Chitin Synthase, a Fungal Glycosyltransferase That Is A

Home , Chitin synthase

Glycochemistry Today CHIMIA 2011, 65, No. 1/2 49 doi:10.2533/chimia.2011.49 Chimia 65 (2011) 49–53 © Schweizerische Chemische Gesellschaft Chitin Synthase, a Fungal Glycosyltrans- ferase that Is a Valuable Antifungal Target

Jean-Bernard Behr*

Abstract: During the last 30 past years, more life-threatening fungal infections have appeared due to the increas- ing frequency of patients with weakened immune systems. Inhibition of fungal enzymes involved in the biosynthesis of sterols is considered to be a safe and effective option for antifungal therapy. However, the intensive use of sterol-biosynthesis inhibitors for years has resulted in resistance development. Consequently, the search for alternative therapeutics must be intensified. In this context, the biosynthesis of chitin, an essential component of the fungal cell wall that is absent in mammalian cells, was envisioned as a safe and selective therapeutic target. We present here recent successes in the inhibition of chitin synthase, the enzyme involved in the last step of the biosynthesis of chitin. Keywords: Antifungal · Glycosyltransferase · Iminosugars · Inhibition · Nikkomycin

Introduction UDP-GlcNAc as the obligate substrate. In Saccharomyces cerevisiae, three distinct (UDP-GlcNAc) chitin synthase activities (ScCSI, ScCSII Fungal cells are protected from the exter- 2n nal environment by the cell wall, an initial and ScCSIII) have been described, which barrier made mostly of carbohydrates.[1] differ in function, localization, cation spec- This cell wall is essential for the mainte- 2n UDP ificity, optimum pH and even sensitivity to nance of cell shape, prevention of lysis and inhibitors.[8] The chemical transformation regulation of biochemical exchange with HO NHAc catalyzed by CS can be seen as a repeti- [2] O HO tive transfer of GlcNac residues from the the environment. The composition of O O the cell walls is made up of three biopoly- HO O activated donor UDP-GlcNAc to the grow- NHAc [3] OH mers: glucans, mannoproteins and chitin. n ing chitin chain (Fig. 1), with concomitant Chitin is a linear polysaccharide com- chitin release of UDP. posed of 2-acetamido-2-deoxy-d-glucose Though all enzymes of the chitin bio- (GlcNAc) units joined through b-(1,4) gly- HO O synthetic pathway could serve as targets [4] HO O in antifungal drug design, most studies cosidic linkages (Fig. 1). It is indispens- HO NH able for fungal survival. Cells in which the AcN H concerned chitin synthase,[9] by analogy formation of chitin is disrupted are affected O O N O with what is found in Nature. Indeed, the O O by osmotic sensitivity, abnormal morphol- P P ‘natural’ control of chitin metabolism is ogy and growth arrest.[5] Since chitin is O O O O achieved by the antifungal antibiotics nik- HO OH absent from vertebrates, inhibition of its komycins and polyoxins that specifically UDP-GlcNAc : (KM=0.5 mM) biosynthesis has been considered as a safe inhibit chitin synthase with Ki values in the and selective option for the development of O low micromolar range.[10] The structural antifungal agents.[6] HO resemblance of these peptidyl nucleosides O NH CO2H with UDP-GlcNAc has been pointed out to N O explain their high potency to bind to the N N O H catalytic site of chitin synthase. Nikko- OH NH2 mycin Z (1, NiZ), one of the most potent HO OH [11] CS inhibitors (Ki = 0.78 mM), showed 1, Nikkomycin Z : good in vitro activity against a number of (IC50=0.50 M; Ki=00.78.34 M, pathogenic fungi (MIC = 12 mg/mL for competitive) Candida albicans, MIC = 0.78 mg/mL for Blastomyces dermatidis) and to date, this Fig. 1. Biosynthesis of chitin from UDP- compound is the only CS inhibitor that un- GlcNAc. derwent clinical trials.[12] However, despite excellent in vitro results, nikkomycins and polyoxins display restricted ability to stop *Correspondence: Dr. J.-B. Behr Formation of chitin is a highly complex fungal growth in vivo due to either inef- Université de Reims Champagne-Ardenne Institut de Chimie Moléculaire de Reims and interconnected series of biochemical ficient transport or degradation before CNRS UMR 6229 reactions that has been intensively studied reaching the target enzyme. The search UFR des Sciences Exactes et Naturelles in the yeast Saccharomyces cerevisiae.[7] for new CS inhibitors with enhanced an- BP 1039 - 51687 Reims Cedex 2, France. The last step of this process is governed tifungal activity follow three strategies: i) Tel.: +33 3 26913238 FIGURE 1, one-column Fax: +33 3 26913166 by an enzymatic activity called chitin identification of new inhibitors from natu- E-mail: [email protected] synthase (CS, EC 2.4.1.16), which uses ral sources by screening programs,[13] ii) 50 CHIMIA 2011, 65, No. 1/2 Glycochemistry Today

inhibition based on substrate analogues using 1 as the lead compound or iii) inhi- O O bition based on mechanism with the design NH NH O OH of transition-state mimetics or bisubstrate H O O N O N N O analogues. Recent results following the N O O O N two last options are discussed here. O O OH O HO OH HO OH

4: IC50=2.9 mM 5: IC50=2.0 mM (Ki=0.79 mM, competitive) The Development of Synthetic NiZ Analogues O R R R compound R11 R22 R33 IC50 (mM) The design of analogues that could NH R O function as inhibitors of chitin synthase 3 H H 0.80 R2O O 6a O N O N CH2 focused on structural features of either the R1O O natural substrate UDP-GlcNAc or the NiZ OH 6b H H 2.0 core and aimed to address the issues limit- HO OH N CH2 6a-c ing in vivo activity: hydrolytic degradation 6c H H 3.2 N CH2 and poor uptake. The most simple non- hydrolyzable analogue of UDP-GlcNAc was prepared by the Finney[14] and Kirk[15] Fig. 3. Inhibition of chitin synthase by hetaryl nucleosides 4–6. groups and features a methylene in place of the glycosidic oxygen (compound 2, Fig. 2). Phosphonate 2 was a very weak O inhibitor of chitin synthase (Ki >10 mM) having much lower affinity for the active NH site than UDP-GlcNAc itself (K = 0.5 H M O N O 59 carboxylic acids 15 isocyanides mM). Interestingly, the replacement of O + + R COOH R2NCCN the sugar residue in 2 by a hetaryl moiety 1 2 (compound 3) restored the inhibitory po- O O [16] 1°) Rink amide resin tency (IC50 = 0.82 mM). The quinoleine + group certainly assumes complementarity 7 2°) FIHGURE 3,two-column between the hydrophobicity of CS and its O inhibitors, as exemplified by NiZ. Using O H 3 as a template, we prepared analogues R N H 2 O NH O N N featuring other nonhydrolizable phosphate O NH mimics such as tartrate, malonate or mono- HN N O O HN N O saccharidic residues (Fig. 3).[17] Whereas R O 1 O quinoleine-malonate 4 (IC = 2.9 mM) or O 50 HO OH NH O tartrate 5 (IC = 2.0 mM, K = 0.79 mM, HO OH 50 i Ph competitive) display weaker activity than μ the corresponding phosphonate, the inhibi- 8 (885 compounds) 8a: IC50=6.1 M tion potencies of monosaccharidic derivatives 6 depend on the substitution pattern Scheme 1. Combinatorial synthesis of NiZ analogues 8.

(IC50 = 0.80–3.2 mM). The 1,3 substituted glucose nucleoside 6a was the most active of the series, which displayed Ki = 0.25 Surprisingly, the replacement of the quin- with strong hydrophobic character, close mM with a competitive inhibition pattern. oleine moiety in the structure of 4 or 5 by to the catalytic site. Thus, 23 novel NiZ a hydroxypyridine to resemble NiZ did not analogues 9 with aromatic groups at the improve the binding interactions.[17] b-position were prepared, some of them A combinatorialFIGU synthesisRE 4, tw ofo- colu a largemn displayed strong anti-CS activity (selected O HO series of peptidyl-nucleosides has been data in Scheme 2). O HO NH described for the construction of library A library of 1,4-disubstituted-1,2,3- HO AcN H of NiZ analogues. The Ugi condensation triazolyluridine derivatives was realized O O N O P P O of the uridinyl aldehyde 7 with a combi- recently using Cu(i)-catalyzed 1,3-dipolar O O O O nation of various carboxylic acids and cycloaddition of the azido-uridine 10 with HO OH isocyanides gave 885 bis-amides of type propargylated aromatics (Scheme 3).[20] 2:IC50>10 mM 8 (Scheme 1).[18] Some hits were identi- The design of these targets was mainly O fied, which were as active as NiZ against based on the replacement of the peptide NH CS from Candida albicans, compound 8a linker of NiZ with a triazole moiety. Chi- (IC = 6.1 mM) being the most promis- tin synthase activity was evaluated for Ben- O O N O 50 N P P O ing candidate for therapeutic application. jaminiella poitrasii cells. Surprisingly, the O O O O Based on a receptor site model calcula- isopropylidene-protected nucleosides 11 HO OH tion for NiZ, a study by Obi and cowork- displayed potent anti-CS activity (80–95% 3: IC50=0.82 mM ers aimed at designing a new series of NiZ inhibition at 4 mg/mL) and good antifungal analogues with simplified chemical struc- potency against human and plant patho- [19] Fig. 2. Inhibition of chitin synthase by substrate ture. Molecular modeling has stressed gens. In an analogous approach, Plant and analogues 2 and 3. out the existence of a binding pocket coworkers designed unprecedented NiZ

FIGURE 2, one-column Glycochemistry Today CHIMIA 2011, 65, No. 1/2 51

analogues bearing an isoxazole ring as the phosphate surrogate.[21] Unfortunately, O O their inhibitory activities toward CS have NH O NH not been reported yet. CO2H CO2H Ar H N N O N O 2 O Ar COOH i,ii N O + H NH2 NHBoc The Design of Mechanism-based HO OH HO OH Inhibitors 9a-d Like other polymerizing transfer- i: N-hydroxysuccinimide/DCC, 16-60%; ii: TFA, then Dowex XFS-43279, 19-98%. ases, chitin synthase is an integral mem- brane protein for which little information is available concerning its structure and compound Ar % inhibition at IC50 ( g/mL) 100 g/mL mode of action. However, by analogy with other glycosyltransferases or with glyco- HO sidases, a direct SN2-type transfer seems NiZ CH 0.4 N to be a reasonable hypothesis. It requires OH the intervention of a metallic cation (Mg2+, 2+ 2+ 9a 96 6.4 Mn , Ni ) as an activating Lewis acid CH2 (Fig. 4).[22] The design of transition state mimetics 9b 99 4.7 or bisubstrate analogues has been applied CH2 with success to glycosyltransferases in the CH 2 [23] 9c 99 0.3 gluco, galacto or fuco series. The goal is to simulate in a stable mimetic, most func- tionalities (charge, shape, polarity) of one 9d HO 34 or both substrates in the transition state of

N CH2 the enzymatic reaction. Since the topology of the active site has evolved to best comple- ment the structure of the reaction transition Scheme 2. Synthesis and chitin synthase inhibition potencies of NiZ analogues 9. state, such an analogue should largely sur- pass a conventional substrate or a substrate mimic in affinity.[24] In our group several of transition state mimics for chitin synthase were designed that incorporate single or multiple features of UDP-GlcNAc and/ O FIGURE 5, two-column O or chitin under formation (Fig. 4). On this NH NH basis, we synthesized polyhydroxypyrro- lidines 12a,b,[25,26] difluoromethyl-phos- N O RO N O N3 i N phonoiminosugar 13,[27] pseudodisaccha- O RO O + N N [28] 50-83% rides 14a,b and pyrrolizidine 15. The O O O O five-membered ring iminosugar mimics the half chair conformation and the charge (through protonation at physiological pH) 10 11 (14 examples) of the glycosyl cation, whereas the phos-

i: CuSO4-5H2O, sodium ascorbate, t-BuOH/H2O, 28°C. phonate moiety and the additional GlcNAc residue promote supplementary binding interactions by mimicking either the py- compound R CS (% inhibition yeast growth rophosphate group or the growing chitin at 4 g/mL) (% inhibition at 4 mg/mL) chain. Iminosugar 12b and phosphonoimi- nosugar 13 are weak inhibitors of CS, with NiZ 96 45 IC50s above 2.6 mM. Surprisingly, the pres- ence of a supplementary GlcNAc residue, 11a 95 34 expected to mimic the acceptor of the gly- Cl cosyltransfer reaction, led to a reduction

of the inhibition of the enzyme (IC50 = 5.5 11b F 84 9 mM for 14a, whereas IC50 = 2.6 mM for the corresponding pyrrolidinol 12b). Interest- H3C ingly, compound 12a, an iminosugar found 11c 82 34 in Nature, is significantly more effective [29] (IC50 = 65 mM, Ki = 38 mM). In spite of O minimal structural complexity, iminosugar 11d 80 51 12a was as active as more elaborated natu- O ral or synthetic CS inhibitors. It is worth noting that the stereochemical integrity of 12a is essential for binding since the C(2) Scheme 3. Synthesis and biological evaluation of triazoles 11. epimer 12b exhibited only weak inhibition

FIGURE 6, two-column 52 CHIMIA 2011, 65, No. 1/2 Glycochemistry Today

other groups.[30] The Finney group has focused on the design of dimeric analogues B NHAc that assume a two-active-site mechanism. HO OR H O O Indeed, the opposed orientation of two ad- HO OH jacent monosaccharidic units in the struc- HO O O ture of chitin (each sugar is rotated ≈180° HO O O NH relative to the preceding one in the chain) AcHN O O O P P N O have suggested that a processive transfer- O O O OCH ase such as chitin synthase operates by AcHN 3 Mg HO transferring two sugar units simultane- HO OH NH O ously with alternating up/down configura- N (CH ) HO HO 2 2 OH tion. One possible explanation is that CS NH HO HO or an associated protein pre-assembles the HO R2 CF2P(O)(OEt)2 required disaccharide donor (UDP-chito- R1 biose), allowing extension by two residues 14b: No inhibition at 4 mM at a time. However, this hypothesis was ex- 12a: R1=(CH2)2OH,R2=H, IC50=0.065 mM cluded by running experiments with radio- HO H [31] 12b: R1=H, R2=(CH2)2OH, IC50=2.6 mM N labelled UDP-chitobiose. An alternative HO proposal is that CS has two active sites in HO H 13: R1=CF2P(O)(OEt)OH, R2=H, IC50=39 mM close proximity, one for each sugar orien- OCH [32] AcHN 3 OH tation (Fig. 5). To probe this hypothesis, HO 14a: R =H, R = (CH2)2NH , IC =5.5 mM Finney and coworkers prepared a series of 1 2 O 50 15: IC50=0.82 mM OH dimeric inhibitors that contain two uridinyl moieties linked by an aliphatic tether.[33] Fig. 4. Design of transition-state analogues for chitin synthase. Though the compounds displayed weak inhibition potencies it appeared clearly that bivalent inhibitors such as 16 or 17 displayed activity approximately 10-fold site 1 higher than that of the corresponding UDP monomer 18, providing some evidence for HO AcHN HO O + O the two-active-site mechanism. HO + FIGUHORE 7, two-column HO HO HO O HO AcHN OH AcHN UDP chitin Conclusion site 2 Chitin synthase, the enzyme that con- HO OH O verts UDP-GlcNAc to chitin, is a member of a class of enzymes known as polymer- O NH O H izing (or processive) glycosyltransferases. O N N O O N N O Since chitin is an essential component of 2 H O 16: IC50=1.1 mM HN O the fungal cell wall, that is absent in mam- mals, CS has been recognized as a promis- O HO OH ing antifungal target. The natural peptidyl HO OH O nucleosides nikkomycins and polyoxins OH O NH are competitive chitin synthase inhibitors O H O N N with Kis in the micromolar range, which N N O H O display strong in vitro antifungal activity. HN O OH 17: 35% inibition at 1 mM The search for new CS inhibitors allowed O HO OH identification of candidates with excellent affinities for CS. To date, no detailed in- O formation is available on the structure of O NH CS or its accurate mechanism of action, H C O impeding the rational design of more ef- 3 O N N O 18: IC50=11.8 mM 2 H O (6% inhibition at 1 mM) fective inhibitors. Alternatively, the design of future inhibitors should explore new HO OH mechanistic proposals or should focus on identifying new functional groups that im- Fig. 5. Design of dimeric inhibitors for chitin synthase. prove binding with the enzyme.

Received: August 10, 2010 towards chitin synthase. 15 was the best inhibitor of the series (IC50 New polyhydroxy-pyrrolizidines were = 820 mM) displaying a non-competitive [1] F. M. Klis, A. Boorsma, P. W. J. De Groot, Yeast designed next as constrainedFIGURE analogues8, two-col inhibitionumn pattern, whereas the diastereo- 2006, 23, 185. [2] J.-P. Latgé, Molecul. Microbiol. 2007, 66, 279. of 6-deoxy-homoDMDP 12a, to improve isomeric pyrrolizidines had IC50 values in binding through pre-orientation of the es- the range 4.3–18.9 mM.[28] [3] R. Kollar, E. Petrakova, G. Ashwell, P. W. Robbins, E. Cabib, J. Biol. Chem. 1995, 270, sential hydroxyethyl substituent. Enzymat- Attempts to prepare mechanism-based 1170. ic assays revealed that 7-deoxycasuarine inhibitors of CS have been reported by Glycochemistry Today CHIMIA 2011, 65, No. 1/2 53

[4] R. Minke, J. Blackwell, J. Mol. Biol. 1978, 120, [18] A. Suda, A. Ohta, M. Sudoh, T. Tsukuda, N. 167. Shimma, Heterocycles 2001, 55, 1023. [5] E. Cabib, S. Roberts, B. Bowers, Annu. Rev. [19] K. Obi, J.-I. Uda, K. Iwase, O. Sugimoto, H. Biochem. 1982, 51, 763. Ebisu, A. Matsuda, Bioorg. Med. Chem. Lett. [6] J. Ruiz-Herrera, G. San-Blas, Curr. Drug 2000, 10, 1451. Targets Infect. Disord. 2003, 3, 77; D. M. Rast, [20] P. M. Chaudhary, S. R. Chavan, F. Shirazi, R. A. Merz, A. Jeanguenat, E. Mösinger, Advan. M. Razdan, P. Nimkar, S. P. Maybhate, A. Chitin Sci. 2000, 4, 479. P. Likhite, R. Gonnade, B. M. Hazara, M. V. [7] C. E. Bulawa, Ann. Rev. Microbiol. 1993, 47, Deshpande, S. R. Deshpande, Bioorg. Med. 505; J. A. Shaw, P. C. Mol, B. Bowers, S. J. Chem. 2009, 17, 2433. Silverman, M. H. Valdiviesco, A. Duran, E. [21] A. Plant, P. Thompson, D. M. Williams, J. Org. Cabib, J. Cell Biol. 1991, 114, 111. Chem. 2008, 73, 3714. [8] W.-J. Choi, E. Cabib, Anal. Biochem. 1994, 219, [22] M. L. Sinnott, Chem. Rev. 1990, 90, 1171. 368; E. Cabib, Antimicrob. Agents Chemother. [23] H. Hashimoto, T. Endo, Y. Kajihara, J. Org. 1991, 35, 195; J. P. Martinez, D. Gozalbo, Chem. 1997, 62, 1914; S. Takayama, S. J. Microb. Sem. 1994, 10, 239. Chung, Y. Igarashi, Y. Ichikawa, A. Sepp, R. I. [9] J.-B. Behr, Curr. Med. Chem. (Anti Infective Lechler, J. Wu, T. Hayashi, G. Siuzdak, C.-H. Agents) 2003, 2, 173. Wong, Bioorg. Med. Chem. 1999, 7, 401; Y.-J. [10] M. Winn, R. J. M. Goss, K. Kimura, T. D. H. Kim, M. Ichikawa, Y. Ichikawa, J. Am. Chem. Bugg, Nat. Prod. Rep. 2010, 27, 279. Soc. 1999, 121, 5829; M. L. Mitchell, F. Tian, [11] J.-B. Behr, I. Gautier-Lefebvre, C. Mvondo- L. V. Lee, C.-H. Wong, Angew. Chem. Int. Ed. Evina, G. Guillerm, N. S. Ryder, J. Enz. Inhib. 2002, 41, 3041; R. Wang, D. H. Steensma, Y. 2001, 16, 107. Takaoka, J. W. Yun, T. Kajimoto, C.-H. Wong, [12] J. R. Graybill, L. K. Najvar, R. Bocanegra, R. Bioorg. Med. Chem. 1997, 5, 661; C. Saotome, F. Hector, M. F. Luther, Antimicrob. Agents Y. Kanie, O. Kanie, C.-H. Wong, Bioorg. Med. Chemother. 1998, 42, 2371; R. K. Li, M. G. Chem. 2000, 8, 2249. Rinaldi, Antimicrob. Agents Chemother. 1999, [24] R. Wolfenden, Bioorg. Med. Chem. 1999, 7, 43, 1401; D. A. Stevens, Antimicrob. Agents 647. Chemother. 2000, 44, 2547. [25] J.-B. Behr, G. Guillerm, Tetrahedron: Asymm. [13] For recent reports concerning the screening of 2002, 13, 111. natural products against chitin synthase, see: T. [26] I. Gautier-Lefebvre, J.-B. Behr, G. Guillerm, M. H. Kang, E. I. Hwang, B. S. Yun, K. D. Park, B. Muzard, Eur. J. Med. Chem. 2005, 1255. M. Kwon, C. S. Shin, S. U. Kim, Biol. Pharm. [27] J.-B. Behr, C. Mvondo-Evina, N. Phung, G. Bull. 2007, 30, 598; T. H. Kang, E. I. Hwang, Guillerm, J. Chem. Soc., Perkin Trans. 1 1997, B. S. Yun, C. S. Shin, S. U. Kim, Biol. Pharm. 1597. Bull. 2008, 31, 755; N. H. Yim, E. I. Hwang, B. [28] J.-B. Behr, A. Gainvors-Claisse, A. Belarbi, S. Yun, K. D. Park, J. S. Moon, S. H. Lee, N. D. Nat. Prod. Res. 2006, 20, 1308. Sung, S. U. Kim, Biol. Pharm. Bull. 2008, 31, [29] M. Djebaili, J.-B. Behr, J. Enz. Inhib. Med. 1041; S. Lee, J. N. Kim, E. Kim, M.-S. Kim, Chem. 2005, 20, 123. H. K. Lee, Bull. Korean Chem. Soc. 2009, 30, [30] L. Devel, A. Vidal-Cros, A. Thellend, 3092. Tetrahedron Lett. 2000, 41, 299. [14] R. Chang, T.-T. Vo, N. S. Finney, Carbohydr. [31] R. Chang, A. R. Yeager, N. S. Finney, Org. Res. 2006, 341, 1998. Biomol. Chem. 2003, 1, 39. [15] J. Hajduch, G. Nam, E. J. Kim, R. Fröhlich, J. [32] I. M. Saxena, R. M. Brown, M. Fèvre, R. A. A. Hanover, K. L. Kirk, Carbohydr. Res. 2008, Geremia, B. Henrissat, J. Bacteriol. 1995, 177, 343, 189. 1419. [16] M. Djebaili, J.-B. Behr, unpublished results. [33] A. R. Yeager, N. S. Finney, J. Org. Chem. 2004, [17] J.-B. Behr, T. Gourlain, A. Helimi, G. Guillerm, 69, 613; A. R. Yeager, N. S. Finney, Bioorg. Bioorg. Med. Chem. Lett. 2003, 13, 1713. Med. Chem. 2004, 12, 6451.