Subscriber access provided by UNIV OF CALIFORNIA SAN DIEGO LIBRARIES
Accelerated Publication
The structure of murine APAO reveals molecular detail of vertebrate polyamine catabolism Tove Sjögren, Carola M Wassvik, Arjan Snijder, Anna Aagaard, Taichi Kumanomidou, Louise Barlind, Tim Patrick Kaminski, Akiko Kashima, Takehiro Yokota, and Ola Fjellström Biochemistry, Just Accepted Manuscript • DOI: 10.1021/acs.biochem.6b01140 • Publication Date (Web): 28 Dec 2016 Downloaded from http://pubs.acs.org on December 29, 2016
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Biochemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 39
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biochemistry
The structure of murine APAO reveals molecular detail of vertebrate polyamine catabolism Tove Sjögren*1, Carola M. Wassvik2, Arjan Snijder1, Anna Aagaard1, Taichi Kumanomidou3, Louise Barlind1, Tim P. Kaminski1, Akiko Kashima3, Takehiro Yokota3 and Ola Fjellström2 1
Discovery Sciences, Innovative Medicines and Early Development, AstraZeneca, Pepparedsleden 1, SE-431 83 Mölndal.
2
Cardiovascular and Metabolic Diseases, Innovative Medicines and Early Development, AstraZeneca, Pepparedsleden 1, SE-431 83 Mölndal
3
Discovery Technology Laboratories, Sohyaku. Innovative Research Division, Mitsubishi
Tanabe Pharma Corporation, 1000, Kamoshida-cho, Aoba-ku, Yokohama, Kanagawa 227-0033, Japan
Keywords: Polyamine oxidase, polyamine catabolism, inhibitor
ACS Paragon Plus Environment
1
Biochemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 2 of 39
Abbreviations APAO, N1-acetylspermine oxidase; SMO, spermine oxidase; SSAT, spermine/spermidine acetyl transferase; FAD, flavin adenine dinucleotide; PAO, polyamine oxidase; RMSD, root mean square deviation; APAOox, oxidized N1-acetylspermine oxidase; APAOred, reduced N1acetylspermine oxidase
ACS Paragon Plus Environment
2
Page 3 of 39
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biochemistry
Abstract
N1-acetylspermine oxidase (APAO) catalyses the conversion of N1-acetylspermine or N1acetylspermidine to spermidine or putrescine with concomitant formation of N-acetyl-3aminopropanal and hydrogen peroxide. Here we present the structure of murine APAO in its oxidized holo form, and in complex with substrate. The structures provide a basis for understanding molecular detail of substrate interaction in vertebrate APAO, highlighting a key role for an asparagine residue in coordinating the N1-acetyl group of the substrate. We applied computational methods to the crystal structures to rationalize previous observations regarding substrate charge state. The analysis suggests that APAO features an active site ideally suited for binding of charged polyamines. We also reveal the structure of APAO in complex with the irreversible inhibitor MDL72527. In addition to the covalent adduct, a second MDL72527 molecule is bound in the active site. Binding of MDL72527 is accompanied by altered conformations in the APAO backbone. Based on structures of APAO, we discuss the potential for development of specific inhibitors.
ACS Paragon Plus Environment
3
Biochemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 4 of 39
Polyamines play critical roles in cellular functions including transcription, translation, metabolism, stress regulation and are essential for life in eukaryotes. They are also important in disease, and dysregulation of polyamine catabolism is linked to disorders such as neurodegeneration and cancer.1
Mammalian polyamine catabolism includes two different polyamine oxidases, spermine oxidase (SMO) and N1-acetyl polyamine oxidase (APAO).1 SMO directly oxidizes spermine to form spermidine together with 3-aminopropanal and hydrogen peroxide (H2O2). APAO oxidizes the N1-acetylated forms of spermine or spermidine to generate spermidine or putrescine together with N-acetyl-3-aminopropanal and hydrogen peroxide (Scheme 1). Acetylation of polyamines is catalyzed by spermine/spermidine acetyl transferase (SSAT). The regulation of SMO, APAO and SSAT is key to maintaining polyamine homeostasis.
APAO is a flavin adenine nucleotide (FAD) dependent amine oxidase, first described by Holtta et al. in 19772 and characterized in greater detail in 2003.3, 4 Human APAO shares 39% sequence identity with SMO. Vertebrate APAO and SMO enzymes have evolved from invertebrate polyamine oxidase (PAO) following a gene duplication event.5 APAO and SMO differs from PAO enzymes in substrate specificity profile as well as type of oxidation products. In plants and bacteria, PAO converts spermidine and spermine to 4-aminobutyraldehyde and N-(3aminopropyl)-4-aminobutyraldehyde, respectively, in addition to 1,3-diaminopropane.6 Yeast polyamine oxidase (Fms1) oxidizes both spermine and N1-acetylspermine similar to the mammalian enzymes.7 Structures of Zea mays PAO (ZmPAO),8 and Fms1,9 have been described. The structures contain two well-defined domains. The FAD binding domain displays a classical
ACS Paragon Plus Environment
4
Page 5 of 39
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biochemistry
Rossman fold, which is observed in a number of dinucleotide-binding enzymes. The substratebinding domain is composed of two large inserts extending from the FAD-binding domain, forming a β-sheet flanked by α-helices. The substrate binding site is located at the interface between the two domains. Although the main architecture of invertebrate PAO is expected to be conserved in SMO and APAO, the low level of sequence identity (around 25%) and differences in substrate specificity makes predictions around details of structure and function in the mammalian enzymes uncertain.
Here we describe the structure of murine APAO in its native oxidized form. We also describe the structure of reduced APAO and the complex of oxidized APAO with the substrate N1acetylspermine. The structures give unprecedented detail about substrate binding and allow for a description of the catalytic mechanism in a three-dimensional context. Moreover, we present the structure of APAO in complex with MDL72527, an irreversible inhibitor which has been used extensively in order to elucidate the role of APAO.10 MDL72527 was originally thought to be highly APAO selective, but is in fact equally potent on SMO.11 Because of the lack of selectivity, MDL72527 has limited value as a tool compound. The structures presented here provide an excellent framework for developing selective APAO inhibitors.
Experimental procedures Expression and purification The coding sequence of murine APAO corresponding to residues 3-504, with residues 451-457 replaced by a single glycine residue (3-450/G/458-504), was inserted into pET24a. BL21 Star™ (DE3) (Life technologies, #44-0049) was used as a host for protein production.
ACS Paragon Plus Environment
5
Biochemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 6 of 39
For large scale bacterial culture, modified TB medium was used. The TB medium was obtained from (VWR, # 89126-192) and supplemented with 3 mM MgCl2, 0.8% (v/v) glycerol, 0.02% (w/v) glucose, 100 µg/ml Kanamycin and 0.25x BME Vitamins (Sigma, #B6891). The medium was inoculated from an overnight culture and was grown to OD600 of 0.6-0.8 at 37°C after which the temperature was reduced to 18°C. Protein production was induced with 0.1 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) overnight. Cells were collected by centrifugation at 4000g for 15 min at 4°C, frozen in liquid nitrogen, and stored at -80°C. For purification, cells were thawed and re-suspended in lysis buffer (50 mM HEPES, pH 7.0, 100 mM NaCl, 20 mM imidazole, 10% (v/v) glycerol, 2.5 mM TCEP, Complete EDTA free protease inhibitors). Cells were lysed by using a Branson sonifier (450W) and the cell debris was pelleted by 45 minutes centrifugation at 38k g at 4°C. Ni-affinity purification was performed in batch mode, by addition of Ni-NTA Superflow resin (Qiagen, # 30450) to the cleared lysate. After incubation, the resin was transferred to a gravity flow column and washed with Buffer A (50 mM HEPES, pH 7.0, 100 mM NaCl, 20 mM imidazole, 10% (v/v) glycerol, 2.5 mM TCEP). The protein was eluted with Buffer B (50 mM HEPES, pH 7.0, 100 mM NaCl, 10% (v/v) glycerol, 2.5 mM TCEP, 300 mM imidazole). The His-tag was cleaved off with TEV protease, during dialysis over night against Buffer A (50 mM HEPES, pH 7.0, 100 mM NaCl, 20 mM imidazole, 10% (v/v) glycerol, 2.5 mM TCEP using SnakeSkin Pleated Dialysis Tubing, Pierce, 3.5K, # 68035). After proteolytic removal of the His-tag, a reverse IMAC was run and the cleaved material was collected. The sample was diluted four-fold with ion exchange (IEX) Buffer A (50 mM HEPES pH 7.0, 0 mM NaCl, 10% (v/v) glycerol and 2.5 mM TCEP). The diluted protein was applied to a pre-equilibrated 1 ml Resource Q column (GE healthcare no 17-1177-01) using an Äkta Explorer. The column was
ACS Paragon Plus Environment
6
Page 7 of 39
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biochemistry
washed with 2% IEX Buffer B (50 mM HEPES pH 7.0, 1 M NaCl, 10% (v/v) glycerol and 2.5 mM TCEP) and PAO was eluted with a gradient of IEX Buffer B. The APAO peak from ion exchange chromatography was pooled and concentrated in an Amicon spin concentrator with a 10K cut-off (Milipore, UFC901024). APAO was then polished on a Superdex200 16/60 prepgrade column (GE healthcare, #28-9893-35) in SEC buffer (20 mM HEPES, pH 7.0, 100 mM NaCl, 5% (v/v) glycerol and 2.5 mM TCEP). The APAO peak was pooled and concentrated in an Amicon spin concentrator with a 10K cut-off (Milipore, UFC901024) to final concentrations ranging from 23 mg/ml to 70 mg/ml as determined by nanodrop UV absorbance measurement. The APAO single site mutations in full length wild-type background (N313L, N313T, N313D, and N313A) used for steady state kinetics studies, were produced and purified similar to wild type APAO.
Structure determination Crystals of APAO (amino acids 3-450/G/458-504, 23 mg/ml) were grown by hanging drop vapor diffusion at 20°C in 2.2 M (NH4)2SO4, 0.2 M NaSCN, 0.1 M Tris pH 8. The structure of reduced APAO was obtained by soaking holo-APAO crystals for 10 minutes in a stabilizing solution consisting of 2.5 M (NH4)2SO4, 0.2 M NaSCN, 0.1 M Tris pH 8.0 supplied with 10 mM of in N1-acetylspermine (Sigma-Aldrich). The complex of oxidized APAO with substrate was obtained by soaking holo-APAO crystals for 20 minutes in a stabilizing solution of 2.5 M (NH4)2SO4, 0.2 M NaSCN, 0.1 M MES pH 5.5 supplied with 10 mM of in N1-acetylspermine. The inhibitor complex was obtained by soaking holo-APAO crystals for 4 days in a stabilizing solution of 2.5 M (NH4)2SO4, 0.2 M NaSCN, 0.1 M Tris pH 8 supplied with 5 mM MDL72527
ACS Paragon Plus Environment
7
Biochemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 8 of 39
(synthesized according to Bey et al.10). The crystals were cryo protected with glycerol before being flash frozen in liquid nitrogen. X-ray diffraction data were collected at the ID23-1 and ID29 beamlines at the ESRF, Grenoble, France or I04 at Diamond light source, Oxford, United Kingdom. For data collection on the reduced APAO crystal, an automatic data collection strategy was used, in which the maximum resolution was set to 1.6 Å. However, unexpectedly, the crystal diffracted far beyond that limit. Due to detector geometry the completeness beyond 1.6 Å is decreasing, but since including the incomplete high resolution data improved electron density maps all data were included. Phasing of APAO was done using molecular replacement with a trimmed version of the FADbinding domain of ZmPAO as a model (pdb id 1BQ58). Following initial phasing iterative rounds of density modification and automatic model building and refinement was applied using the programs Parrot12 and Buccaneer13 of the CCP4 suite of programs14 as well as Autobuster.15 Manual model building was done in Coot.16 Structures of APAO in complex with N1acetylspermine and MDL72527 were determined by molecular replacement using the refined holo-APAO structure as a search model. Data collection and refinement statistics can be found in Table 1. Figures were prepared using Pymol.17
Steady state kinetics studies For KS,app determination a coupled enzymatic assay was used. Substrate solutions were prepared as 15-fold 2:1 serial dilution series. The highest concentration of N1-acetylspermine was 70 mM and 500 mM for spermine. Tha assay was performed in a Greiner flat bottom black 384-well plate with total assay volume of 20 µl. Assay buffer contained 0.3 M Glycine, 0.01 % Triton X-100 and 0.001 % BSA at pH 9. All dilutions were carried out in assay buffer. Firstly, 2
ACS Paragon Plus Environment
8
Page 9 of 39
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biochemistry
µl of enzyme solution containing 10 nM enzyme, 600 µM AmplexRed for N1-acetylspermine and 300 µM AmplexRed for spermine and 20 µg/ml horseradish peroxidase were added to each well on ice. Subsequently, 18 µl substrate solution at 4°C was added. For each concentration N=3 measurements were performed. Measurements were performed using a Sapphire2 plate reader (Tecan) at 30°C with top read mode. Velocity was determined by linear fitting using a custom Matlab script during steady state. Kinetic fitting was done with Matlab using MichaelisMenten kinetics.
Computational modelling To better reflect hydrogen bonding and ligand-protein interactions in the APAO-N1acetylspermine structure at physiological pH, a model was prepared using MOE modelling suite.18 Protonation states of side-chains were sampled at pH 7.4 and hydrogen bonding network optimized. The ligand, N1-acetylspermine was used in the mono-protonated form with N5 and N10 neutral and N14 positively charged. The complex was subject to a restrained minimization using the Amber10:EHT force field and the Reaction field solvation model with a dielectric constant of 80.
ACS Paragon Plus Environment
9
Biochemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 10 of 39
Results Protein production and structure determination Murine APAO was produced in E.coli and purified to homogeneity using a combination of metal affinity chromatography, ion exchange, and finally size-exclusion chromatography. Despite the apparent homogeneity and purity, crystallization attempts were unsuccessful. Based on a sequence alignment of mammalian APAO with polyamine oxidases from maize and yeast, we identified a region where mammalian APAO sequences display an insert. The available structures of the distantly related yeast Fms19 and ZmPAO,8 indicated that this insert was creating a flexible loop. A number of constructs containing loop deletions to address potential crystal packing interference, as well as various N- and C-terminal truncations were designed. A construct containing a three amino acid deletion at the N-terminus as well as a loop deletion in which amino acids 451-457 were removed and replaced by a glycine readily generated diffraction quality crystals. The structure was solved by molecular replacement using a trimmed version of the ZmPAO FAD binding domain. The structure was built through iterative rounds of density modification, automatic model building and refinement. The final model contains one molecule in the asymmetric unit and was refined to a resolution of 1.85 Å with R/Rfree factors of 17.1% and 21%, respectively (Table 1).
The overall structure of APAO The overall structure is similar to Fms1 and ZmPAO and the final model contains 466 residues and one FAD molecule (Figure 1). Similar to Fms1 and ZmPAO, the substrate binding domain is composed of two fragments extending from the FAD binding domain. Four loops could not be modelled: Glu 89-Leu 102 (14 residues), Phe 136-Met 144 (9 residues), Ala 160-Asp 166 (7
ACS Paragon Plus Environment
10
Page 11 of 39
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biochemistry
residues) and Ser 365- Ser 368 (4 residues). The Root mean square deviation (RMSD) for 392 Cα atoms in an optimal overlay of the APAO and ZmPAO structures is 1.93Å. For 368 Cα atoms of Fms1, superimposed on APAO, the RMSD value is 2.13 Å. The structure similarity of the FAD binding domains are slightly higher with RMSD value of 1.67 Å and 1.52 Å for 217 residues from ZmPAO and Fms1 respectively (Figure S1). The isoalloxasine ring of the FAD adopts a conformation with a slight bend of approximately 7°. This is in contrast to ZmPAO, which features a highly bent isoalloxasine ring.8 For Fms1 the isoalloxasine ring was modelled as flat in wild type structures which were determined to 2.3-2.4 Å.9 However, in a high resolution structure (2.0 Å) a similar bend to that seen in APAO was reported.19 The isoalloxasine ring conformation reflects the surrounding protein environment and is linked to different redox properties.20
Active site structure The active site cavity is very open compared to the tunnel-shaped cavities described for ZmPAO and Fms1.8, 9 It should be noted that this region harbours an extended region comprising residues Glu 89-Leu 102, which, if structured, may limit the access to the active site. Following complete refinement of the model, an electron density peak in the active site remained. This peak was interpreted as a glycerol molecule. The modelled glycerol molecule is capable of creating several hydrogen bonds to the protein and well resolved solvent molecules in the active site (Figure 2A). Hence, the structure does not represent a truly un-liganded form of APAO. The Lshaped cavity is divided in two parts, one narrow solvent filled groove and one wider sub pocket lined with hydrophobic residues. The solvent filled groove is flanked by the phenol ring of Tyr 204, Ser 473 and the C-terminal end of a helix spanning the interior of the substrate binding domain (α6). The hydrophobic subpocket is lined with hydrophobic residues including Trp 62,
ACS Paragon Plus Environment
11
Biochemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 12 of 39
Phe 331, Phe 375 and Ile 183 as well as the polar Asn 313 (Figure 2A). It also contains the highly conserved Lys 315 coordinating the N5 nitrogen of the isoalloxasine ring via a water molecule. Lys 315 has been implicated in the oxidative half reaction,21 and similar roles have been proposed for the corresponding residues in SMO (Lys367)22 and ZmPAO (Lys 300)23. The site of catalysis is located in the junction between the two sub pockets where the catalytically important His 6424 and the highly conserved Tyr 4305 line a narrow groove. While the electrostatic character of the hydrophobic sub-pocket is neutral, the other sub-pocket holds two distinguishable patches of negative charge. The negative charge originates from Glu 184, together with the negative dipole of the helix spanning the substrate binding domain (residues 167-188), and the partial negative charge from the Oγ of Ser 473 (Figure 2A). These features can help direct the positively charged substrate into the binding pocket. In conclusion, the active site shape and charge distribution is well suited to bind to the positively charged polyamines.
The structure of reduced APAO Due to the presence of the bound FAD co-factor in its oxidized form, crystals of APAO are bright yellow. Addition of substrate to the crystals grown at pH 8.0 causes bleaching of the crystals, indicating that catalysis involving reduction of the FAD group is taking place. The structure of reduced APAO (APAOred) determined from N1-acetylspermine soaked crystals was solved to 1.4 Å resolution. The isoalloxasine ring displays a similar bend to the one seen in the native, oxidized APAO (APAOox) structure. However, the N5 lacks the interaction with the Lys 315-coordinated water, which has been displaced by a shift of the side chain of His 61. In addition, Nδ of His 61 displaces a second solvent atom and accepts a hydrogen bond from Nζ of Lys 315. Moreover, there is a 180° flip of the Trp 62 side chain further limiting the space
ACS Paragon Plus Environment
12
Page 13 of 39
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biochemistry
between Lys 315 and isoalloxasine N5. The active site features an extended electron density, which was interpreted as N1-acetylspermine. The N1-acetyl group of the substrate is coordinated by Asn 313 and the polyamine tail is located in the narrow groove where three solvent molecules are displaced (Figure 2B). The N10 nitrogen is located in a similar position to a water in the APAOox structure and makes the corresponding interactions to the hydroxyl group of Ser 473 and the main chain carbonyl of Val 187. The primary amine N14 is making hydrogen bonds to three solvent atoms. The central part of the substrate (C9-N5) is slightly less well defined than the extremes (Figure S2A, S2B). In the best model, the N5 of N1-acetylspermine is making a hydrogen bond to His 64, and is located directly above the FAD molecule (distance 3.2 Å) (Figure 2B).
Substrate binding to oxidized APAO Although the APAOred-N1-acetylspermine complex provide some detail on substrate interaction, the catalytically more relevant structure is that of APAOox in complex with N1acetylspermine. The enzyme activity of APAO displays a strong pH dependency with an activity maximum at pH 8.0.22 At pH 5.5 the activity is less than 5 %, and indeed, soaking of crystals with substrate at pH 5.5 did not cause bleaching, hence the protein is assumed to still be in its oxidized state. The structure of APAOox in complex with N1-acetylspermine at low pH was solved to 1.6 Å. Similar to what is observed in the APAOred- N1-acetylspermine complex the N1acetyl group of the substrate is coordinated by Asn 313 and the polyamine tail is located in the narrow solvent filled groove (Figure 2C, Figure S2C, S2D). The placement of the central part of N1-acetylspermine is somewhat ambiguous. However, extensive testing of various starting conformations suggest that a binding mode where the substrate N5 is within hydrogen bonding
ACS Paragon Plus Environment
13
Biochemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 14 of 39
distance (3.4 Å) to His 64, and the adjacent C4 carbon 3.1 Å from the N5 of the isoalloxasine ring provides the best fit to the experimental data. The only difference within the protein in the active site observed in the substrate bound structure compared to the APAOox structure is a flip of the Trp 62 side chain, analogous to what is observed in APAOred. Moreover, His 61 adopts a dual conformation, where the two conformations are similar to those observed in holo APAOox and APAOred, respectively.
Mutational studies of substrate binding The APAO-substrate complex structures indicate a role for Asn 313 in coordinating the substrate, N1-acetylspermine, by providing a hydrogen bond to the N1-acetyl group. In order to assess the importance of Asn 313 in substrate binding we mutated this residue to threonine, alanine, leucine and aspartic acid. The kinetic parameters were determined in a coupled enzyme assay where the production of hydrogen peroxide was assessed using the chromogenic substrate AmplexRed. With N1-acetylspermine as a substrate, all mutants displayed a dramatic reduction of kcat /Ks. Assays were run at pH 9 where Ks,app for the wild type enzyme was 3 µM. Ks,app for the mutants ranged from 90 to 230 µM. In contrast, with spermine as substrate the mutation of Asn 313 showed no impact on the kinetic parameters (Table 2).
Computational analysis of substrate binding The APAOox-N1-acetylspermine complex was determined at pH 5.5 where N1-acetylspermine can be expected to carry a charge of +3.25 In order to analyze specific interactions in the precatalytic complex, the APAOox-N1-acetylspermine structure was subjected to computational preparation to reflect pH 7.4, i.e. protonation states of side-chains were sampled and hydrogen-
ACS Paragon Plus Environment
14
Page 15 of 39
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biochemistry
bonding network optimized followed by a restrained minimization. It has been shown that APAO requires its substrates N1-acetylspermine/spermidine to be monoprotonated and that the nitrogen in the reactive C-N bond (N5) needs to be deprotonated for catalysis to occur. For N1acetylspermidine the nitrogen carrying the charge could be assigned to N10, but for N1acetylspermine it could be either N10 or N14.25 In both APAOox and APAOox complexed with N1-acetylspermine, three water molecules form a pyramidal arrangement together with either a fourth water molecule (APAOox) or the N14 in the in N1-acetylspermine (Figure 2A, 2C, Figure S3). Such a fully coordinated site is in good agreement with the suggestion that N14 is the charged atom when N1-acetylspermine is bound to the enzyme to undergo catalysis. The position of N10 in N1-acetylspermine is versatile in the respect that it can harbour a charged or a neutral secondary amine. This is mainly because the Ser 473 hydroxyl group can act as either a hydrogen bond donor or acceptor. In the case of N1-acetylspermidine, a charged primary amine could therefore be accommodated.
The exact location of the substrate N5 from APAOox-N1-acetylspermine crystallographic data is uncertain. The best fit to the experimental data places it between Tyr 430 and His 64 with N5OH and N5-Nε2 distances of 3.5 Å and 3.4, Å respectively. This position is suboptimal for a charged secondary amine since it would be difficult to make good substrate-enzyme interactions for both amine protons simultaneously. It is most likely a reflection of the crystallization pH where both His 64 and N5 are expected to be charged and therefore repel one another. In contrast, at physiological pH with His 64 in its neutral form and with N5 deprotonated (as required for catalysis25) the two are likely to engage in a hydrogen bond.
ACS Paragon Plus Environment
15
Biochemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 16 of 39
Inhibitor binding The putrescine-derived APAO inhibitor MDL7252710 has been extensively used to elucidate the role of APAO in vivo.26 MDL72527 inhibits APAO through an irreversible mechanism.27 Soaking of crystals of oxidized holo-APAO with MDL72527 resulted in bleaching of the APAO crystals. The 1.9 Å electron density map reveals a novel electron density extending from the N5 atom of the isoalloxasine ring (Figure 3, Figure S2E, S2F) suggesting that this atom has been covalently modified by the inhibitor. A flavocyanine adduct previously proposed by Wu et al.27was modelled into the extending electron density, although the exact nature of the covalent product formed between APAO and MDL72527 is uncertain. In the unbiased omit map (Figure S2E), very little electron density is observed beyond the imine nitrogen, N5, indicating that this part of the molecule is disordered. However, it cannot be ruled out that the imine has been hydrolysed between N5 and C4, leaving an aldehyde as the covalent adduct. In the reaction mechanism proposed by Wu et al.27 (Scheme S1), which is not described in detail, the MDL72527 allene acts as nucleophile when reacting with the FAD. An active site base is indicated for the initial removal of the amino α-proton. The only basic residues within reasonable distance of the amino α-carbon is His 64 (d = 3.9 Å). This amino acid takes part in the only noncovalent interaction between the covalently bound MDL72527 and the protein, which is a hydrogen bond with the imine nitrogen N5 (Figure 3A). Here the His 64 imidazole ring has flipped by 180 degrees to allow this hydrogen bond to form. As a consequence, there is no longer a hydrogen bond with Asp 211 and the loop holding this residue shows increased flexibility as suggested by weak electron density and high B- factors.
ACS Paragon Plus Environment
16
Page 17 of 39
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biochemistry
In addition to the bound inhibitor there is an elongated feature in the electron density roughly overlapping with N1-acetylspermine binding. A second molecule of MDL72527 was modelled into the electron density (Figure 3B), but due to the lack of features in the electron density map the confidence in exact placement is low. However, in the best model the N5 and N10 of MDL72527 occupy positions very similar to that of N10 and N14 in N1-acetylspermine in the APAOox - N1-acetylspermine structure (Fig3B). There is only very weak density for the allene pointing out of the active site. However, there is implicit evidence of an extended molecule in the site as Arg 134 assumes a conformation not observed in any of the other APAO structures. In addition, there is an increased level of order in the adjacent loop (residues 138-144) which is fully resolved in the APAO-MDL72527 complex.
ACS Paragon Plus Environment
17
Biochemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 18 of 39
Discussion We here describe the first structure of a mammalian polyamine oxidase. We also describe the structure of APAO in its reduced form as well as complexes with APAOox and the substrate N1acetylspermine and a covalently bound inhibitor. All structures are determined to high resolution. The structure has been elusive and several failed attempts to determine the structure have been reported.22 In our hands, the protein readily formed diffraction quality crystals following surface engineering. However, we do observe several disordered loops, and the structure of APAO in complex with MDL72527 provide some insights into protein flexibility, which may contribute to the lack of success of crystallisation in previous attempts.
The structure of APAOred in complex with N1-acetylspermine displays the expected features of a reduced flavin where the water molecule bridging the N5 of FAD and Lys 315 observed in the holo APAO structure has moved away as a result of the N5 changing its hydrogen bonding properties together with conformational changes of Trp 62 and His 61 reducing the space. Surprisingly, an extended electron density peak which was interpreted as N1-acetylspermine was found in the active site pocket. We believe this is a result of the experimental conditions where excess N1-acetylspermine was present in the crystal. Another possibility, which cannot be ruled out without further studies, is that the electron density, at least in part, represents a reaction intermediate (imine) and/or product (spermidine), both of which should be present in the crystal at an unknown concentration.
ACS Paragon Plus Environment
18
Page 19 of 39
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biochemistry
Although the APAOred-N1-acetylspermine complex provides some insights to substrate binding to APAO, the ambiguities regarding the bound molecule in combination with the reduced FAD makes it a non-ideal model for substrate binding. Instead, we solved the structure of APAO soaked with N1-acetylspermine at pH 5.5. At this pH, catalysis is extremely slow and APAO is therefore considered to be in its oxidized state. N1-acetylspermine binding to APAOox displays a similar hydrogen-bonding network to that observed in the APAOred structure with exception for the central part of the molecule. In the APAOred complex the N5 nitrogen of N1-acetylspermine is within hydrogen bonding distance to the N5 nitrogen of FAD. In contrast, in APAOox-N1acetylspermine complex, the C4 carbon of N1-acetylspermine is the atom closest to the FAD N5 with a distance of 3.1 Å.
The exact reaction mechanism used by enzymes in the amine oxidase family is still a matter of scientific debate. The topic was recently reviewed in detail.28,
29
The three most commonly
proposed mechanisms include single electron transfer with radical formation,30 a polar nucleophilic mechanism where a substrate-flavin intermediate is formed,31 and direct hydride transfer (Scheme 1).28 Common to all three is that a hydrogen must be abstracted from C4 of the amine substrate in the form of a hydrogen atom, a proton or a hydride and transferred to N5 of the flavin cofactor. The binding mode of N1-acetyl spermine in the APAOox complex is in agreement with any of these three models.
The structure of APAO in complex with N1-acetylspermine identified Asn 313 as providing a key interaction to the acetyl group of the substrate. By mutating this residue to alanine, leucine or aspartic acid the possibility to donate a hydrogen bond to the substrate is lost, but only minor
ACS Paragon Plus Environment
19
Biochemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 20 of 39
steric and electrostatic repulsions are expected. The resulting dramatic 30-70 fold decrease in kcat / Ks confirms the importance of the hydrogen bond provided by Asn 313 to N1-acetylspermine for binding to APAO. The residue corresponding to Asn 313 in invertebrate PAO is poorly conserved, but is strictly conserved as an asparagine in APAO and as a threonine in vertebrate SMO orthologues.5 The Asn 313 Thr mutation also resulted in a reduced kcat / Ks. Although the hydroxyl group of threonine could act as a hydrogen bond donor to N1-acetylspermine the reduced side chain length makes this unlikely (d=4 Å for Oγ1 to acetyl O). Spermine is a rather poor substrate for APAO, but notably it is not affected by any of the four mutations, suggesting that this residue is not contributing to spermine binding in APAO. In order to further probe this possible hot-spot for substrate selectivity, corresponding mutational analysis of Thr 365 in SMO would be of great interest. In a previous study using a homology model of SMO, residues Glu 216/Ser 218 in SMO corresponding to Leu 195/Ala 197 in APAO were implicated in substrate selectivity.22 Mutational studies showed that when the residues in SMO were replaced with the corresponding residues in APAO the ability to oxidize N1-acetylspermine was improved.32 However, from the present crystal structure it is evident that Leu 195 and Ala 197 are located more than 8 Å from the bound substrate. The APAO structures offers no evident explanation for the reported shift of substrate specificity in these mutants.
A histidine within the binding site, which is highly conserved across many polyamine oxidases, has been suggested to play a role in both directing the substrate to the catalytic site (His 64 in APAO)24 as well as participating in a hydrogen bond network important for maintaining overall active site structure (His 67-Asn 195-Asp 94 in Fms1).33 Analysis of the APAO-substrate complex shows that His 64 is indeed positioned close enough to interact with N5 of N1-
ACS Paragon Plus Environment
20
Page 21 of 39
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biochemistry
acetylspermine through a hydrogen bond at physiological pH. It also reveals that Asp 211, corresponding to Asn 195 in Fms1, interacts with both the side chain of His 64 and the backbone of Gly 65 maintaining the overall structure (Figure 4A, C, D). Although the N1-acetylspermineAPAO structure provides unique insights into the architecture of substrate binding in the catalytic site, none of the three reaction mechanisms listed above could be ruled out based solely on this structural information. In order to develop further understanding of the exact nature of APAO amine oxidation, additional mutagenesis studies could provide insights in the catalytic mechanism. Aspartate 211 mutations would be of particular interest to establish the role of His 64. Several computational studies of the catalytic mechanism for other members of the amine oxidase family have recently been published.34-36 With the substrate crystal structure described here, analogous theoretical investigations of the catalytic mechanism could be performed for murine APAO as well. The oxidative half-reaction is less studied, but in a review Chaiyen et al. conclude that the oxygen reaction likely occurs through a direct contact between O2 and the reduced flavine N5-C4a locus.37 Structural changes in the surrounding protein in this region are likely to affect the FAD re-oxidation rate. Residues Trp 62, His 61 and Lys 315 are all located within a 5-6 Å radius of the FAD N5. In fact, Lys 315 has proved important for the reaction of the reduced flavin with oxygen21 and Trp 62 and His 61 adopt distinct conformations in the structures disclosed in the present work. Taken together, further studies of these three amino acids in the context of the oxidative half-reaction would likely provide interesting insights in its mechanism at a molecular level. Despite the similar fold and conservation of key residues, the overall shape of the active site cavity differs remarkably from both Fms1 and ZmPAO. In ZmPAO and Fms1 the active site resides in a narrow U-shaped tunnel, making it inaccessible to surrounding solvent due to bulky
ACS Paragon Plus Environment
21
Biochemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 22 of 39
sidechains extending over the cavity (Glu 170 in ZmPAO and Trp 174 in Fms1, Figure 4). The corresponding residue in APAO is a valine (Val 187). In APAO there is an additional hydrophobic cavity in the interface between the domains adjacent to the conserved Lys 315. This sub-pocket is not present in ZmPAO or Fms1. The structures of APAO, ZmPAO and Fms1 in complex with substrates provide additional details on the differences between vertebrate and invertebrate polyamine oxidases (Figure 4). In APAO, the N1-acetyl group of the substrate extends into the more hydrophobic part of the active site, making a hydrogen bond to Asn 313. In ZmPAO and Fms1 where this sub-pocket is not present, substrates adopt conformations extending towards residues Val 196 (ZmPAO) or Asp 94 and Asn 195 (Fms1). It should be noted that although two of the invertebrate PAO structures have been deposited in the Protein Data Bank,38 they are not described in the literature. However, the significance of the binding mode in Fms1 is supported by mutations of Asp 94 and Asn 195, resulting in a reduced rate constant for flavin reduction.33 The differences in active site shape and substrate binding are not surprising given the distant relationship between vertebrate APAO and invertebrate PAO, but highlights the challenges in using distantly related models for detailed prediction of structure and function.
MDL72527 (Scheme 1B) is an irreversible inhibitor which has been extensively used in order to elucidate the role of APAO.10 The structure of APAO in complex with MDL72527 supports the formation of the proposed irreversible covalent adduct.27 Surprisingly MDL72527 is a noncovalent inhibitor of ZmPAO.8 While the exact role of His 64 in the reaction between APAO and MDL72527 (Scheme S1) has not been experimentally established, we speculate that the lack of a base in a similar position in the ZmPAO active site could be responsible for the observed difference in mode of inhibition. MDL72527 was originally thought to be highly APAO
ACS Paragon Plus Environment
22
Page 23 of 39
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biochemistry
selective, but was later demonstrated to be almost equally potent on SMO.11 Because of the lack of selectivity, MDL72527 has limited value as a selective tool compound. To date no specific inhibitors of SMO or PAO have been described. Previously described inhibitors of polyamine oxidases tend to be extended flexible polyamine derivatives, such as guazatine and 1,8diaminoctane.11 The structures of APAO show indeed an active site ideal for binding polyamines with basic nitrogens. Such compounds are likely to suffer from poor bioavailability and cytotoxicity, which would make them unsuitable as tool compounds for in vivo studies. However, the present structures disclose opportunities for more elaborate inhibitors. For example, the hydrophobic part, which harbours the N1-acetyl group of the substrate, should have the capacity to accommodate large groups that could have productive van Der Waals interactions with the surrounding residues. There are also opportunities to gain selectivity over SMO by, for example, exploring interactions to Asn 313 or any of the residues Val 187 (Glu 208 in SMO) and Ile 358 (Cys 409 in SMO) (Figure S4).
The structures presented here represent the first structure of a vertebrate polyamine oxidase. As such, it provides a valuable basis for building molecular understanding of polyamine catabolism and offers unprecedented possibilities for developing selective tools for studying this essential biochemical pathway in health and disease.
ACS Paragon Plus Environment
23
Biochemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 24 of 39
Figure 1. Structure and domain organization of murine APAO. A) Schematic view of the polypeptide sequence and domain organization. The three fragments constituting the FAD binding domain are shown in shades of pink, and the substrate domain fragments are green. Missing loops are indicated in black. (B) The overall structure of murine APAO is colored according to the color scheme in (A). Missing residues are drawn as dashed lines. The FAD molecule is shown in yellow stick representation.
ACS Paragon Plus Environment
24
Page 25 of 39
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biochemistry
Figure 2. The active site of murine APAO. A) The native, oxidized APAO structure featuring a bound glycerol molecule (green) in the active site. B) APAOred in complex with N1acetylspermine C) APAOox in complex with N1-acetylspermine. N1-acetylspermine is shown in orange stick representation and FAD in yellow stick representation. Residues with a black label are conserved in human APAO while red labels indicate a different residue type. Dashes indicate hydrogen bonds. For clarity, only hydrogen bonds commented in the text are included.
ACS Paragon Plus Environment
25
Biochemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 26 of 39
Figure 3. Inhibition of APAO by MDL72527. FAD and MDL72527 are shown in yellow and orange stick representation respectively. Secondary structure element and some key residues of APAO is shown in yellow. The structure of APAOox in complex with N1-acetylspermine is shown in blue. (A) The active site of APAO showing the covalent FAD-MDL72527 product. The hydrogen between the covalent adduct and His 64 (yellow dashes) cause the imidazole to twist 180. As a consequence, the hydrogen bond to Asp 211 seen in APAOox in complex with N1-acetylspermine is lost. (B) The substrate binding domain showing covalent MDL72517 adduct as well as the non-covalently bound MDL72527. Binding of MDL72527 is accompanied by a shift of the guanidine group of Arg 134 of almost 10 Å.
ACS Paragon Plus Environment
26
Page 27 of 39
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biochemistry
Figure 4. Comparison of substrate binding in APAO, ZmPAO and Fms1. A) APAOox in complex with N1-acetylspermine (this work) B) ZmPAO in complex with spermine (pdb id 3ku9, unpublished) C) Fms1 in complex with N1-acetylspermine (pdb id 3cnd, unpublished) and D) spermine (pdb id 1xpq9). The FAD co-factor and substrate are shown in yellow and orange stick representation, respectively.
ACS Paragon Plus Environment
27
Biochemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 28 of 39
Scheme 1. Catalytic mechanism and inhibitor structure. The proposed mechanism for APAOcatalyzed oxidation of N1-actylspermine as described by the hydride transfer mechanism. A hydride is abstracted from C4 of the amine substrate and transferred to N5 of the flavin cofactor producing a positively charged imine and a reduced flavin. The imine spontaneously hydrolyses to yield N-acetyl-3-aminopropanal and spermidine while the oxidized flavin is regenerated by reacting with molecular oxygen giving hydrogen peroxide (H2O2) as a product (not shown).
1
A
4a 5
14 1 10
5
N1-Acetylspermine APAO
H2O
Spermidine
N-Acetyl-3-aminopropanal
B 10 5
MDL72572
ACS Paragon Plus Environment
28
Page 29 of 39
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biochemistry
Table 1. Data collection and refinement statistics from APAO structures. All data sets were collected from single crystals. Values in parenthesis refer to highest-resolution shell. All atoms, which were not refined with dual conformations, were refined with full occupancy. APAOox
Description
APAOred
APAOox + N1-
APAO + MDL72527
acetylspermine
PDB accession code
5lae
5mbx
5lfo
5lgb
ESRF/ID23-1
Diamond/I04
ESRF/ID29
ESRF/ID29
Pilatus
Pilatus
Pilatus
Pilatus
P61
P61
P61
P61
a=b=121.98, c=55.09
a=b=121.99, c=54.91
a=b=122.29, c=54.86
a=b=120.08, c=55.03
49-1.85 (1.89-1.85)
40.81-1.40 (1.48-1.40)
48.7-1.66 ( 1.72-1.66)
0739.54-1.8 (1.9-1.8)
Observed reflections
395873
817460
561204
433803
Unique reflections
40245
81328
55475
42850
Completeness (%)
99.2 (95)
88.8 (51.5)
100 (100)
99.9 (99.3)
Mean I/ σI
14.7 (1.9)
20.1 (4.5)
14.5 (1.6
23.5 (3.5)
10.9 (87.4)
6.6 (41.7)
9.8 (148)
5.4 (70)
9.8 (6.5)
10.1 (8.4)
10.1 (10.1)
10.1 (10.0)
47-1.85
40-1.40
47-1.66
40-1.8
3886
4061
4001
3786
17.1 (21.0)
18.1 (20.1)
19.8 (22.1)
17.6 (19.9)
Data collection statistics
Radiation source
Radiation detector
Space group
Cell dimensions
Resolution (Å)
Rsym%
Redundancy
Refinement statistics
Resolution (Å)
No.
protein
+
ligand
atoms
R (%), Rfree (%)
ACS Paragon Plus Environment
29
Biochemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Wilson B (Å2), Refined
Page 30 of 39
25.0, 31.0
18.4, 21.9
20.8, 29.0
30.7, 39.0
Protein
31.1
21.4
29.8
39.0
FAD
16.2
11.5
16.1
22.5
36.0 (GOL)
22.3 (SP5)
35.7 (SP5)
64.1 64.9 (MDL72527
(Å2)
B-factors
Active site ligand
covalent/non-covalent)
Solvent
39.9
29.1
37.4
46.9
ACS Paragon Plus Environment
30
Page 31 of 39
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biochemistry
Table 2. Kinetic parameters for wild-type and mutant murine polyamine oxidases. All measurements were done in air saturated buffers at pH 9 at 30°C. All kinetic parameters are apparent parameters in air saturated buffer. Errors given are the 95% confidence interval of global non-linear fitting Michaelis-Menten kinetics to reaction velocity vs. substrate
Kinetic parameter
Wild-type PAO
N313D
N313A
N313L
N313T
Ks,app (mM)
0.003±0.001
0.2±0.15
0.1±0.03
0.2±0.1
0.12±0.06
kcat,app (s-1)
27±2
27±5
25±3
25±3
22±2
kcat /Ks (mM-1s-1)
9600±3840
137±128
277±116
144±90
175±105
Ks,app (mM)
8±4
6±3
9±1
7±2
11±3
kcat,app (s-1)
8±1
5.2±1
8.8±0.4
7.5±0.5
8.7±2
kcat/Ks (mM-1s-1)
1±0.6
0.9±0.6
1±0.2
1.1±0.4
0.8±0.4
N1-acetylspermine
Spermine
ACS Paragon Plus Environment
31
Biochemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 32 of 39
Supporting Information. Scheme S1, Figures S1-S4: Scheme S1. The proposed mechanism for the formation of a covalently bound flavocyanine product when MDL72527 reacts with APAO-bound FAD. Figure S1. Stereo views showing the APAO structure compared to ZmPAO and Fms1. Figure S2. Stereo views showing Fo-Fc omit density and 2Fo-Fc maps for APAO crystals soaked in N1-acetylspermine at pH 8.0 and pH 5.5 and MDL72527. Figure S3. Overlay of structures of the active site in oxidized and reduced APAO. Figure S4. Sequence differences between murine APAO and human SMO.
Corresponding Author: * E-mail address
[email protected]
Author Contributions TS, OF, AS, TK, AK and TY designed the work. TS, AS, CW, AA, LB and TK performed the experiments. TS and CW analysed the data. TS, AS, CW and OF drafted the article. All authors critically revised the article and approved the final version
Acknowledgements:
ACS Paragon Plus Environment
32
Page 33 of 39
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biochemistry
Dr. Lovisa Holmberg Schiavone is acknowledged for purification of mutant APAO. We would also like to thank all beam line staff at the ESRF, Grenoble, and especially Dr. Stephanie Monaco, for support during data collection.
References (1) Casero, R. A., and Pegg, A. E. (2009) Polyamine catabolism and disease, Biochem J 421, 323-338. (2) Holtta, E. (1977) Oxidation of spermidine and spermine in rat liver: purification and properties of polyamine oxidase, Biochemistry 16, 91-100. (3) Wu, T., Yankovskaya, V., and McIntire, W. S. (2003) Cloning, sequencing, and heterologous expression of the murine peroxisomal flavoprotein, N1-acetylated polyamine oxidase, J Biol Chem 278, 20514-20525. (4) Vujcic, S., Liang, P., Diegelman, P., Kramer, D. L., and Porter, C. W. (2003) Genomic identification and biochemical characterization of the mammalian polyamine oxidase involved in polyamine back-conversion, Biochem J 370, 19-28. (5) Polticelli, F., Salvi, D., Mariottini, P., Amendola, R., and Cervelli, M. (2012) Molecular evolution of the polyamine oxidase gene family in Metazoa, BMC Evol Biol 12, 90. (6) Federico, R., Ercolini, L., Laurenzi, M., and Angelini, R. (1996) Oxidation of acetylpolyamines by maize polyamine oxidase, Phytochemistry 43, 339-341. (7) Landry, J., and Sternglanz, R. (2003) Yeast Fms1 is a FAD-utilizing polyamine oxidase, Biochem Biophys Res Commun 303, 771-776.
ACS Paragon Plus Environment
33
Biochemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 34 of 39
(8) Binda, C., Coda, A., Angelini, R., Federico, R., Ascenzi, P., and Mattevi, A. (1999) A 30angstrom-long U-shaped catalytic tunnel in the crystal structure of polyamine oxidase, Structure 7, 265-276. (9) Huang, Q., Liu, Q., and Hao, Q. (2005) Crystal structures of Fms1 and its complex with spermine reveal substrate specificity, J Mol Biol 348, 951-959. (10) Bey, P., Bolkenius, F. N., Seiler, N., and Casara, P. (1985) N-2,3-Butadienyl-1,4butanediamine derivatives: potent irreversible inactivators of mammalian polyamine oxidase, J Med Chem 28, 1-2. (11) Bianchi, M., Polticelli, F., Ascenzi, P., Botta, M., Federico, R., Mariottini, P., and Cona, A. (2006) Inhibition of polyamine and spermine oxidases by polyamine analogues, FEBS J 273, 1115-1123. (12) Zhang, K. Y., Cowtan, K., and Main, P. (1997) Combining constraints for electron-density modification, Methods Enzymol 277, 53-64. (13) Cowtan, K. (2008) Fitting molecular fragments into electron density, Acta Crystallogr D Biol Crystallogr 64, 83-89. (14) Winn, M. D., Ballard, C. C., Cowtan, K. D., Dodson, E. J., Emsley, P., Evans, P. R., Keegan, R. M., Krissinel, E. B., Leslie, A. G., McCoy, A., McNicholas, S. J., Murshudov, G. N., Pannu, N. S., Potterton, E. A., Powell, H. R., Read, R. J., Vagin, A., and Wilson, K. S. (2011) Overview of the CCP4 suite and current developments, Acta Crystallogr D Biol Crystallogr 67, 235-242. (15) Bricogne, G., Blanc, E., Brandl, M., Flensburg, C., Keller, P., Paciorek, W., Roversi, P., Sharff, A., Smart, O. S., Vonrhein, C., and Womack, T. O. . (2016) BUSTER version 2.11.6, Global Phasing Ltd. , Cambridge, United Kingdom.
ACS Paragon Plus Environment
34
Page 35 of 39
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biochemistry
(16) Emsley, P., Lohkamp, B., Scott, W. G., and Cowtan, K. (2010) Features and development of Coot, Acta Crystallogr D Biol Crystallogr 66, 486-501. (17) Schrodinger, LLC. (2015) The PyMOL Molecular Graphics System, Version 1.8. (18) (2016) Molecular Operating Environment (MOE),
2013.08 ed., Chemical Computing
Group Inc., 1010 Sherbooke St. West, Suite #910, Montreal, QC, Canada, H3A 2R7. (19) Adachi, M. S., Torres, J. M., and Fitzpatrick, P. F. (2010) Mechanistic studies of the yeast polyamine oxidase Fms1: kinetic mechanism, substrate specificity, and pH dependence, Biochemistry 49, 10440-10448. (20) Mathews, F. S. (1991) New flavoenzymes, Current Opinion in Structural Biology 1, 954967. (21) Henderson Pozzi, M., and Fitzpatrick, P. F. (2010) A lysine conserved in the monoamine oxidase family is involved in oxidation of the reduced flavin in mouse polyamine oxidase, Arch Biochem Biophys 498, 83-88. (22) Tavladoraki, P., Cervelli, M., Antonangeli, F., Minervini, G., Stano, P., Federico, R., Mariottini, P., and Polticelli, F. (2011) Probing mammalian spermine oxidase enzymesubstrate complex through molecular modeling, site-directed mutagenesis and biochemical characterization, Amino Acids 40, 1115-1126. (23) Fiorillo, A., Federico, R., Polticelli, F., Boffi, A., Mazzei, F., Di Fusco, M., Ilari, A., and Tavladoraki, P. (2011) The structure of maize polyamine oxidase K300M mutant in complex with the natural substrates provides a snapshot of the catalytic mechanism of polyamine oxidation, FEBS J 278, 809-821.
ACS Paragon Plus Environment
35
Biochemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 36 of 39
(24) Tormos, J. R., Henderson Pozzi, M., and Fitzpatrick, P. F. (2012) Mechanistic studies of the role of a conserved histidine in a mammalian polyamine oxidase, Arch Biochem Biophys 528, 45-49. (25) Henderson Pozzi, M., Gawandi, V., and Fitzpatrick, P. F. (2009) pH dependence of a mammalian polyamine oxidase: insights into substrate specificity and the role of lysine 315, Biochemistry 48, 1508-1516. (26) Dai, H., Kramer, D. L., Yang, C., Murti, K. G., Porter, C. W., and Cleveland, J. L. (1999) The polyamine oxidase inhibitor MDL-72,527 selectively induces apoptosis of transformed hematopoietic cells through lysosomotropic effects, Cancer Res 59, 49444954. (27) Wu, T., Ling, K. Q., Sayre, L. M., and McIntire, W. S. (2005) Inhibition of murine N1acetylated polyamine oxidase by an acetylenic amine and the allenic amine, MDL 72527, Biochem Biophys Res Commun 326, 483-490. (28) Fitzpatrick, P. F. (2010) Oxidation of amines by flavoproteins, Arch Biochem Biophys 493, 13-25. (29) Gaweska, H., and Fitzpatrick, P. F. (2011) Structures and Mechanism of the Monoamine Oxidase Family, Biomol Concepts 2, 365-377. (30) Silverman, R. B., Zhou, J. J. P., Ding, C. Z., and Lu, X. (1995) Monoamine oxidasecatalyzed
oxidative
decarboxylation
of
cis-
and
trans-5-aminomethyl-3-(4-
methoxyphenyl)dihydrofuran-2(3H)-one. Evidence for the intermediacy of an .alpha.radical, Journal of the American Chemical Society 117, 12895-12896.
ACS Paragon Plus Environment
36
Page 37 of 39
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biochemistry
(31) Binda, C., Mattevi, A., and Edmondson, D. E. (2002) Structure-function relationships in flavoenzyme-dependent amine oxidations: a comparison of polyamine oxidase and monoamine oxidase, J Biol Chem 277, 23973-23976. (32) Cervelli, M., Angelucci, E., Stano, P., Leboffe, L., Federico, R., Antonini, G., Mariottini, P., and Polticelli, F. (2014) The Glu(2)(1)(6)/Ser(2)(1)(8) pocket is a major determinant of spermine oxidase substrate specificity, Biochem J 461, 453-459. (33) Adachi, M. S., Taylor, A. B., Hart, P. J., and Fitzpatrick, P. F. (2012) Mechanistic and structural analyses of the role of His67 in the yeast polyamine oxidase Fms1, Biochemistry 51, 4888-4897. (34) Karasulu, B., Patil, M., and Thiel, W. (2013) Amine oxidation mediated by lysine-specific demethylase 1: quantum mechanics/molecular mechanics insights into mechanism and role of lysine 661, J Am Chem Soc 135, 13400-13413. (35) Zapata-Torres, G., Fierro, A., Barriga-Gonzalez, G., Salgado, J. C., and Celis-Barros, C. (2015) Revealing Monoamine Oxidase B Catalytic Mechanisms by Means of the Quantum Chemical Cluster Approach, J Chem Inf Model 55, 1349-1360. (36) Repic, M., Vianello, R., Purg, M., Duarte, F., Bauer, P., Kamerlin, S. C., and Mavri, J. (2014) Empirical valence bond simulations of the hydride transfer step in the monoamine oxidase B catalyzed metabolism of dopamine, Proteins 82, 3347-3355. (37) Chaiyen, P., Fraaije, M. W., and Mattevi, A. (2012) The enigmatic reaction of flavins with oxygen, Trends in Biochemical Sciences 37, 373-380. (38) Berman, H. M., Battistuz, T., Bhat, T. N., Bluhm, W. F., Bourne, P. E., Burkhardt, K., Feng, Z., Gilliland, G. L., Iype, L., Jain, S., Fagan, P., Marvin, J., Padilla, D., Ravichandran, V.,
ACS Paragon Plus Environment
37
Biochemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 38 of 39
Schneider, B., Thanki, N., Weissig, H., Westbrook, J. D., and Zardecki, C. (2002) The Protein Data Bank, Acta Crystallogr D Biol Crystallogr 58, 899-907.
ACS Paragon Plus Environment
38
Page 39 of 39
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biochemistry
For table of contents use only The structure of murine APAO reveals molecular detail of vertebrate polyamine catabolism Tove Sjögren, Carola M. Wassvik, Arjan Snijder, Anna Aagaard, Taichi Kumanomidou, Louise Barlind, Tim P. Kaminski, Akiko Kashima, Takehiro Yokota and Ola Fjellström
APAO FADox APAO FADred
H2O
ACS Paragon Plus Environment
39
