Bioorganic & Medicinal Chemistry Letters 24 (2014) 390–393
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters journal homepage: www.elsevier.com/locate/bmcl
Inhibition of serine and proline racemases by substrate-product analogues Matthew Harty a, Mitesh Nagar a, Logan Atkinson a, Christina M. LeGay b, Darren J. Derksen b,⇑, Stephen L. Bearne a,c,⇑ a b c
Department of Biochemistry and Molecular Biology, Dalhousie University, Halifax, Nova Scotia B3H 4R2, Canada Department of Chemistry, St. Francis Xavier University, Antigonish, Nova Scotia B2G 2W5, Canada Department of Chemistry, Dalhousie University, Halifax, Nova Scotia B3H 4R2, Canada
a r t i c l e
i n f o
Article history: Received 22 September 2013 Revised 28 October 2013 Accepted 28 October 2013 Available online 4 November 2013 Keywords: Racemases D-Proline D-Serine Inhibition Substrate-product analogues
a b s t r a c t D-Amino
acids can play important roles as specific biosynthetic building blocks required by organisms or act as regulatory molecules. Consequently, amino acid racemases that catalyze the formation of D-amino acids are potential therapeutic targets. Serine racemase catalyzes the reversible formation of D-serine (a modulator of neurotransmission) from L-serine, while proline racemase (an essential enzymatic and mitogenic protein in trypanosomes) catalyzes the reversible conversion of L-proline to D-proline. We show the substrate-product analogue a-(hydroxymethyl)serine is a modest, linear mixed-type inhibitor of serine racemase from Schizosaccharomyces pombe (Ki = 167 ± 21 mM, Ki0 = 661 ± 81 mM, cf. Km = 19 ± 2 mM). The bicyclic substrate-product analogue of proline, 7-azabicyclo[2.2.1]heptan-7-ium1-carboxylate is a weak inhibitor of proline racemase from Clostridium sticklandii, giving only 29% inhibition at 142.5 mM. However, the more flexible bicyclic substrate-product analogue tetrahydro1H-pyrrolizine-7a(5H)-carboxylate is a noncompetitive inhibitor of proline racemase from C. sticklandii (Ki = 111 ± 15 mM, cf. Km = 5.7 ± 0.5 mM). These results suggest that substrate-product analogue inhibitors of racemases may only be effective when the active site is capacious and/or plastic, or when the inhibitor is sufficiently flexible. Ó 2013 Elsevier Ltd. All rights reserved.
Although L-amino acids predominate in nature, D-amino acids may be employed as specific metabolites, regulators, or used as specific biosynthetic building blocks required by organisms. DAmino acids may be generated or degraded through the action of racemases, which catalyze the inversion of stereochemistry at the a-carbon. That some of these molecules of unusual stereochemistry play unique roles in organisms or lead to particular pathologies if under- or overproduced, has led to the recognition of racemases (and epimerases) as potential therapeutic targets.1– 8 Our studies on mandelate racemase9–12 revealed that mimicking the structure of both enantiomeric substrates (substrate-product analogues) may serve as a useful design strategy for developing inhibitors of racemases. For example, benzilate competitively inhibits mandelate racemase with Ki = 0.7 mM (cf. KS = 0.7 mM).11 Inspection of the X-ray crystal structures of serine racemase13 and proline racemase14 suggested that these two enzymes might be susceptible to inhibition by substrate-product analogues. Serine racemase (SR, EC 5.1.1.18) is a pyridoxal-50 -phosphate (PLP)-dependant enzyme that catalyzes both the racemization and a,b-elimination reaction of the amino acid serine ⇑ Corresponding authors. E-mail addresses: [email protected] (D.J. Derksen), [email protected] (S.L. Bearne). 0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.bmcl.2013.10.061
(Scheme 1A).15 SR is activated by ATP and divalent metal ions such as Mg2+ and Ca2+, and utilizes a two-base mechanism wherein one enantiospecific Brønsted base abstracts the proton from the L-SerPLP aldimine and the conjugate acid of a second enantiospecific Brønsted base protonates the intermediate to form the D-Ser-PLP aldimine, and vice versa (Scheme 1A). SR also catalyzes the a,b-elimination of L- and D-serine to yield pyruvate and ammonia. SR activity in the brain serves as the only endogenous mechanism for generating D-serine, a co-agonist of the N-methyl-D-aspartate receptor (NMDAR) which mediates glutamatergic neurotransmission.16 Inhibition of SR has been suggested as a means of controlling D-serine levels to limit the extent of NMDAR-mediated neurotoxicity17 in Alzheimer’s disease18–20 and amyotrophic lateral sclerosis,21 to ameliorate neuropathic pain,22 and to protect against cerebral ischemia.23 Indeed, a number of SR racemase inhibitors have been developed including dicarboxylic acids,24 small peptides containing a 3-phenylpropanoic acid moiety,25 and hydroxamic acids.26 However, the development of potent and selective SR inhibitors remains a challenge.7 The X-ray crystal structures of human,16 rat,16 and Schizosaccharomyces pombe (fission yeast)13,27 SRs have been solved. The amino acid sequence of SR from S. pombe (SpSR) shares 35% identity with human SR, and the overall structure appears to be conserved for
M. Harty et al. / Bioorg. Med. Chem. Lett. 24 (2014) 390–393
Scheme 1. Two-base racemization mechanisms of serine racemase (A) and proline racemase (B); and substrate-product analogues (C). For SpSR, B1 and B2 are Lys 57 and Ser 82, respectively.13 For TcPR, B1 and B2 are Cys 130 and Cys 300, respectively.14
391
mammalian SRs.13 To delineate the binding of serine within the active site of SpSR, Goto et al.13 generated a model using the structure of lysino-D-alanyl-modified SpSR and rat serine dehydratase with bound O-methyl-L-serine. This model suggested that the hydroxymethyl side chain of the substrate could occupy one of two binding pockets depending on which enantiomer was bound (Fig. 1A). Consequently, we anticipated that the substrate-product analogue a-(hydroxymethyl)serine (1) might be an inhibitor of SR. We sub-cloned the open reading frame encoding SpSR from S. pombe and expressed and purified the enzyme as a fusion protein bearing an N-terminal hexahistidine (His6) tag (see Supplementary data). Inhibition studies with 1 (prepared by the copper-catalyzed condensation of formaldehyde with glycine)28 were conducted by following the production of pyruvate (dehydratase activity) using a coupled assay with L-lactic dehydrogenase.27 As shown in Fig. 2, 1 is a modest, linear mixed-type inhibitor of SpSR. (We also verified the mode of inhibition by plotting the initial velocity data in the form of Cornish–Bowden plots.)29 The inhibition constants for binding to the free enzyme (Ki) and the enzyme-substrate complex (Ki0 ) are equal to 167 ± 21 mM and 661 ± 81 mM, respectively. Thus, 1 binds to the free enzyme with an 8-fold weaker affinity relative to that of the substrate (i.e., Km = 19 ± 2 mM). Although the model presented by Goto et al. (Fig. 1A) suggests that SpSR should be able to accommodate the hydroxymethyl side chains of L- and D-serine simultaneously, as would be the case for binding of 1, our results suggest that such a binding scenario is unfavorable. It is conceivable that the polar b-OH of serine is bound at only a single site and therefore an inhibitor bearing two hydroxyl functions is not tolerated. Alternatively, the packing arrangement within the active site is not flexible enough to accommodate the added steric bulk of 1, or a conformational change occurs in the enzyme as it converts between the substrate-bound L- and D-forms such that only a single side chain can be bound at the active site.
Figure 1. Stereoviews (wall-eyed) showing putative motion of amino acid side chains during racemization (red arrows). Panel A shows a model of SpSR with bound PLP-Lserine aldimine (green) and PLP-D-serine aldimine (blue) superimposed at the active site.13 This model was constructed using the X-ray structures of SpSR (PDB 2ZR8) and rat serine dehydratase complexed with O-methyl-L-serine (PDB 1PWH).30 Panel B shows the active site structure of TcPR with the transition state analogue P2C bound (PDB 1W61) with space available on either side of the planar ring to accommodate the side chains of D- and L-proline.14 The van der Waals surface corresponding to the active site Cys residues is shaded yellow, while surfaces corresponding to Phe 102 and Phe 290 are shaded magenta. This figure was prepared using MacPyMOL.31
392
M. Harty et al. / Bioorg. Med. Chem. Lett. 24 (2014) 390–393
Scheme 2. Synthetic route to 2. Reagents and conditions: Key: (a) Boc2O, NaOH, dioxane, water; (b) (i) MsCl, NEt3, DCM, (ii) t-BuOK, THF; (c) s-BuLi, TMEDA, DMF; (d) 2-methyl-2-butene, NaClO2, NaH2PO4, t-BuOH, H2O; (e) TFA, DCM.
Figure 2. Inhibition of SpSR by 1. A representative double-reciprocal plot is shown in panel A. Assays were conducted as described in the Supplementary data with concentrations of 1 equal to 0 mM (s), 100 mM (4), 200 mM (h), and 300 mM (}). Representative replots of the apparent 1/Vmax (panel B) and apparent Km/Vmax (panel C) values obtained from fitting the initial velocity curves as a function of Lserine concentration to eqn. 1S are shown and the x-intercepts yield the values of Ki0 and Ki according to eqn. 3S, respectively.
Proline racemase (PR, EC 5.1.1.4) catalyzes the reversible interconversion of L- and D-proline. Like SR, PR is a two-base racemase (Scheme 1B); however, the enzyme is not PLP-dependent.32–34 PR from the eukaryotic parasite Trypanosoma cruzi (TcPR) plays a role in immune evasion and in the regulation of intracellular metabolism.4,35 Consequently, the enzyme has been identified as a target for the development of drugs directed against Chagas’ disease.36 Development of inhibitors of PR has been particularly challenging.8 Recent attempts to develop more soluble analogues of the transition state analogue inhibitor pyrrole-2-carboxylate (P2C) proved unsuccessful; however, in silico screening revealed two lead compounds, (E)-4-oxopent-2-enoic acid and its derivative (E)-5-bromo-4-oxopent-2-enoic acid, that were shown to be irreversible competitive inhibitors of TcPR.36 The X-ray crystal structure of TcPR with bound P2C revealed that the carboxylate is held firmly in place by 5 H-bonds and the ligand is completely buried in a predominately hydrophobic active site with no solvent channel connecting the active site to the bulk solvent.14 In addition, the quaternary aminium group of the substrate is held in place by 2 H-bonds.8,14,37 Crystallographic evidence suggests that the enzyme undergoes a significant conformational transition upon binding ligands, with the loop containing Cys 300 being highly mobile.14 Considering the conformational flexibility of PR and that the a-carboxylate and a-aminium groups are held in place via H-bonds, we anticipated that the b-, c-, and dcarbons of the proline ring of the substrate may move during catalysis and that the hydrophobic pocket surrounding the proline ring site might be flexible enough to accommodate substrate-product analogues such as the bicyclic compounds 2 and 3 (Scheme 1C). The synthesis of 2 was accomplished using a modification of the route described by Xiong et al. (Scheme 2; see Supplementary data).38 In our hands, the direct carboxylation of the deprotonated bridgehead carbon with carbon dioxide proved problematic in terms of yield and purification. To overcome these challenges, a
two-step procedure incorporating a formylation-oxidation sequence was employed. The formyl substituent was introduced by quenching the lithiated carbanion, generated in situ, with DMF. After prompt but careful purification of the formylated intermediate 7, the final oxidation to the carboxylic acid 8 was completed in high yield using buffered sodium chlorite conditions (Pinnick oxidation). By optimization of this key carboxylation sequence, we were able to prepare the target bicyclic analogue 2 on a sufficient scale required for the enzyme assays. Compound 3 was obtained from a commercial source. To test the inhibition properties of the bicyclic substrate-product analogues 2 and 3, we sub-cloned the open reading frame encoding PR from Clostridium sticklandii (CsPR) and expressed and purified the enzyme as a fusion protein bearing an N-terminal hexahistidine (His6) tag. Removal of the His6-tag did not significantly alter the kinetic properties of the enzyme (see Supplementary data) and, therefore, the His6-tagged enzyme was used for the inhibition studies. At the highest concentration of 2 compatible with the circular dichroism-based assay (142.5 mM), only 29 (±2)% inhibition was observed (see Supplementary data). Assuming 2 acts as a noncompetitive inhibitor (cf. 3, vide infra), the Ki value can be estimated to be 349 ± 22 mM (with Km = 5.7 mM, [LPro] = 5.0 mM), or 187 ± 12 mM if one assumes competitive inhibition. The bicyclic analogue 2 is therefore only a weak inhibitor of CsPR, binding with an apparent affinity 61-fold weaker than that exhibited for the substrate. Since it was possible that the weak binding could have arisen because of the rigid nature of the bicyclic system, we tested 3 as a potential inhibitor. We anticipated that the two 6-membered rings in 3 might afford greater conformational flexibility relative to the two 5-membered rings in 2. Indeed, the IC50 value for 3 was 67 ± 5 mM (Fig. 2S, Supplementary data).
Figure 3. Inhibition of CsPR by 3. A representative double-reciprocal plot is shown in panel A. Assays were conducted as described in the Supplementary data with concentrations of 3 equal to 0 mM (s), 25 mM (4), 50 mM (h), and 75 mM (}). A representative replot of the apparent 1/Vmax values obtained from fitting the initial velocity curves as a function of L-proline concentration to eqn. 1S is shown in panel B.
M. Harty et al. / Bioorg. Med. Chem. Lett. 24 (2014) 390–393
Although Lineweaver–Burk plots could be fit quite well to either noncompetitive inhibition or linear mixed-type inhibition, Cornish–Bowden plots suggested that 3 was a noncompetitive inhibitor. The Ki value was estimated as 111 ± 15 mM (Fig. 3). Hence, some flexibility of the bicyclic analogue is required for binding. Considering our success at generating a substrate-product analogue inhibitor for mandelate racemase,11 which catalyzes a similar reaction wherein the moiety of the substrate undergoing motion (i.e., phenyl ring)12 is nonpolar and is bound at a plastic and capacious hydrophobic binding pocket,39 we were surprised that 2 and 3 were such poor inhibitors of PR. The active site is either too small (100 Å3)8 relative to the molecular volumes of 2 and 3 (148 Å3 and 161 Å3 for the zwitterions, respectively), not flexible enough to accommodate the bicyclic structures, and/or the proline side chain does not move substantially during catalysis. Indeed, computational studies on TcPR by Rubinstein and Major37 suggest that PR places conformational restraints on the proline substrate to narrow regions of pseudorotational space with Phe 102 and Phe 190 (Fig. 1B) limiting the mobility of the five-membered ring. In addition to demonstrating that the hydrophobic pocket of the active site of PR is not overly plastic, our results furnish experimental support for the conclusions of the computational studies. These are important features of PR that should inform future strategies for inhibitor design. Acknowledgments We thank Professor Richard Singer (Dalhousie University, NS) for kindly providing cultures of S. pombe, and Professor Ken Hirotsu (RIKEN SPring-8 Center, Harima Institute, Japan) for kindly providing the PDB file for the model of SpSR with bound substrate (Fig. 1A). This work was supported by a Discovery Grant from the Natural Sciences and Engineering Research Council (NSERC) of Canada (S.L.B.), an NSERC-USRA (L.A.), a Nova Scotia Health Research Foundation Scotia Grant (S.L.B. & M.H.), and funds from the St. Francis Xavier University Council for Research (D.J.D. & C.M.L.). Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.bmcl.2013.10.061. References and notes 1. Bugg, T. D. H.; Walsh, C. T. Nat. Prod. Rep. 1993, 9, 199. 2. Cox, R. J.; Sutherland, A.; Vederas, J. C. Bioorg. Med. Chem. 2000, 8, 843. 3. Lloyd, M. D.; Darley, D. J.; Wierzbicki, A. S.; Threadgill, M. D. FEBS J. 2008, 275, 1089. 4. Chamond, N.; Cosson, A.; Coatnoan, N.; Minoprio, P. Mol. Biochem. Parasitol. 2009, 165, 170.
393
5. Whalen, K. L.; Pankow, K. L.; Blanke, S. R.; Spies, M. A. ACS Med. Chem. Lett. 2010, 1, 9. 6. Kowatz, T.; Morrison, J. P.; Tanner, M. E.; Naismith, J. H. Protein Sci. 2010, 19, 1337. 7. Jiraskova-Vanickova, J.; Ettrich, R.; Vorlova, B.; Hoffman, H. E.; Lepsik, M.; Jansa, P.; Konvalinka, J. Curr. Drug Targets 2011, 12, 1037. 8. Conti, P.; Tamborini, L.; Pinto, A.; Blondel, A.; Minoprio, P.; Mozzarelli, A.; De Micheli, C. Chem. Rev. 2011, 111, 6919. 9. St. Maurice, M.; Bearne, S. L. Biochemistry 2004, 43, 2524. 10. Burley, R. K.; Bearne, S. L. Bioorg. Med. Chem. Lett. 2005, 15, 4342. 11. Siddiqi, F.; Bourque, J. R.; Jiang, H.; Gardner, M.; Maurice, M. S.; Blouin, C.; Bearne, S. L. Biochemistry 2005, 44, 9013. 12. Lietzan, A. D.; Nagar, M.; Pellmann, E. A.; Bourque, J. R.; Bearne, S. L.; St. Maurice, M. Biochemistry 2012, 51, 1160. 13. Goto, M.; Yamauchi, T.; Kamiya, N.; Miyahara, I.; Yoshimura, T.; Mihara, H.; Kurihara, T.; Hirotsu, K.; Esaki, N. J. Biol. Chem. 2009, 284, 25944. 14. Buschiazzo, A.; Goytia, M.; Schaeffer, F.; Degrave, W.; Shepard, W.; Gregoire, C.; Chamond, N.; Cosson, A.; Berneman, A.; Coatnoan, N.; Alzari, P. M.; Minoprio, P. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 1705. 15. Foltyn, V. N.; Bendikov, I.; De Miranda, J.; Panizzutti, R.; Dumin, E.; Shleper, M.; Li, P.; Toney, M. D.; Kartvelishvily, E.; Wolosker, H. J. Biol. Chem. 2005, 280, 1754. 16. Smith, M. A.; Mack, V.; Ebneth, A.; Moraes, I.; Felicetti, B.; Wood, M.; Schonfeld, D.; Mather, O.; Cesura, A.; Barker, J. J. Biol. Chem. 2010, 285, 12873. 17. Panizzutti, R.; De Miranda, J.; Ribeiro, C. S.; Engelender, S.; Wolosker, H. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 5294. 18. Hashimoto, K.; Fukushima, T.; Shimizu, E.; Okada, S.; Komatsu, N.; Okamura, N.; Koike, K.; Koizumi, H.; Kumakiri, C.; Imai, K.; Iyo, M. Prog. Neuropsychopharmacol. Biol. Psychiatry 2004, 28, 385. 19. Inoue, R.; Hashimoto, K.; Harai, T.; Mori, H. J. Neurosci. 2008, 28, 14486. 20. Wu, S. Z.; Bodles, A. M.; Porter, M. M.; Griffin, W. S.; Basile, A. S.; Barger, S. W. J. Neuroinflammation 2004, 1, 2. 21. Sasabe, J.; Chiba, T.; Yamada, M.; Okamoto, K.; Nishimoto, I.; Matsuoka, M.; Aiso, S. EMBO J. 2007, 26, 4149. 22. Miraucourt, L. S.; Peirs, C.; Dallel, R.; Voisin, D. L. Pain 2011, 152, 1340. 23. Mustafa, A. K.; Ahmad, A. S.; Zeynalov, E.; Gazi, S. K.; Sikka, G.; Ehmsen, J. T.; Barrow, R. K.; Coyle, J. T.; Snyder, S. H.; Dore, S. J. Neurosci. 2010, 30, 1413. 24. Strisovsky, K.; Jiraskova, J.; Mikulova, A.; Rulisek, L.; Konvalinka, J. Biochemistry 2005, 44, 13091. 25. Dixon, S. M.; Li, P.; Liu, R.; Wolosker, H.; Lam, K. S.; Kurth, M. J.; Toney, M. D. J. Med. Chem. 2006, 49, 2388. 26. Hoffman, H. E.; Jiraskova, J.; Cigler, P.; Sanda, M.; Schraml, J.; Konvalinka, J. J. Med. Chem. 2009, 52, 6032. 27. Yamauchi, T.; Goto, M.; Wu, H. Y.; Uo, T.; Yoshimura, T.; Mihara, H.; Kurihara, T.; Miyahara, I.; Hirotsu, K.; Esaki, N. J. Biochem. 2009, 145, 421. 28. Otani, T. T.; Winitz, M. Arch. Biochem. Biophys. 1960, 90, 254. 29. Eisenthal, R.; Cornish-Bowden, A. Biochem. J. 1974, 139, 715. 30. Yamada, T.; Komoto, J.; Takata, Y.; Ogawa, H.; Pitot, H. C.; Takusagawa, F. Biochemistry 2003, 42, 12854. 31. DeLano, W. L. DeLano Scientific LLC 2007, Palo Alto, CA, USA (http:// www.pymol.org). 32. Stadtman, T. C.; Elliott, P. J. Biol. Chem. 1957, 228, 983. 33. Cardinale, G. J.; Abeles, R. H. Biochemistry 1968, 7, 3970. 34. Rudnick, G.; Abeles, R. H. Biochemistry 1975, 14, 4515. 35. Tonelli, R. R.; Silber, A. M.; Almeida-de-Faria, M.; Hirata, I. Y.; Colli, W.; Alves, M. J. Cell. Microbiol. 2004, 6, 733. 36. Berneman, A.; Montout, L.; Goyard, S.; Chamond, N.; Cosson, A.; d’Archivio, S.; Gouault, N.; Uriac, P.; Blondel, A.; Minoprio, P. PLoS ONE 2013, 8, e60955. 37. Rubinstein, A.; Major, D. T. J. Am. Chem. Soc. 2009, 131, 8513. 38. Xiong, H.; Frietze, W.; Andisik, D. W.; Ernst, G. E.; Palmer, W. E.; Hinkley, L.; Varnes, J. G.; Albert, J. S.; Veale, C. A. Tetrahedron Lett. 2010, 51, 6741. 39. Neidhart, D. J.; Howell, P. L.; Petsko, G. A.; Powers, V. M.; Li, R. S.; Kenyon, G. L.; Gerlt, J. A. Biochemistry 1991, 30, 9264.
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters journal homepage: www.elsevier.com/locate/bmcl
Inhibition of serine and proline racemases by substrate-product analogues Matthew Harty a, Mitesh Nagar a, Logan Atkinson a, Christina M. LeGay b, Darren J. Derksen b,⇑, Stephen L. Bearne a,c,⇑ a b c
Department of Biochemistry and Molecular Biology, Dalhousie University, Halifax, Nova Scotia B3H 4R2, Canada Department of Chemistry, St. Francis Xavier University, Antigonish, Nova Scotia B2G 2W5, Canada Department of Chemistry, Dalhousie University, Halifax, Nova Scotia B3H 4R2, Canada
a r t i c l e
i n f o
Article history: Received 22 September 2013 Revised 28 October 2013 Accepted 28 October 2013 Available online 4 November 2013 Keywords: Racemases D-Proline D-Serine Inhibition Substrate-product analogues
a b s t r a c t D-Amino
acids can play important roles as specific biosynthetic building blocks required by organisms or act as regulatory molecules. Consequently, amino acid racemases that catalyze the formation of D-amino acids are potential therapeutic targets. Serine racemase catalyzes the reversible formation of D-serine (a modulator of neurotransmission) from L-serine, while proline racemase (an essential enzymatic and mitogenic protein in trypanosomes) catalyzes the reversible conversion of L-proline to D-proline. We show the substrate-product analogue a-(hydroxymethyl)serine is a modest, linear mixed-type inhibitor of serine racemase from Schizosaccharomyces pombe (Ki = 167 ± 21 mM, Ki0 = 661 ± 81 mM, cf. Km = 19 ± 2 mM). The bicyclic substrate-product analogue of proline, 7-azabicyclo[2.2.1]heptan-7-ium1-carboxylate is a weak inhibitor of proline racemase from Clostridium sticklandii, giving only 29% inhibition at 142.5 mM. However, the more flexible bicyclic substrate-product analogue tetrahydro1H-pyrrolizine-7a(5H)-carboxylate is a noncompetitive inhibitor of proline racemase from C. sticklandii (Ki = 111 ± 15 mM, cf. Km = 5.7 ± 0.5 mM). These results suggest that substrate-product analogue inhibitors of racemases may only be effective when the active site is capacious and/or plastic, or when the inhibitor is sufficiently flexible. Ó 2013 Elsevier Ltd. All rights reserved.
Although L-amino acids predominate in nature, D-amino acids may be employed as specific metabolites, regulators, or used as specific biosynthetic building blocks required by organisms. DAmino acids may be generated or degraded through the action of racemases, which catalyze the inversion of stereochemistry at the a-carbon. That some of these molecules of unusual stereochemistry play unique roles in organisms or lead to particular pathologies if under- or overproduced, has led to the recognition of racemases (and epimerases) as potential therapeutic targets.1– 8 Our studies on mandelate racemase9–12 revealed that mimicking the structure of both enantiomeric substrates (substrate-product analogues) may serve as a useful design strategy for developing inhibitors of racemases. For example, benzilate competitively inhibits mandelate racemase with Ki = 0.7 mM (cf. KS = 0.7 mM).11 Inspection of the X-ray crystal structures of serine racemase13 and proline racemase14 suggested that these two enzymes might be susceptible to inhibition by substrate-product analogues. Serine racemase (SR, EC 5.1.1.18) is a pyridoxal-50 -phosphate (PLP)-dependant enzyme that catalyzes both the racemization and a,b-elimination reaction of the amino acid serine ⇑ Corresponding authors. E-mail addresses: [email protected] (D.J. Derksen), [email protected] (S.L. Bearne). 0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.bmcl.2013.10.061
(Scheme 1A).15 SR is activated by ATP and divalent metal ions such as Mg2+ and Ca2+, and utilizes a two-base mechanism wherein one enantiospecific Brønsted base abstracts the proton from the L-SerPLP aldimine and the conjugate acid of a second enantiospecific Brønsted base protonates the intermediate to form the D-Ser-PLP aldimine, and vice versa (Scheme 1A). SR also catalyzes the a,b-elimination of L- and D-serine to yield pyruvate and ammonia. SR activity in the brain serves as the only endogenous mechanism for generating D-serine, a co-agonist of the N-methyl-D-aspartate receptor (NMDAR) which mediates glutamatergic neurotransmission.16 Inhibition of SR has been suggested as a means of controlling D-serine levels to limit the extent of NMDAR-mediated neurotoxicity17 in Alzheimer’s disease18–20 and amyotrophic lateral sclerosis,21 to ameliorate neuropathic pain,22 and to protect against cerebral ischemia.23 Indeed, a number of SR racemase inhibitors have been developed including dicarboxylic acids,24 small peptides containing a 3-phenylpropanoic acid moiety,25 and hydroxamic acids.26 However, the development of potent and selective SR inhibitors remains a challenge.7 The X-ray crystal structures of human,16 rat,16 and Schizosaccharomyces pombe (fission yeast)13,27 SRs have been solved. The amino acid sequence of SR from S. pombe (SpSR) shares 35% identity with human SR, and the overall structure appears to be conserved for
M. Harty et al. / Bioorg. Med. Chem. Lett. 24 (2014) 390–393
Scheme 1. Two-base racemization mechanisms of serine racemase (A) and proline racemase (B); and substrate-product analogues (C). For SpSR, B1 and B2 are Lys 57 and Ser 82, respectively.13 For TcPR, B1 and B2 are Cys 130 and Cys 300, respectively.14
391
mammalian SRs.13 To delineate the binding of serine within the active site of SpSR, Goto et al.13 generated a model using the structure of lysino-D-alanyl-modified SpSR and rat serine dehydratase with bound O-methyl-L-serine. This model suggested that the hydroxymethyl side chain of the substrate could occupy one of two binding pockets depending on which enantiomer was bound (Fig. 1A). Consequently, we anticipated that the substrate-product analogue a-(hydroxymethyl)serine (1) might be an inhibitor of SR. We sub-cloned the open reading frame encoding SpSR from S. pombe and expressed and purified the enzyme as a fusion protein bearing an N-terminal hexahistidine (His6) tag (see Supplementary data). Inhibition studies with 1 (prepared by the copper-catalyzed condensation of formaldehyde with glycine)28 were conducted by following the production of pyruvate (dehydratase activity) using a coupled assay with L-lactic dehydrogenase.27 As shown in Fig. 2, 1 is a modest, linear mixed-type inhibitor of SpSR. (We also verified the mode of inhibition by plotting the initial velocity data in the form of Cornish–Bowden plots.)29 The inhibition constants for binding to the free enzyme (Ki) and the enzyme-substrate complex (Ki0 ) are equal to 167 ± 21 mM and 661 ± 81 mM, respectively. Thus, 1 binds to the free enzyme with an 8-fold weaker affinity relative to that of the substrate (i.e., Km = 19 ± 2 mM). Although the model presented by Goto et al. (Fig. 1A) suggests that SpSR should be able to accommodate the hydroxymethyl side chains of L- and D-serine simultaneously, as would be the case for binding of 1, our results suggest that such a binding scenario is unfavorable. It is conceivable that the polar b-OH of serine is bound at only a single site and therefore an inhibitor bearing two hydroxyl functions is not tolerated. Alternatively, the packing arrangement within the active site is not flexible enough to accommodate the added steric bulk of 1, or a conformational change occurs in the enzyme as it converts between the substrate-bound L- and D-forms such that only a single side chain can be bound at the active site.
Figure 1. Stereoviews (wall-eyed) showing putative motion of amino acid side chains during racemization (red arrows). Panel A shows a model of SpSR with bound PLP-Lserine aldimine (green) and PLP-D-serine aldimine (blue) superimposed at the active site.13 This model was constructed using the X-ray structures of SpSR (PDB 2ZR8) and rat serine dehydratase complexed with O-methyl-L-serine (PDB 1PWH).30 Panel B shows the active site structure of TcPR with the transition state analogue P2C bound (PDB 1W61) with space available on either side of the planar ring to accommodate the side chains of D- and L-proline.14 The van der Waals surface corresponding to the active site Cys residues is shaded yellow, while surfaces corresponding to Phe 102 and Phe 290 are shaded magenta. This figure was prepared using MacPyMOL.31
392
M. Harty et al. / Bioorg. Med. Chem. Lett. 24 (2014) 390–393
Scheme 2. Synthetic route to 2. Reagents and conditions: Key: (a) Boc2O, NaOH, dioxane, water; (b) (i) MsCl, NEt3, DCM, (ii) t-BuOK, THF; (c) s-BuLi, TMEDA, DMF; (d) 2-methyl-2-butene, NaClO2, NaH2PO4, t-BuOH, H2O; (e) TFA, DCM.
Figure 2. Inhibition of SpSR by 1. A representative double-reciprocal plot is shown in panel A. Assays were conducted as described in the Supplementary data with concentrations of 1 equal to 0 mM (s), 100 mM (4), 200 mM (h), and 300 mM (}). Representative replots of the apparent 1/Vmax (panel B) and apparent Km/Vmax (panel C) values obtained from fitting the initial velocity curves as a function of Lserine concentration to eqn. 1S are shown and the x-intercepts yield the values of Ki0 and Ki according to eqn. 3S, respectively.
Proline racemase (PR, EC 5.1.1.4) catalyzes the reversible interconversion of L- and D-proline. Like SR, PR is a two-base racemase (Scheme 1B); however, the enzyme is not PLP-dependent.32–34 PR from the eukaryotic parasite Trypanosoma cruzi (TcPR) plays a role in immune evasion and in the regulation of intracellular metabolism.4,35 Consequently, the enzyme has been identified as a target for the development of drugs directed against Chagas’ disease.36 Development of inhibitors of PR has been particularly challenging.8 Recent attempts to develop more soluble analogues of the transition state analogue inhibitor pyrrole-2-carboxylate (P2C) proved unsuccessful; however, in silico screening revealed two lead compounds, (E)-4-oxopent-2-enoic acid and its derivative (E)-5-bromo-4-oxopent-2-enoic acid, that were shown to be irreversible competitive inhibitors of TcPR.36 The X-ray crystal structure of TcPR with bound P2C revealed that the carboxylate is held firmly in place by 5 H-bonds and the ligand is completely buried in a predominately hydrophobic active site with no solvent channel connecting the active site to the bulk solvent.14 In addition, the quaternary aminium group of the substrate is held in place by 2 H-bonds.8,14,37 Crystallographic evidence suggests that the enzyme undergoes a significant conformational transition upon binding ligands, with the loop containing Cys 300 being highly mobile.14 Considering the conformational flexibility of PR and that the a-carboxylate and a-aminium groups are held in place via H-bonds, we anticipated that the b-, c-, and dcarbons of the proline ring of the substrate may move during catalysis and that the hydrophobic pocket surrounding the proline ring site might be flexible enough to accommodate substrate-product analogues such as the bicyclic compounds 2 and 3 (Scheme 1C). The synthesis of 2 was accomplished using a modification of the route described by Xiong et al. (Scheme 2; see Supplementary data).38 In our hands, the direct carboxylation of the deprotonated bridgehead carbon with carbon dioxide proved problematic in terms of yield and purification. To overcome these challenges, a
two-step procedure incorporating a formylation-oxidation sequence was employed. The formyl substituent was introduced by quenching the lithiated carbanion, generated in situ, with DMF. After prompt but careful purification of the formylated intermediate 7, the final oxidation to the carboxylic acid 8 was completed in high yield using buffered sodium chlorite conditions (Pinnick oxidation). By optimization of this key carboxylation sequence, we were able to prepare the target bicyclic analogue 2 on a sufficient scale required for the enzyme assays. Compound 3 was obtained from a commercial source. To test the inhibition properties of the bicyclic substrate-product analogues 2 and 3, we sub-cloned the open reading frame encoding PR from Clostridium sticklandii (CsPR) and expressed and purified the enzyme as a fusion protein bearing an N-terminal hexahistidine (His6) tag. Removal of the His6-tag did not significantly alter the kinetic properties of the enzyme (see Supplementary data) and, therefore, the His6-tagged enzyme was used for the inhibition studies. At the highest concentration of 2 compatible with the circular dichroism-based assay (142.5 mM), only 29 (±2)% inhibition was observed (see Supplementary data). Assuming 2 acts as a noncompetitive inhibitor (cf. 3, vide infra), the Ki value can be estimated to be 349 ± 22 mM (with Km = 5.7 mM, [LPro] = 5.0 mM), or 187 ± 12 mM if one assumes competitive inhibition. The bicyclic analogue 2 is therefore only a weak inhibitor of CsPR, binding with an apparent affinity 61-fold weaker than that exhibited for the substrate. Since it was possible that the weak binding could have arisen because of the rigid nature of the bicyclic system, we tested 3 as a potential inhibitor. We anticipated that the two 6-membered rings in 3 might afford greater conformational flexibility relative to the two 5-membered rings in 2. Indeed, the IC50 value for 3 was 67 ± 5 mM (Fig. 2S, Supplementary data).
Figure 3. Inhibition of CsPR by 3. A representative double-reciprocal plot is shown in panel A. Assays were conducted as described in the Supplementary data with concentrations of 3 equal to 0 mM (s), 25 mM (4), 50 mM (h), and 75 mM (}). A representative replot of the apparent 1/Vmax values obtained from fitting the initial velocity curves as a function of L-proline concentration to eqn. 1S is shown in panel B.
M. Harty et al. / Bioorg. Med. Chem. Lett. 24 (2014) 390–393
Although Lineweaver–Burk plots could be fit quite well to either noncompetitive inhibition or linear mixed-type inhibition, Cornish–Bowden plots suggested that 3 was a noncompetitive inhibitor. The Ki value was estimated as 111 ± 15 mM (Fig. 3). Hence, some flexibility of the bicyclic analogue is required for binding. Considering our success at generating a substrate-product analogue inhibitor for mandelate racemase,11 which catalyzes a similar reaction wherein the moiety of the substrate undergoing motion (i.e., phenyl ring)12 is nonpolar and is bound at a plastic and capacious hydrophobic binding pocket,39 we were surprised that 2 and 3 were such poor inhibitors of PR. The active site is either too small (100 Å3)8 relative to the molecular volumes of 2 and 3 (148 Å3 and 161 Å3 for the zwitterions, respectively), not flexible enough to accommodate the bicyclic structures, and/or the proline side chain does not move substantially during catalysis. Indeed, computational studies on TcPR by Rubinstein and Major37 suggest that PR places conformational restraints on the proline substrate to narrow regions of pseudorotational space with Phe 102 and Phe 190 (Fig. 1B) limiting the mobility of the five-membered ring. In addition to demonstrating that the hydrophobic pocket of the active site of PR is not overly plastic, our results furnish experimental support for the conclusions of the computational studies. These are important features of PR that should inform future strategies for inhibitor design. Acknowledgments We thank Professor Richard Singer (Dalhousie University, NS) for kindly providing cultures of S. pombe, and Professor Ken Hirotsu (RIKEN SPring-8 Center, Harima Institute, Japan) for kindly providing the PDB file for the model of SpSR with bound substrate (Fig. 1A). This work was supported by a Discovery Grant from the Natural Sciences and Engineering Research Council (NSERC) of Canada (S.L.B.), an NSERC-USRA (L.A.), a Nova Scotia Health Research Foundation Scotia Grant (S.L.B. & M.H.), and funds from the St. Francis Xavier University Council for Research (D.J.D. & C.M.L.). Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.bmcl.2013.10.061. References and notes 1. Bugg, T. D. H.; Walsh, C. T. Nat. Prod. Rep. 1993, 9, 199. 2. Cox, R. J.; Sutherland, A.; Vederas, J. C. Bioorg. Med. Chem. 2000, 8, 843. 3. Lloyd, M. D.; Darley, D. J.; Wierzbicki, A. S.; Threadgill, M. D. FEBS J. 2008, 275, 1089. 4. Chamond, N.; Cosson, A.; Coatnoan, N.; Minoprio, P. Mol. Biochem. Parasitol. 2009, 165, 170.
393
5. Whalen, K. L.; Pankow, K. L.; Blanke, S. R.; Spies, M. A. ACS Med. Chem. Lett. 2010, 1, 9. 6. Kowatz, T.; Morrison, J. P.; Tanner, M. E.; Naismith, J. H. Protein Sci. 2010, 19, 1337. 7. Jiraskova-Vanickova, J.; Ettrich, R.; Vorlova, B.; Hoffman, H. E.; Lepsik, M.; Jansa, P.; Konvalinka, J. Curr. Drug Targets 2011, 12, 1037. 8. Conti, P.; Tamborini, L.; Pinto, A.; Blondel, A.; Minoprio, P.; Mozzarelli, A.; De Micheli, C. Chem. Rev. 2011, 111, 6919. 9. St. Maurice, M.; Bearne, S. L. Biochemistry 2004, 43, 2524. 10. Burley, R. K.; Bearne, S. L. Bioorg. Med. Chem. Lett. 2005, 15, 4342. 11. Siddiqi, F.; Bourque, J. R.; Jiang, H.; Gardner, M.; Maurice, M. S.; Blouin, C.; Bearne, S. L. Biochemistry 2005, 44, 9013. 12. Lietzan, A. D.; Nagar, M.; Pellmann, E. A.; Bourque, J. R.; Bearne, S. L.; St. Maurice, M. Biochemistry 2012, 51, 1160. 13. Goto, M.; Yamauchi, T.; Kamiya, N.; Miyahara, I.; Yoshimura, T.; Mihara, H.; Kurihara, T.; Hirotsu, K.; Esaki, N. J. Biol. Chem. 2009, 284, 25944. 14. Buschiazzo, A.; Goytia, M.; Schaeffer, F.; Degrave, W.; Shepard, W.; Gregoire, C.; Chamond, N.; Cosson, A.; Berneman, A.; Coatnoan, N.; Alzari, P. M.; Minoprio, P. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 1705. 15. Foltyn, V. N.; Bendikov, I.; De Miranda, J.; Panizzutti, R.; Dumin, E.; Shleper, M.; Li, P.; Toney, M. D.; Kartvelishvily, E.; Wolosker, H. J. Biol. Chem. 2005, 280, 1754. 16. Smith, M. A.; Mack, V.; Ebneth, A.; Moraes, I.; Felicetti, B.; Wood, M.; Schonfeld, D.; Mather, O.; Cesura, A.; Barker, J. J. Biol. Chem. 2010, 285, 12873. 17. Panizzutti, R.; De Miranda, J.; Ribeiro, C. S.; Engelender, S.; Wolosker, H. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 5294. 18. Hashimoto, K.; Fukushima, T.; Shimizu, E.; Okada, S.; Komatsu, N.; Okamura, N.; Koike, K.; Koizumi, H.; Kumakiri, C.; Imai, K.; Iyo, M. Prog. Neuropsychopharmacol. Biol. Psychiatry 2004, 28, 385. 19. Inoue, R.; Hashimoto, K.; Harai, T.; Mori, H. J. Neurosci. 2008, 28, 14486. 20. Wu, S. Z.; Bodles, A. M.; Porter, M. M.; Griffin, W. S.; Basile, A. S.; Barger, S. W. J. Neuroinflammation 2004, 1, 2. 21. Sasabe, J.; Chiba, T.; Yamada, M.; Okamoto, K.; Nishimoto, I.; Matsuoka, M.; Aiso, S. EMBO J. 2007, 26, 4149. 22. Miraucourt, L. S.; Peirs, C.; Dallel, R.; Voisin, D. L. Pain 2011, 152, 1340. 23. Mustafa, A. K.; Ahmad, A. S.; Zeynalov, E.; Gazi, S. K.; Sikka, G.; Ehmsen, J. T.; Barrow, R. K.; Coyle, J. T.; Snyder, S. H.; Dore, S. J. Neurosci. 2010, 30, 1413. 24. Strisovsky, K.; Jiraskova, J.; Mikulova, A.; Rulisek, L.; Konvalinka, J. Biochemistry 2005, 44, 13091. 25. Dixon, S. M.; Li, P.; Liu, R.; Wolosker, H.; Lam, K. S.; Kurth, M. J.; Toney, M. D. J. Med. Chem. 2006, 49, 2388. 26. Hoffman, H. E.; Jiraskova, J.; Cigler, P.; Sanda, M.; Schraml, J.; Konvalinka, J. J. Med. Chem. 2009, 52, 6032. 27. Yamauchi, T.; Goto, M.; Wu, H. Y.; Uo, T.; Yoshimura, T.; Mihara, H.; Kurihara, T.; Miyahara, I.; Hirotsu, K.; Esaki, N. J. Biochem. 2009, 145, 421. 28. Otani, T. T.; Winitz, M. Arch. Biochem. Biophys. 1960, 90, 254. 29. Eisenthal, R.; Cornish-Bowden, A. Biochem. J. 1974, 139, 715. 30. Yamada, T.; Komoto, J.; Takata, Y.; Ogawa, H.; Pitot, H. C.; Takusagawa, F. Biochemistry 2003, 42, 12854. 31. DeLano, W. L. DeLano Scientific LLC 2007, Palo Alto, CA, USA (http:// www.pymol.org). 32. Stadtman, T. C.; Elliott, P. J. Biol. Chem. 1957, 228, 983. 33. Cardinale, G. J.; Abeles, R. H. Biochemistry 1968, 7, 3970. 34. Rudnick, G.; Abeles, R. H. Biochemistry 1975, 14, 4515. 35. Tonelli, R. R.; Silber, A. M.; Almeida-de-Faria, M.; Hirata, I. Y.; Colli, W.; Alves, M. J. Cell. Microbiol. 2004, 6, 733. 36. Berneman, A.; Montout, L.; Goyard, S.; Chamond, N.; Cosson, A.; d’Archivio, S.; Gouault, N.; Uriac, P.; Blondel, A.; Minoprio, P. PLoS ONE 2013, 8, e60955. 37. Rubinstein, A.; Major, D. T. J. Am. Chem. Soc. 2009, 131, 8513. 38. Xiong, H.; Frietze, W.; Andisik, D. W.; Ernst, G. E.; Palmer, W. E.; Hinkley, L.; Varnes, J. G.; Albert, J. S.; Veale, C. A. Tetrahedron Lett. 2010, 51, 6741. 39. Neidhart, D. J.; Howell, P. L.; Petsko, G. A.; Powers, V. M.; Li, R. S.; Kenyon, G. L.; Gerlt, J. A. Biochemistry 1991, 30, 9264.
