate efforts to fine-tune their specificities and/or phaπnacological properties, and to exploit their use in chemical genetics (S. L. Schreiber, Chem. Eng. News 2003, 51-61 ). Here we have taken a multi-disciplinary approach to identifying a primary target of the polyketide kendomycin in the model mammalian cell line, U-937. U-937 lymphoma cells are strongly susceptible to kendomycin treatment with an ICso of 0.8 μM. The cytotoxicity of kendomycin towards U-937 cells results, at least in part, from the induction of apoptotic pathways, as evidenced by a decrease in mitochondrial transmembrane potential, activation of caspases 3 and 8, and DNA fragmentation, within a few hours of treatment. We could further show by FACS analysis that exposure to kendomycin yielded a substantial population of sub-Gl cells relative to untreated controls. These findings prompted us to carry out a global proteomics analysis of kendomycin-treated U- 937 cells in order to determine possible triggers for apoptosis.

Our proteomics data revealed that kendomycin caused substantial changes in the expression levels of many cellular proteins (both up- and down-regulation), including those involved in the stress response, protein folding, cytoskeletal organization, intracellular signal transduction, nucleotide metabolism and protein degradation (Table 2). Analysis of such multivariate high-throughput data is notoriously laborious, but we were able to take advantage of the new gene analysis package GeneTrail to efficiently identify the proteasome as one likely primary target of kendomycin. Indeed, we could show by direct in vitro analysis, that kendomycin inhibits the chymotryptic activity of rabbit reticulocyte proteasomes by 75%, a level comparable to the known proteasome inhibitor MG-132. Studies to determine the specificity and mechanism of inhibition are currently underway in the laboratory. However, it is tempting to speculate that kendomycin forms a covalent adduct with the active site threonine of the proteasome chymotryptic site (M. Groll et al., Structure 2006, 14, 451-456), as the C20 position (Figure 1 ) of the metabolite is known to be particularly susceptible to nucleophilic attack (H. G. Bode, personal communication). Thus, C20 may be part of the pharmacophore.

The finding that kendomycin is a proteasome inhibitor explains many of the other data obtained in this example. The overall proteomic response of the U-937 cells closely resembles that of several other cell lines when exposed to proteasome inhibitors (N. Mitsiades et al., Proc. Natl. Acad. Sci. USA 2002, 99, 14374-14379; B. F. Jin et al., Oncogene 2003, 22, 4819^830; S. Meiners et al., J. Biol. Chem. 2003, 278, 21517— 21525). For example, the most pronounced change observed in multiple myeloma cells following treatment with Bortezomib, was an up-regulation of hsp transcripts, including members of the hsp40, hsp70 and hsp90 families (N. Mitsiades et al., Proc. Natl. Acad. Sci. USA 2002, 99, 14374-14379). In kendomycin-treated U-937 cells, expression of Hsp27, DnaJ (a Hsp40 homologue), Hsp70, Hsp90 and Hsp 105 was uniformly increased. In both cell types, the changes are consistent with the protective role of molecular chaperones in response to cytotoxic agents (N. Mitsiades et al., Proc. Natl. Acad. Sci. USA 2002, 99, 14374-14379; C. Jolly and R. I. Morimoto, J. Natl. Cancer Inst. 2000, 92, 1564-1572; C. Garrido et al., Cancer Res. 1997, 57, 2661-2667; C. Paul, et al., MoI. Cell Biol. 2002, 22, 816-834). Exposure of vascular smooth muscle cells (VSMCs) from rat to proteasome inhibitors c-lactacystin and MG-132 resulted in a coordinated up-regulation of proteasome subunits at both the transcriptional and translational levels (S. Meiners et al., J. Biol. Client. 2003, 278, 21517-21525). Likewise, we also observed increased expression of both the 19S ATPase subunit 3, and the 20S subunit α.5. Our finding that expression of the PA28 proteins α and γ decreased is perhaps surprising, as inhibition of VSMC cells had almost no effect on these regulators (S. Meiners et al., J. Biol. Chem. 2003, 278, 21517-21525). Nevertheless, the substantial up-regulation of the proteasomal subunits likely reflects an attempt by the cells to compensate for the loss of proteasomal activity. We also observed increased expression of heterogeneous nuclear ribonucleoproteins, the cytoskeletal proteins actin and tubulin, as well as a significant accumulation of the eukaryotic initiation factor 5A. Although we cannot at present explain these findings, expression of the same proteins was also up-regulated upon exposure of the human megakaryoblastic leukemic cell line Mo7e to proteasome inhibitors MG-132 and lactacystin (B. F. Jin et al., Oncogene 2003, 22, 4819-4830). Finally, our analysis of PtK₂ cells by both light and electron microscopy, reveals morphological changes to both the ER and mitochondria which are also consistent with proteasome inhibition. Although the evidence for proteasome inhibition is compelling, kendomycin may have additional targets within U-937 cells (M. Groll, R. Huber, Biochim. et Biophys. Acta - Molecular Cell Research 2004, 1695, 33^14). Indeed, Western blot analysis (Supporting Information) shows that caspase 8 - an enzyme involved in death-receptor mediated apoptosis - is activated following kendomycin treatment. As one of its downstream effects, caspase 8 can initiate the mitochondrial apoptotic pathway though proteolytic cleavage of the effector protein Bid (BH3-interacting domain death agonist) into its activated form, tBid (truncated Bid) (S. Desagher and J. -C. Martinou, Trends Cell Biol. 2000, 10, 369-377). However, in U-937 cells, caspase 8 is activated 40-60 minutes after exposure to kendomycin, either simultaneously with or following depolarization of the mitochondrial membrane (Figure 2). Therefore, it appears that kendomycin may independently activate both caspase 8- and mitochondria-dependent apoptotic pathways in U-937 cells, as has been observed previously with Bortezomib (N. Mitsiades et al., Proc. Natl. Acad Sa. USA 2002. 99, 14374-14379). Nonetheless, we cannot at present rule out a different cellular trigger for caspase 8 activation. Naturally occurring and synthetic inhibitors of the proteasome characterized to date are largely peptidic in nature (aldehydes (e.g. MG-132), boronates (e.g. Bortezomib), vinyl sulfones (e.g. AdaAhx3L3VS), epoxy ketones (e.g. epoxomicin), cyclic (e.g. TMC- 95A)), but also include several β-lactones (e.g. lactacystin and salinosporamide), the triterpene celastrol and the polyphenol curcumin. In common with many of these inhibitors (L. J. Crawford et al., Cancer Res. 2006, 66, 6379-6386), kendomycin targets the chymotryptic-like activity of the proteasome, but it is, to our knowledge, the first macrocyclic polyketide to show inhibitory activity. Therefore, kendomycin not only represents an attractive lead structure for further development in cancer chemotherapy (C. Montagut et al., Clin. Transl. Oncol. 2006, 8, 313-317), but it may also serve as a useful tool for revealing further aspects of proteasome-mediated cell biology (M. Groll and R. Huber, Biochim. et Biophys. Acta - Molecular Cell Research 2004, 1695, 33-44). To summarise, we have used a number of approaches including microscopy, proteomics, and bioinfoπnatics, to investigate the mode of action of kendomycin in mammalian cell cultures. In response to kendomycin treatment, human U-937 tumor cells exhibit depolarization of the mitochondrial membrane, caspase-3 activation and DNA laddering, consistent with induction of the intrinsic apoptotic pathway. To elucidate possible apoptotic triggers, DIGE and MALDI-TOF were used to identify proteins that are differently regulated in U-937 cells relative to controls. Statistical analysis of the proteomics data by the new web-based application GeneTrail highlighted several significant changes in protein expression, most notably among proteasomal regulatory subunits. Overall, the profile of altered expression closely matches that observed with other tumor cell lines in response to proteasome inhibition. Direct assay in vitro further shows that kendomycin inhibits the chymotrypsin-like activity of the rabbit reticulocyte proteasome, with comparable efficacy to the established inhibitor MG- 132. Finally, microscopy and immunofluoresence studies reveal that kendomycin induces extensive vacuolization of the endoplasmic reticulum as well as mitochondrial swelling in a second cell line derived from kangaroo rat epithelial (PtK:) cells, phenotypes associated with inhibition of the proteasome. This study therefore provides evidence that kendomycin mediates its cytotoxic effects, at least in part, through proteasome inhibition.

Experimental Section

Cell cultures: Cell line PtKi was obtained from the American Type Culture Collection (ATCC-56), and cell lines U-937 and KB-3- 1 from the German Collection of Microorganisms and Cell Cultures (DMSZ). All cell lines were cultivated under conditions recommended by their respective depositors. Cell culture reagents were purchased from Life Technologies Inc (GIBCO BRL). Plastic ware was obtained from Nunc and Saarstedt.

Growth inhibition assay (MTT test): Growth inhibition was measured in 96-well plates. Aliquots ( 120 μl) of cells (5 x 10⁴ cells ml^{" 1} ) were incubated with 60 μl kendomycin at various concentrations (313 pg mL^"'-55.5 μg mL^{" 1} ). After 5 days, cell viability was measured using the MTT assay (T. Mosmann, J. Immunol. Methods 1983, (55, 55-63). Determination of mitochondrial membrane potential (Δψ): The mitochondrial membrane potential were measured by the method described by Cossariza (S. Salvioli et al., FEBS Lett. 1997, 411, 77-82; A. Cossarizza et al., Biochem. Biophys. Res. Commun. 1993, 197, 40^15). 0.5 x 10⁶ cells were incubated with JC-I ( 10 μg) at room temperature for 10 min. Cells were then washed with cold phosphate-buffered saline (PBS), resuspended in PBS (500 μl) and analyzed with a FacScan flow cytometer (Becton Dickinson) equipped with a single 488-nm argon laser.

Flow cytometric analysis: Cells were treated with a range of kendomycin concentrations ( 1-3 μM) for 24 h and then harvested by centrifugation. The cells were then fixed in 80% methanol at -20 ⁰C for 30 min, washed with PBS and then treated with 0.1% (w/v) saponin in PBS. Finally, the cells were treated with RNAse and the nuclei were stained with propidium iodide (20 μg ml^{" 1}) for 30 min at 37 ⁰C. The DNA content was measured using a FACS-Calibur instrument (Becton Dickinson); 30,000 events were collected for each experiment. Data were analyzed using CellQuest software (Becton Dickinson).

Caspase-3 assay: Caspase-3 activity was measured using an Apo-One Homogeneous Caspase 3/7 Assay Kit (Promega). Briefly, cells were seeded into a 384 well plate (2500 cells/well) and incubated with kendomycin (1 μm) for various time periods. The fluorescence was measured with a Fusion Fluorescence Reader (Perkin Elmer; excitation 485 nni, emission 535 run).

SDS-PAGE and detection by Western blot of caspase 8 activation: U-937 cells were treated with 1 -2 μM kendomycin for 0, 10, 20, 30, 40 and 60 min. The cells were immediately lysed by sonication in buffer (30 mM Tris-HCl (pH 7.5), 150 mM NaCl, 10% glycerol, 1% triton-X and protease inhibitors). The concentration of protein in each sample was then determined by Bradford (S. T. Smiley et al., Cell Biol. 1978, 88, 3671- 3675), and equal amounts of protein separated by SDS-PAGE. The proteins were then transferred onto immune-Blot PVDF-membrane (Bio-Rad) in buffer (25 mM Tris, 192 mM glycine, 20% methanol, 0.05 % SDS) in a semidry system (Bio-Rad) for 30 minutes (15 V) at room temperature. Residual binding sites on the membrane were blocked by incubating the membrane in Tris-buffered saline (TBS; 20 mM Tris-HCl (pH 7.4), 0.15 M NaCl), containing 5% nonfat dry milk and 0.1 % Tween-20, at room temperature for 2 hours. Membranes were incubated with mouse monoclonal anti-caspase-8 antibodies (12F5; Alexis Biochemicals; I - I OOO) at 4°C for 16 hours. The blots were washed 3 times in TBS, and incubated for 1 hour with anti-mouse horseradish peroxidase (HRP)- conjugated antibodies (Sigma) (diluted 1 : 10000 in blocking buffer). After washing with TBS, the protein bands were analyzed by enhanced chemiluminescence using the Supersignal West Pico Chemiluminescent Substrate System (Pierce).

Detection of DNA laddering: Genomic DNA of treated and untreated cells was isolated using a PURGENE DNA purification system (Gentra, USA). The DNA was then analyzed by electrophoresis (1% agarose, 0.7% hydroxyethylcellulose (HEC) in I x TBE buffer at 100 V). Protein sample preparation for labeling with Cy Dyes: Treated and untreated cells were collected by centπfugation After washing the cell pellets with PBS buffer, cells were lysed in buffer (7 M urea, 2 M thiourea, 4% CHAPS, 1 % SDS, 20 mg ml^{" 1} DTT) To reduce protein degradation, all steps were performed at 4⁰C After sonication, cell extracts were centπfuged for 30 mm at 14000 rpm and 4 ⁰C The supernatant was removed and proteins were precipitated overnight with a mixture of acetone and methanol ( 1 .9) at -20 ⁰C. The protein pellets obtained after precipitation were washed with acetone and dissolved in labeling buffer (7 M urea, 2 M thiourea, 4% CHAPS, 30 ITiM Tπs (pH S 5)) Protein concentration was determined by the method of Bradford (S T Smiley et al , CW/ _9ιo/ 1978, 88, 3671-3675). Cy Dye labeling: Four independent samples of both treated and untreated cells were analyzed along with a pooled standard which contained equal amounts of proteins from all of the samples (Table 1 ) A total of 50 μg protein in DIGE labeling buffer (7 M urea, 2 M thiourea, 4% CHAPS, 30 mM Tπs (pH 8 5)) was minimally labeled with one of three Cy Dye DIGE Fluors. Cy2, Cy3 and Cy5 (Amersham Biosciences) (Table 1 ). The Cy Dyes were reconstituted in anhydrous DMF and combined with protein samples at a ratio of 400 pmol of Cy Dye to 50 μg protein. Labeling was performed on ice in the dark for 30 mm. The reactions were then quenched by incubating with lysine on ice in the dark for 10 mm Three labeled protein samples (treated, untreated and the pooled standard) were combined and loaded into the IPG-stπpes For the purpose of protein identification, 600 μg of unlabelled pooled standard samples were separately processed by 2D gel electrophoresis Alternatively, 300 μg of unlabelled pooled standard were spiked into each gel 2D-eIectrophoresis, gel imaging and data analysis: Proteins were loaded into 18 cm, 4-7 immobilized pH gradient (IPG) stripes by passive re-hydration for approximately 14 hours. The IPGphor focusing apparatus (Amersham Biosciences) was used for separation, with the following program: 500 V gradient for 2 kVHs (kilovolt hours), 500 V step for 0.5 KVHs, 1000 V gradient for 2 KVHs, 1000 V step for 1 KVHs, 8000 V gradient for 10 KVHs, and 8000 V step for 30 KVHs. Following the isoelectric focusing, the IPG stripes were equilibrated for 15 min in equilibration buffer (2% SDS, 50 mM Tris-HCl (pH 8.8), 6 M urea, 3% glycerol, 0.002% bromophenol blue, 10 mg ml^{" 1} DTT) for 15 min. The IPG stripes were equilibrated again for 15 min in the same buffer except that the DTT was replaced with iodacetamide (25 mg ml^{' 1}). Equilibrated IPGs were then transferred onto 12.5% homogenous SDS-polyacrylamide gels for 5 h. Gels were run using an Ettan ALT 12 at 2W/gel for 45 min and then for 17 W/gel. After SDS-PAGE, the gels were scanned at 100 μm resolution, using a Typhoon 9410 imager (Amersham). The excitation/emission wavelengths for Cy2, Cy3 and Cy5 are 488/520, 532/580 and 633/670 nm, respectively. The 2D-gel containing 600 μg of unlabelled pooled standard sample was stained with colloidal Coomassie blue and scanned with the same imager using an excitation wavelength of 633 nm. Relative protein quantification across all samples was performed using DeCyder Differential in Gel Analysis (DIA) and Biological Variance Analysis (BVA). Student^'s /-Test and one-way analysis of variance (ANOVA) were used to calculate significant differences in relative abundances of protein spot-features in treated cells compared with control cells.

Protein spot picking and protein identification: Protein spots exhibiting more than a 2-fold change in expression level (P < 0.001 ) were excised directly from the gels either using an automated Ettan spot picker (Amersham) or manually. After trypsin in-gel digestion, the digested peptides were mixed ( 1 : 1 ) with matrix solution (5 mg ml^{" 1} α- cyano-4-hydroxy-cinammic acid in 50% acetonitile and 0.1% TFA) and spotted on stainless steel MALDI sample plates. Analysis was performed using an Applied Biosystems Instrument 4800 MALDI TOF/TOF Analyzer™. Database searches were performed using Mascot [http://www.matrixscience.com/search intro. html]. Identifications were based on protein spot features if the protein score was calculated to be greater than 70, correlating to a confidence interval of 99%.

Analysis by GeneTrail: GeneTrail is a web-based application which detects functional categories that are enriched in a set of genes compared to a reference set. The significance values (p-values) of the over-representation analysis (ORA) are computed by the Hypergeometric distribution. Although GeneTrail supports many biological categories including many biochemical pathways, we focused our analysis on Gene Ontology (GO). We compared the set of up- and down-regulated genes after kendomycin treatment separately to the set of all human genes using GeneTrail's standard parameters. Since many GO categories were tested for statistical significance, we adjusted the p-values for multiple testing. GeneTrail is freely available at genetrail.bioinf.uni-sb.de.

Proteasome activity assay: The chymotrypsin-like activity of the isolated rabbit reticulocyte proteasome was measured in microplates using the SDS-activated proteasome kit (Calbiochem) according to the manufacturer's instructions. Briefly, proteasome (end concentration 1.2 μg mL ¹) was incubated in reaction buffer ( 100 mM

HEPES (pH 7.6), 0.5 mM EDTA, 0.03% SDS) for 10 min at 37 ⁰C. The methanol, kendomycin and MG- 132 were added and the mixtures were further incubated for 15 min at 37 ⁰C. The reaction was initiated by adding the fluorogenic substrate Suc-Leu-

Leu-Val-Try-AMC, and the rate of AMC release was measured using a Victor 1420 multilabel counter (excitation 380 nm, emission 460 nm).

Staining of the ER in PtK₂ cells: PtK.₂ cells were grown on glass cover slips (13 mm diameter) in four-well-plates. Exponentially growing cells were incubated with kendomycin ( 1 -2 μM) for various time periods (Figure 8). Cells were then fixed in freshly prepared 3.75% formaldehyde for 15 min at room temperature. The fixed cells were washed twice with PBS and treated with anti-GRP antibodies (1 :500 dilution; Transduction Lab) for 30 min at 37 ⁰C. The cells were then washed with PBS and treated with anti-mouse-Alexa Flour 488 (1 : 1000 dilution; Molecular Probes). The nuclei were stained by incubating with DAPI (4'-6-diamidino-2-phenylindole) solution ( 1 μg mL^"' ) in PBS for 3 min.

Staining of mitochondria in PtK₂ cells: PtK: cells were grown on glass cover slips ( 13 mm diameter) in four-well-plates. Exponentially growing cells were incubated with the inhibitor for 2-6 hours and stained for mitochondria with 75 nM MitoTracker Green FM (Molecular Probes) at 37 ⁰C for 30 min. The nuclei were stained using Hoechst 33258 (5 μg ml^{" 1} ). The cover slips were mounted upside down in PBS, fixed with nail polish, and observed under a fluorescent microscope. Transmission electron microscopy: PtK₂ cells were seeded onto 12-mm cover glasses in 4-well plates. After appropriate incubation times, treated and untreated cells were fixed with 3.7% formaldehyde in PBS at room temperature for 15 min.. The cells were washed twice with cacodylate buffer (100 mM sodium cacodylate (pH 6.9), 90 mM sucrose, 10 mM MgCb, 10 mM CaCl₂) and post-fixed in 1% osmium oxide in 0.1 M cacodylate buffer. After an additional washing step with cacodylate buffer, samples were embedded into 1.7% agar, cubed, dehydrated in an ascending acetone concentration series (from 10 to 100%) and finally embedded in Spurr resin. Samples were then transferred into gelatine capsules and allowed to polymerize for 8 hours at 70 ⁰C. After polymerization, specimens were cut as ultra thin slices using a diamond knife. To increase the contrast, samples were post-stained with 4% uranyl acetate for 5 min, and examined with a Zeiss TEM910 transmission electron microscope at an acceleration voltage of 8O kV.

Supplementary Information. Western blot analysis of caspase 8 activation. U-937 cells were treated with 1-2 μM kendomycin for 0, 10, 20, 30, 40 or 60 min, and then the cells were lysed. Equal amounts of protein from each sample were then subjected to

Western blotting using monoclonal anti caspase 8 antibodies, and the protein bands visualized by enhanced chemiluminescence using the Supersignal West Pico

Chemiluminescent Substrate System. As evidenced by the decrease in intensity of the band corresponding to pro-caspase 8, and the appearance of a band corresponding to an intermediate fragment in the cleavage pathway, activation of caspase 8 occurs after 40-

60 minutes of kendomycin treatment.

Example 2; Isolation and sequence analysis of the kendomycin biosynthetic gene cluster Genomic DNA was isolated from S. violaceoruber strain 3844-33C (Jendrossek, D. et al. Bacterial degradation of natural rubber: a privelege of actinomycetes? FEMS Microbiol. Lett. 1997, 150, 179-188) using an established protocol (Keiser T. et al. Practical Streptomyces Genetics, Norwich, England, The John Innes Foundation, 2000).

The DNA was partially digested with Saιι3A using a serial dilution method and ligated in BamHl hydrolysed SuperCos- 1 (Stratagene, La Jolla). Ligations were packaged using the Gigapack III Gold Packaging Extract (Stratagene, La Jolla) and transfected into

Escherichia coli SURE (Stratagene, La Jolla) according to the manufacturer's protocol. 2300 colonies were picked in 96 well microtiterplates and cultivated in LB medium amended with 50 μg/ml ampicillin at 37⁰C over night. Duplicates of the clones were transferred onto nylon membranes (Roche Molecular Biochemicals) as described in the DIG systems user guide for filter hybridization. For the screening of the cosmid library with homologous probes, degenerate oligonucleotides (EAKSC 5 -MGIGARGCIYTIGCIATGGA YCCICARCARMG-3 ^' and LCK5NC 5 -GGRTCNCCIARYTGIGTICCIGTICCRTGIGC-3 ) and Taq DNA polymerase (GIBCO BRL, Eggenstein) were used to amplify approx. 750 bp ketosynthase (KS) fragments from genomic DNA of 5. violaceoruber strain 3844-33C. PCR conditions were as follows: initial denaturation 5 min at 97 ⁰C; denaturation 30 s at 97 ⁰C; annealing 30 s at 55 ⁰C; extension 50 s at 72 ⁰C; 30 cycles; addition of 5 % dimethylsulphoxide to the PCR mixture. The generated product was labelled with 50 μCi α-33PdCTP (Hartmann, Braunschweig) and used for hybridization experiments in a buffer containing 50% formamide at 42 ⁰C with stringent washes at 68 ⁰C. For the detection, Fuji imaging plates and a phosphoimager were used. From approx. 2300 cosmid clones, 96 gave strong signals. Cosmid DNA from all 96 clones was isolated and electroporated into E. coli XL-I Blue MRF (Jerpseth B et al., XLl-Blue MRF^' E. coli cells: MrcA-, MrcB-, Mrr-, HsdR-derivative of XLl -Blue cells. Strategies 1992, 5:81 -83) according to standard protocols (Sambrook J, Russell DW: Molecular cloning: A laboratory manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press; 2001 ). After reisolation from E. coli XL-I Blue MRF^", the cosmid DNA was hydrolyzed with EcoRl/BamUl.

Southern analysis was performed using a heterologous probe based on the dpgA, dpgB, dpgC and dpgD genes from the balhimycin pathway of Amycolatopsis mediterranei (Pfeifer V. et al., A polyketide synthase in glycopeptide biosynthesis - The biosynthesis of the non-proteinogenic amino acid (S)-3,5-dihydroxyphenylglycine. J.Biol. Chem. 2001 , 276:38370-38377). The probe was generated by EcoRl/Clal restriction of pVP5 (Pfeifer V. et al., A polyketide synthase in glycopeptide biosynthesis - The biosynthesis of the non-proteinogenic amino acid (S)-3,5-dihydroxyphenylglycine. J. Biol. Chem. 2001 , 276:38370-38377), gel purification of the 0.96 kb and 2.95 kb fragments using the Nucleospin Extract Kit (Macherey-Nagel, Duren) and labelling of the fragments with 50 μCi α-33PdCTP (Hartmann, Braunschweig). From the 96 KS cosmids, 5 gave strong signals after hybridization with the dpgA-dpgB- dpgC-dpgD probe (cosmid A7, C7, E 12, F3 and H4). DNA from these 5 cosmids was isolated using the Nucleospin Plasmid Kit (Macherey-Nagel, Dϋren), submitted for end sequencing and the data was analyzed by BLAST (http://www.ncbi.nlm.nih.gov/BLAST/). Cosmid H4, which was found to encode a type I PKS at the T3 end and a protein homologous to DpgC (Pfeifer V. et al., A polyketide synthase in glycopeptide biosynthesis - The biosynthesis of the non-proteinogenic amino acid (S)-3,5-dihydroxyphenylglycine. J.Biol. Chem. 2001 , 276:38370-38377) at the T7 end, was completely sequenced (GATC Biotech, Konstanz). Detailed sequence analysis indicated that the 36.1 kb insert from cosmid H4 does not harbour the complete kendomycin biosynthetic gene cluster. Cosmids harbouring fragments, which overlap with the T7 end of cosmid H4, have already been identified in the hybridization experiment using the dpgA-dpgB-dpgC-dpgD probe (cosmid F3 and A7, see Figure 10).

To detect cosmids overlapping with the T3 end of cosmid H4, the sublibrary containing the 96 PKS cosmids (see above) was hybridized with a probe homologous to the T3 end of cosmid H4. The probe was generated by PCR using the oligonucleotides H4,T3.3 (5 -

CTGTTCGCGCCAGGTCAC _{V) and H4T34 (5}_ _{GTCCGCGACGGCTGGTAC}_3') and HotStarTaq DNA polymerase (Qiagen, Hilden); PCR conditions were as follows: initial denaturation and activation of the DNA polymerase 15 min at 95°C; denaturation 20 s at 94 ⁰C; annealing 30 s at 57 ¹¹C; extension 40 s at 72 ⁰C; 32 cycles. The 456 bp PCR product was gel purified using the Nucleospin Extract Kit (Macherey-Nagel, Duren). For the hybridization experiment the DIG High Prime DNA labelling and detection Kit (Boehringer, Mannheim) was used. Strong signals could be detected for the cosmids B9, Dl 1 , F3, Fl O and G2. DNA from these cosmids was isolated using the Nucleospin Plasmid Kit (Macherey-Nagel, Duren) and submitted for end sequencing.

Based on detailed sequence and restriction analyses, cosmids (A7, F3, H4, Dl 1 and FlO) were mapped (see Figure 10). As cosmid H4 (which was completely sequenced) does not harbour the complete kendomycin biosynthetic gene cluster, fragments upstream and downstream to the cosmid H4 insert were subcloned for sequencing according to standard protocols (Sambrook J, Russell DW: Molecular cloning: A laboraton* manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press; 2001 ) (pKen l4, pKenl 5, pKen22, pKen24, pKen26; see Figure 10 and Figure 1 1 ). Sequencing was performed after in vitro transposon mutagenesis using GPS-I (NEB, Beverly) according to the manufacturer^'s protocol and/or by primer walking.

The complete nucleotide sequence of the kendomycin biosynthetic gene cluster was assembled using the Lasergene software package (DNASTAR Inc.) and is shown in Figure 12. The sequence shown in Figure 12 is also referred to herein as SEQ ID NO: 1. Sequence analysis was performed using FramePlot 2.3.2 (http://www.nih.go.jp/~jun/cgi- bin/frameplot.pl), BLAST (http://www.ncbi.nlm.nih.gov/BLAST/), Pfam 5.5 (http://www.mrc-lmb.cam. ac.uk/genomes/madanm/pres/pfaml. htm) and the PKS/NRPS Analysis Web-server (http://www, tigr.org/iravel/nrps/). 20 genes (ken 1-20) transcribed in two putative operons (kenl-11 and ken 12-20) and a gene transcribed in the opposite direction of ken 12-20 operon (ken21) could be identified (see Figure 10). For a detailed description of the identified genes and the deduced function of the encoded proteins see Figure 13.

Example 3: Heterologous expression of genes involved in the biosynthesis of the kendomycin PKS starter unit

A part of the kenl-11 operon (encoding genes involved in the biosynthesis of the kendomycin starter unit) was subcloned into the E. colilStreptomyces shuttle vector pSET152 (Bierman M. et al., Plasmid cloning vectors for the conjugal transfer of DNA from Escherichia coli to Streptomyces spp. Gene 1992, 1 16:43-49) featuring an oriT RK.2 for conjugation, the apramycin resistance gene aac(3)IV for selection and the ΦC31 derived attachement site (attP ΦC31 ) and integrase (/H/ ΦC31 ) for site-specific integration into the Streptomyces chromosome.

The resulting plasmid pKenl S harbouring the genes ken 1-8 (see FIG. 2 and FIG. 3) was transferred into Streptomyces coelicolor by conjugation using an established protocol

(Keiser T. et al. Practical Streptomyces Genetics, Norwich, England, The John Innes

Foundation, 2000). The S. coelicolor :pKenl 8 mutants were selected and cultivated on

MS medium (Keiser T. et al. Practical Streptomyces Genetics, Norwich, England, The

John Innes Foundation, 2000) amended with apramycin 60 μg/ml. After cultivation in baffled flasks for 3-4 days on a rotary shaker ( 180 rpm, 30 "C), the mycelium was separated by centrifugation and extracted with acetone. The extract was evaporated, redissolved in methanol and analyzed for the production of 3,5-dihydroxy benzoic acid (an intermediate in the biosynthesis of the kendomycin PKS starter unit (Bode HB. Chemical and Biosynthetic Investigations of Selected Polyketides, 1 17. 2000. Gottingen, Georg-August-Universitat, Mathematisch-Naturwissenschaftliche Fakultat. Ref Type: Thesis/Dissertation). High-pressure liquid chromatography-mass spectrometry (HPLC-MS) was used to analyze the mutant extracts in comparison to an extract from S.coelicolor wildtype grown under the same conditions. An Agilent 1 100 series solvent delivery system coupled to Bruker HCTplus ion trap mass spectrometer was used. Chromatographic separation was earned out on an RP column Nucleodur Cl 8 (125 by 2 mm, 3 μm particle size; Macherey and Nagel) equipped with a precolumn Cl 8 (8 x 3 mm, 5 μm). The mobile-phase gradient (solvent A: water + 0.1% formic acid and solvent B: acetonitrile + 0.1 % formic acid) was linear from 5 % B at 2 min to 10 % B at 5 min and from 10 % B at 5 min to 95 % B at 9 min, followed by 1 min with 95 % B at flow rate of 0.4 ml/min. Diode array detection was carried out at 254 nm and mass detection in negative ionization mode. 3,5-Dihydroxy benzoic acid was identified by comparison to the retention time and the MS' pattern of the authentic reference standard (m/z [M-H]^" = 153; MS²: m/z [M-H-COT]^" = 109). Quantitation was canned out in manual MS^" mode. Ions of m/z [M-H]^" = 153 were collected and subjected to fragmentation. Peak integration of the characteristic fragment ions m/z 109 was earned out utilizing the Bruker Quant-Analysis vl .6 software package. A calibration curve was established from serial dilutions of 3,5-dihydroxy benzoic acid down to 1 μg/ml. The production of the kendomycin biosynthesis intermediate 3,5-dihydroxy benzoic acid in the S. coelicolor: :pKenl 8 mutants averages 30 μg/1 culture. No 3,5-dihydroxy benzoic acid could be identified in extracts of S. coelicolor wildtype.

Example 4: Assays for test compounds with anti-cancer activity

Cytotoxic assays, such as those described in Example 1 , will be performed in the nanomolar range and against a large panel of cancer cell lines. In the next stage, the biosynthetically and synthetically accessed analogues will be subjected to the same cytotoxicity assays. The invention has been described above by way of example only. It will be appreciated that modifications in details may be made to the invention whilst still falling within the scope of the claims.

Claims

Claims:

1. A method of inhibiting the proteasome comprising contacting the proteasome with a polyketide 2. Use of a polyketide as a proteasome inhibitor.

3. A polyketide for use in treating a disease or disorder in which the proteasome is involved

4 Use of a polyketide in the manufacture of a medicament for the treating a disease or disorder in which the proteasome is involved. 5 A method of treating a patient having a disease or disorder in which the proteasome is involved, wherein said treating comprises inhibiting the proteasome using a polyketide.

6 A polyketide, use or method according to any one of claims 3 to 5, wherein the disease or disorder in which the proteasome is involved is selected from the list consisting of diseases which involve angiogenesis (for example metastatic cancer, diabetic retinopathy and rheumatoid arthritis), cancer such as myeloid myeloma, prostate cancer, pancreatic cancer, breast cancer, lung cancer, ovaπan cancer, solid turnouts and lymphomas such as Non-Hodgkin's lymphoma, mantle cell lymphoma and follicular lymphoma; retroviral diseases such as HlV; chronic inflammatory conditions such as asthma, ischemia and reperfusion injury, multiple sclerosis, rheumatoid arthritis; psoπasis; inflammatory and degenerative conditions such as Alzheimer's disease, amyotrophic lateral sclerosis, autoimmune thyroid disease, cachexia; Crohn's disease, hepatitis B; inflammatory bowel disease; sepsis, systemic lupus erythematosus; acute stroke, myocardial infections and transplantation rejection such as graft-versus-host disease.

7 A polyketide, use or method according to any one of claims 3 to 6, wherein the patient to be treated has been found to have abnormal proteasome activity.

5 A polyketide, use or method according to claim 7, where the disease to be treated is cancer

9. A polyketide, use or method according to any one of claims 3 to 8, wherein during treatment, the polyketide forms a complex with the proteasome chymotryptic site.

10. A polyketide, use or method according to any one of claims 3 to 9, wherein during treatment, caspase is activated. 1 1. Use of a polyketide as an adjuvant in a peptide-based vaccine.

12. The use of claim 1 1 , wherein the peptide is displayed on dendritic cells.

13. The use of claim 1 1 or claim 12, wherein the vaccine is an anti-cancer vaccine.

14. A polyketide, method or use according to any preceding claim, wherein the polyketide is a macrocyclic polyketide. 15. A polyketide, method or use according to claim 14, wherein the macrocyclic polyketide has at least one of an E-trisubstituted olefin, an aliphatic ansa chain, a tetrahydropyran ring and a quinone-methide chromophore.

16. A polyketide, method or use according to claim 14 or claim 15, wherein the macrocyclic polypeptide is according to Formula I: Formula I:

wherein: R is selected from the group consisting of alkyl, alkenyl, alkynyl, and X is a heteroatom

17 A polyketide, method or use according to claim 16, wherein the macrocyclic polypeptide is kendomycin 18 A method for screening for polyketides that have activity against a disease or disordei in which the proteasome is involved, wherein the method involves determining the ability of the analogue or deπvative to inhibit a proteasome

19 A method according to claim 18, which comprises the steps ot i) selecting as a test compound a polyketide, π) incubating the test compound with SDS-activated 2OS proteasomes from rabbit reticulocytes to produce a test combination, in) incubating a known proteasome inhibitor with SDS-activated 2OS proteasomes from rabbit ieticulocytes to produce a control combination, iv) measuring the release of free 7-amino-4-methylcoumaπn (AMC) from the fluorogenic peptide Suc-Leu-Leu-Val-Tyr-AMC in the test and control combinations, v) comparing the measurements made in step iv) between test and control combinations, wherein it the amount of release of free AMC is comparable between the test and control combinations, the test compound is a pioteasome inhibitor

20 A method according to claim I S or claim 19, wherein the polyketide is an analogue or derivative ot kendomycin

21 A method according to any one of claims 18 to 20, wherein the disease or disorder is as recited in claim 6 22 An analogue or deπvative of kendomycin obtained using a scieening method as described in any one of claims 18 to 21 tor use in inhibiting the proteasome

2^ A nucleic acid encoding kendomycin, wherein the nucleic acid comprises the sequence pro\ ided in SEQ ID NO 1

24. A variant of a nucleic acid according to claim 23, wherein the variant has a level of sequence identity of 80% or more to the sequence shown in SEQ ID NO: 1.

25. A nucleic acid comprising or consisting of a fragment of a nucleic acid according to claim 23 or claim 24, wherein the fragment is at least 200 nucleotides in length. 26. A method for the heterologous production of kendomycin by generating it in a recombinant system.

27. A method according to claim 26, comprising expressing one or more of the proteins encoded by a nucleic acid according to any one of claims 23 to 25 from said nucleic acid in the recombinant system. 28. A method according to claim 27, comprising expressing all of the proteins encoded by a nucleic acid according to claim 23 from said nucleic acid in the recombinant system.

29. A method according to any one of claims 26 to 28, comprising: i) generating in a first host cell, a single vector comprising the component genes of the biosynthetic pathway for kendomycin; ii) transforming a second host cell with the vector; iii) culturing the second host cell under conditions which are suitable for synthesis of kendomycin; and wherein the genes of the biosynthetic pathway are transcribed under the control of promoters that are functional in the second host cell.