A comprehensive binding study illustrates ligand recognition in the periplasmic binding protein PotF

•Binding analysis utilizing computational, biochemical, and structural techniques•PotF binds several compounds from polyamine biosynthesis•Crystal structures of closed open apo states emphasize PotF’s plasticity•Binding pocket hydration pattern affects thermodynamic profile for the ligands Periplasmic binding proteins (PBPs) are ubiquitous receptors in gram-negative bacteria. They sense solutes play key roles nutrient uptake. Escherichia coli's putrescine receptor PotF has been reported to bind spermidine. We reveal that similar biogenic polyamines recognized by PotF. Using isothermal titration calorimetry paired with X-ray crystallography different complexes, we unveil PotF's modes detail. The site PBPs is located between two lobes undergo a large conformational change upon ligand recognition. Hence, analyzing influence on complex formation crucial. Therefore, solved crystal an state used them as basis molecular dynamics simulations. In addition, accessed behavior solution all complexes 1H-15N HSQC NMR spectroscopy. This combined provides robust framework understanding future developments drug design protein engineering. A myriad protein-ligand interactions their mechanisms indispensable most processes life (Dunn, 2006Dunn M.F. Protein-ligand Interactions: General Description. ELS, 2006Google Scholar) finely tuned organism's needs. For this purpose, differ: some specifically tailored toward unique partners, while others more flexible recognize wide range molecules. promiscuity recurring feature, where molecules structurally chemically associated native usually bound or lower affinities (Schreiber Keating, 2011Schreiber G. Keating A.E. Protein specificity versus promiscuity.Curr. Opin. Struct. Biol. 2011; 21: 50-61Crossref PubMed Scopus (171) Google Scholar). insights into what determines fine-tunes affinity specificity. such promising candidates changing thermodynamics promiscuous pockets. approach paves way rational engineering, ultimately leading advanced therapeutics diagnostics (Huang et al., 2016Huang P.S. Boyken S.E. Baker D. coming age de novo design.Nature. 2016; 537: 320-327Crossref (630) Scholar; Tiwari 2012Tiwari M.K. Singh R. R.K. Kim I.-W. Lee J.-K. Computational approaches novel functionalities.Comput. Biotechnol. J. 2012; 2: e201209002Crossref (41) Wilson, 2015Wilson C.J. Rational design: Developing next-generation biological nanobiotechnological tools.Wiley Interdiscip. Rev. Nanomed. Nanobiotechnol. 2015; 7: 330-341Crossref (9) work, analyze periplasmic (PBP) E. coli, which belongs PBP class II superfamily. It forms ?/? lobes, connected hinge region (Quiocho Ledvina, 1996Quiocho F.A. Ledvina Atomic structure bacterial active transport chemotaxis: Variation common themes.Mol. Microbiol. 1996; 20: 17-25Crossref (447) ligand-binding at interface adjacent hinge. Upon recognition, twisting closing motion often compared Venus flytrap (Felder 1999Felder C.B. Graul R.C. A.Y. Merkle H.P. Sadee W. proteins: ancient module present multiple receptors.AAPS Pharm. Sci. 1999; 1: E2Crossref (179) One representative family functionally well-characterized maltose-binding maltose/maltodextrin system coli (Mächtel 2019Mächtel Narducci A. Griffith D.A. Cordes T. Orelle C. An integrated mechanism maltose ABC importer.Res. 2019; 170: 321-337Crossref (26) Recent biosensors numerous derived demonstrate fold's versatility applicability sciences (De Lorimier 2009De R.M. Smith J.J. Dwyer M.A. Looger L.L. Sali K.M. Paavola C.D. Rizk S.S. Sadigov S. Conrad D.W. Loew L. al.Construction fluorescent biosensor family.Protein 2009; 11: 2655-2675Crossref (252) Marvin 2011Marvin J.S. Schreiter E.R. Echevarría I.M. genetically encoded, high-signal-to-noise sensor.Proteins Funct. Bioinform. 79: 3025-3036Crossref (61) Scholar, 2013Marvin Borghuis B.G. Tian Cichon Harnett M.T. Akerboom Gordus Renninger S.L. Chen T.W. Bargmann C.I. al.An optimized probe visualizing glutamate neurotransmission.Nat. Methods. 2013; 10: 162-170Crossref (501) 2018Marvin Scholl B. Wilson D.E. Podgorski K. Kazemipour Müller J.A. Schoch Quiroz F.J.U. Rebola N. Bao H. al.Stability, affinity, chromatic variants sensor iGluSnFR.Nat. 2018; 15: 936-939Crossref (140) Especially, evolutionary relationship ionotropic domains (Lee 2019Lee J.Y. Krieger Herguedas García-Nafría Dutta Shaikh S.A. Greger I.H. Bahar I. Druggability simulations GluA3 AMPA N-terminal domain.Structure. 27: 241-252Abstract Full Text PDF (11) illustrates relevance opportunities PBP-based tools neurosciences. previously showed its endogenous (PUT) additionally spermidine (SPD), consists PUT moiety extended propylamine (Scheib 2014Scheib U. Shanmugaratnam Farías-Rico Höcker Change through grafting.J. 2014; 185: 186-192Crossref (16) To gain further knowledge determinants selectivity, systematically characterized event analogs biosynthesis. detail, performed (ITC) experiments ligands, thereby gathering valuable forces drive (Keserü Swinney, 2015Keserü G.M. Swinney D.C. Thermodynamics Kinetics Drug Binding. Wiley-VCH Verlag GmbH & Co. KGaA, 2015Crossref (14) high-resolution these well apo-PotF state, provided details about recognition affinity. holo striking similarities. ligands' amines always coordinated same manner, sometimes even cost adopting strained conformations. study unusual behavior, (MD) simulations, revealed preferred conformations how each extents. probed using nuclear magnetic resonance (NMR) experiments. accordance MD degree appears correlate summary, show extensive capabilities beyond reported. highly comprehensive study, detail crystallographic modes, properties, intrinsic dynamics, pathways, equilibrium distributions systems. As such, believe our platform understand lays groundwork work. analogs, SPD, cadaverine (CDV), spermine (SPM), agmatine (AGM). Binding PotF-ligand were analyzed ITC (Table 1 Figure S1). prevent carryover expression was unfolded refolded during purification. Previous measurements SPD low nanomolar micromolar range, respectively has, expected, high (68 nM) CDV (1.95 ?M), one methylene unit longer than PUT. Similar prior studies shows moderate (29.71 also case SPM (42.69 ?M). Interestingly, second highest AGM (0.22 ?M) test set, despite molecule being bulkier because guanidino group. corresponding amino acids other tested, but did not measurable S3). Although many amines, it seems carboxyl group obstructs binding.Table 1Thermodynamic signatures five obtained (n = 2)LigandKD (?M)?H (kcal × mol?1)?T?S mol?1)?G mol?1)NPutrescine0.068 ± 0.040?23.02 0.1213.28 0.52?9.74 0.400.90 0.02Cadaverine1.95 0.16?14.19 0.026.53 0.07?7.66 0.050.93 0.02Spermidine29.71 1.15?3.65 0.06?2.42 0.04?6.07 0.020.92 0.00Spermine42.69 2.13?3.43 0.01?2.43 0.02?5.86 0.030.97 0.01Agmatine0.22 0.08?13.05 0.214.09 0.42?8.96 0.211.02 0.04 Open table new tab Another interesting property amines. wild-type PUT, contributions enthalpy driven, whereas shift together decreasing entropy-supported 1). observations made crystallized (PDB: 6YED) 6YEB) (Figures 1B 2) (Figure 3). These seven observable differences conformations, evidenced when projecting onto subspace defined opening angles (see STAR Methods definition dihedral angles). summarizes projections work along comparison, PotF:PUT (Vassylyev 1998Vassylyev D.G. Tomitori Kashiwagi Morikawa Igarashi Crystal mutational receptor. Structural substrate specificity.J. Chem. 1998; 273: 17604-17609Abstract (60) variant PotF_SPD see Tables 2 S2.Figure 2Comparison conformationsShow full captionOpen conformation shown light gray surface. conformation, green cartoon sticks. Alignment done PyMOL over C? atoms residues 31–132 bottom lobe (RMSD 0.585 Å).View Large Image ViewerDownload Hi-res image Download (PPT)Figure 3Comparison complexesShow captionBinding pockets form (A; PDB: (B; 6YE0), (C; 6YE6), (D; 6YE8), (E; 6YEC), (F; 6YE6). Ligand forming 2mFo-DFc maps exported ccp4map files phenix.mtz2map water mesh contoured ? PyMOL. Water depicted blue spheres polar contacts dashed lines.View (PPT) Å). lines. adopt various almost linearly correlate. have sharper lie part diagram. 6YE7). perfectly superposes structures, ?119° ?33° (opening, twisting). Slightly PotF:SPM 6YEC) PotF_SPD, defining set intermediate open. widest structure, around 156° 69°. addition similarities global view offer well. network 3A). Most internal least three hydrogen bonds be thermodynamically favorable (Schiebel 2018Schiebel Gaspari Wulsdorf Ngo Sohn Schrader T.E. Cavalli Ostermann Heine Klebe Intriguing role studied neutron trypsin complexes.Nat. Commun. 9: 3559Crossref Ten appear positions, showing bonds. All proximal (section formed S38, D39, D247) distal parts E185 D278) can four bonds, waters aromatic box (formed W37, Y40, W244, F276, Y314) only result protein. characterization solvation crucial binding, unligated ligated gives idea important formation. superposition reveals overlap coordinates polyamine's nitrogens carbons S2D). before Scholar), available date. Our aligns root-mean-square deviation (RMSD) 0.35 Å 341 atoms. Furthermore, 1.63 2), determined roughly 0.6 higher resolution shares space structure; hence, allows better comparability throughout dataset.Table 2Data collection refinement statisticsApo openApo closedPotF:PUTPotF:CDVPotF:SPDPotF:SPMPotF:AGMPDB ID6YED6YEB6YE06YE76YE86YEC6YE6Data collectionSpace groupP 21 1P 32 1Cell dimensions a, b, c (Å)76.2, 53.5, 88.171.1, 71.1, 272.471.0, 71.0, 272.670.9, 70.9, 271.270.8, 70.8, 272.170.4, 70.4, 273.571.2, 71.2, 272.9 ?, ?, ? (°)90, 112, 9090, 90, 12090, 120Molecules per ASU2222222Resolution (Å)45.90–2.18 (2.26–2.18)45.68–1.97 (2.04–1.97)45.64–1.63 (1.69–1.63)45.50–1.60 (1.66–1.60)40.71–1.50 (1.55–1.50)45.58–2.09 (2.17–2.09)45.73–1.56 (1.62–156)Rmerge0.157 (0.970)0.124 (3.324)0.122 (2.482)0.091 (2.547)0.071 (2.663)0.108 (3.334)0.065 (2.922)I/? (I)13.4 (4.7)12.2 (0.7)10.1 (0.7)14.9 (0.7)17.8 (0.8)14.6 (0.7)18.6 (0.7)CC1/20.997 (0.871)0.999 (0.230)0.998 (0.274)0.999 (0.231)0.999 (0.264)0.999 (0.229)0.999 (0.272)Completeness (%)99.9 (99.9)99.5 (99.9)99.9 (99.8)99.2 (94.3)100.0 (100.0)100.0 (100.0)99.8 (99.9)Redundancy10.9 (9.7)10.9 (11.3)10.7 (9.9)10.8 (9.0)11.5 (11.0)11.0 (11.4)11.0 (11.1)No. total reflections735,129 (33,409)629,899 (64,561)1,077,788 (98,256)1,131,843 (88,835)1,471,775 (138,718)525,454 (53,487)1,262,884 (125,863)RefinementRwork/Rfree0.228/0.273 (0.326/0.352)0.191/0.219 (0.323/0.305)0.173/0.201 (0.324/0.325)0.168/0.194 (0.328/0.378)0.162/0.182 (0.334/0.345)0.211/0.239 (0.344/0.362)0.171/0.186 (0.338/0.359)No. Protein5,3815,3735,4115,3775,5105,3515,393 LigandNANA1214202818 Water307261588612606141494B factors (Å2) Protein43.0454.9436.8835.5732.3665.1137.21 Ligand23.8028.5939.0166.1929.65 Water40.1257.4244.1042.6041.5466.9244.35Root-mean-square deviations Bond length (Å)0.0020.0090.0120.0090.0090.0020.010 (°)0.500.940.930.941.090.501.00 PotF, tightly interactions. protonated amine N1 interacts via S38 D39 backbone carbonyls, Y314 hydroxyl, involved salt bridge D247. central strong cation-? N2 C-H-? C4 molecule, enclosing PUT's aliphatic moiety. On side pocket, D278 3B). bridges main driving force series compounds, seem take order coordinate D247 D278. visible CDV, tested does PUT-like element within 1A). downside bending partially disrupting stacking fit position inside 3C S2A). Aside alkyl chain, remarkably side-chain superposing perfectly. additional strain put could account magnitude expressed comparison C2-C3-C4-N2 (?162°) C2-C3-C4-C5 (?93°) related already described compound n-butane (Woller Garbisch, 1972Woller P.B. Garbisch E.W. n-Butane.J. Am. Soc. 1972; 94: 5310-5314Crossref (66) presents skeleton 3D). Its mode resembles S2B), interacting secondary extra extends side, displacing well-coordinated displacement explain loss SPD’s profile. entropically driven well, hydrophobic character interaction. largest series, symmetric polycation physiological pH ones interact D278, terminal extend sites, networks both cases. PotF:SPM's significantly deviate those 1B). Due ligand’s able close entirely, R254 R91 E184 affected displaces pocket. More very Despite slightly larger unfavorable fact, superpose nitrogen NE, which, N2, reflected angle (?160°) S2C). rest nicely 3F). fundamental distribution absence presence all-atom, explicit-solvent totaling 595 ?s simulation time ran six batches. First, simulated concentration, initial input taken (STAR Methods). Then, subsequent ensembles ligands. cases, initially placed considerable distance batches run external forces, allowing freely stabilize angles. data constructing Markov models (MSMs) ensemble trajectories. MSMs calculate slow proven successful (Ferruz 2015Ferruz Harvey M.J. Mestres De Fabritiis Insights fragment hit assays simulations.J. Inf. Model. 55: 2200-2205Crossref (20) Ferruz 2018Ferruz Doerr Vanase-Frawley Zou Y. X. Marr E.S. Nelson R.T. Kormos B.L. Wager T.T. Hou al.Dopamine D3 antagonist cryptic aminergic GPCRs.Sci. Rep. 8: 897Crossref (30) Plattner Noé, 2015Plattner Noé F. plasticity kinetics explored atomistic models.Nat. 6: 7653Crossref (235) started 1A) Being interested opening, projected trajectories discretized 1,000 clusters lumped contain (Prinz 2011Prinz J.-H. Wu Sarich M. Keller Senne Held Chodera J.D. Schütte kinetics: generation validation.J. Phys. 134: 174105Crossref (716) 4). choice number macrostates based inspecting implied timescales S3) suggested previous (Bowman 2009aBowman G.R. Huang Pande V.S. generalized identify states.Methods. 49: 197-201Crossref (225) Prinz describe populated salts concentration. State lowest (3.7% 1.4%) includes narrowest average (120.4° 29.2°). Conformations 4A). converts 2, (139.2°, 50.5°) (39.5% 2.9%). conversion occurs rapidly, 51.0 3.3 ns. Reversion orders slower, 11.5 4.3 ?s, 1's population. rapidly 3 (in 188.7 22.4 ns), (45.0% 2.9%), turn reverts timescale (223.4 17.1 ns). states, 4 transient quickly (68.1 3.4 reverse process slower leads population (11.8% 1.5%). 165.0° 0.2° 87.2° 0.9°, respectively. slowest exchange identified 11.8 4.4 describes transition (4) closest (1). had fold. Specifically, Tang al. apo-MBP rapid nanosecond microsecond major (?95%) minor species (?5%) (Tang 2007Tang Schwieters Clore Open-to-closed observed paramagnetic NMR.Nature. 2007; 449: 1078-1082Crossref (317) recently, Silva found LAO