Subscriber access provided by UNIV OF CALIFORNIA SAN DIEGO LIBRARIES
Letter
Structural Insight into Multivalent Galactoside Binding to Pseudomonas aeruginosa lectin LecA Ricardo Visini, Xian Jin, Myriam Bergmann, Gaelle Michaud, Francesca Pertici, Ou Fu, Aliaksei Pukin, Thomas R Branson, Dominique M.E. Thies-Weesie, Johan Kemmink, Emilie Gillon, Anne Imberty, Achim Stocker, Tamis Darbre, Roland Pieters, and Jean-Louis Reymond ACS Chem. Biol., Just Accepted Manuscript • DOI: 10.1021/acschembio.5b00302 • Publication Date (Web): 21 Aug 2015 Downloaded from http://pubs.acs.org on August 25, 2015
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.
ACS Chemical Biology 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 19
ACS Chemical Biology
1 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
Structural Insight into Multivalent Galactoside Binding to Pseudomonas aeruginosa lectin LecA Ricardo Visini,a Xian Jin,a Myriam Bergmann,a Gaelle Michaud,a Francesca Pertici,b Ou Fu,b Aliaksei Pukin,b Thomas R. Branson,b Dominique M. E. Thies-Weesie,c Johan Kemmink,b Emilie Gillon,d Anne Imberty,d Achim Stocker,a Tamis Darbre,a Roland Pietersb and Jean-Louis Reymonda* a
Department of Chemistry and Biochemistry, University of Berne, Freiestrasse 3, 3012 Berne Switzerland;
[email protected] bDepartment of Medicinal Chemistry & Chemical Biology, Utrecht Institute for Pharmaceutical Sciences, Utrecht University, NL-3508 TB Utrecht, The Netherlands, cVan ’t Hoff Laboratory for Physical and Colloid Chemistry, Debye Institute for Nanomaterials Science, Utrecht University, Padualaan 8, 3584 CH Utrecht, the Netherlands, d Centre de Recherches sur les Macromolécules Végétales, UPR5301, CNRS and Université Grenoble Alpes, 601 rue de la Chimie, F38041 Grenoble, France KEYWORDS Galactosides, Pseudomonas aeruginosa, lectins, X-ray crystallography
Abstract Multivalent galactosides inhibiting Pseudomonas aeruginosa biofilms may help controlling this problematic pathogen. To understand the binding mode of tetravalent glycopeptide dendrimer GalAG2 (Gal-β-OC6H4CO-Lys-Pro-Leu)4(Lys-Phe-Lys-Ile)2Lys-His-Ile-NH2 to its target lectin LecA, crystal structures of LecA complexes with divalent analog GalAG1 (Gal-β-OC6H4CO-LysPro-Leu)2Lys-Phe-Lys-Ile-NH2 and related glucose-triazole linked bis-galactosides 3u3 Gal-βO(CH2)n-(C2HN3)-4-Glc-β-(C2HN3)-[β-Glc-4-(N3HC2)]2-(CH2)n-O-β-Gal (n = 1) and 5u3 (n = 3) were obtained, revealing a chelate bound 3u3, cross-linked 5u3 and monovalently bound GalAG1. Nevertheless a chelate bound model better explaining their strong LecA binding and the absence of lectin aggregation was obtained by modeling for all three ligands. A model of chelate bound GalAG2.LecA complex was also obtained rationalizing its unusually tight LecA binding (KD = 2.5 nM) and aggregation by lectin cross-linking. The very weak biofilm inhibition with divalent LecA inhibitors suggests that lectin aggregation is necessary for biofilm inhibition by GalAG2, pointing to multivalent glycoclusters as a unique opportunity to control P. aeruginosa biofilms.
ACS Paragon Plus Environment
ACS Chemical Biology
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
2
Introduction The galactose specific lectin LecA1 mediates biofilm formation in the opportunistic pathogen P. aeruginosa responsible for lethal airways infections in immunocompromised and cystic fibrosis patients.2 Various multivalent galactosides have been reported as tight binding LecA ligands exploiting the cluster glycoside effect,3-10 a phenomenon which generally allows to obtain high affinity multivalent ligands to various carbohydrate binding proteins.11-26 One such ligand is the tetravalent glycopeptide dendrimer GalAG2 which not only binds LecA tightly but also blocks biofilm formation and disperses established biofilms in vitro (Figure 1B).27 Binding is partly mediated by an unusual CH-π interaction between C(ε)-H of histidine 50 on LecA and the aromatic ring of the phenyl galactoside ligand on GalAG2,28 which also explains the tight binding of aromatic galactosides previously reported with LecA.29-31 On the other hand interactions between the peptide portion of the dendrimer and LecA seem to play only a limited role in affinity.32 In the course of our studies with GalAG2 the smaller bis-galactoside analog GalAG1 was found to be a much weaker LecA binder with only weak biofilm inhibition activity. The crystal structure of the terminal tripeptide GalAG0 in complex with LecA suggested that GalAG1 and GalAG2 were too small to engage in a chelate binding mode with LecA, implying that the multivalency effect on binding affinity simply resulted from a local concentration effect acting on rebinding kinetics.27 However tight LecA binding comparable to that of GalAG2 was recently reported with the divalent inhibitor 3u3 and to a lesser extent 5u3 featuring a pair of galactosides linked by a rigid rod spacer with a length comparable to the distance between two adjacent LecA binding sites (Figure 1A).4, 33, 34 Tight LecA binding by 3u3 and 5u3 was attributed to a chelate binding mode and supported by molecular modeling. Considering that galactoside LecA complexes usually yield to crystallization, we initiated a structural study to obtain an experimental confirmation of the chelate binding mode of 3u3 and 5u3 and to test whether GalAG1 and GalAG2 might also engage in a similar chelate binding mode. ACS Paragon Plus Environment
Page 2 of 19
Page 3 of 19
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
3
ACS Chemical Biology
A
HO OH O HO O OH
n
HO O
OH OH N N N HO O N N O N HO O N N N OH OH HO
N N N HO
C
OH HO
OH
O HO
n
LecA
3u3 (n = 1) 5u3 (n = 3)
HO OH O HO O OH
3u3.LecA 5u3.LecA
OH
N N N
GalTZ
LecA
B HO OH O
HO
OH O
H N
Sugar 4
O O
N
O
Sugar 3
NH O
N NH H
O
GalAG1.LecA
NH
O
O
O
N
NH3
O +O
NH
+
HN
H N O
O
O
N NH H
N
O H N
N H
N
NH3+
GalAG0
GalA-KPL
GalAG1 (GalA-KPL)2KFKI GalAG2 (GalA-KPL)4(KFKI)2KHI
O O OH HO
N
NH HN O
GalAG2.LecA
HN
O NH3+ O
HN HN
LecA
O
O
HN
O
LecA
HO
NH2 H3N
NH
Sugar 1 HO OH O HO O OH
OH HO O
O
NH NH3+
HO O
H N
O
NH
O
O
NH
HO O
O HO
O NH3+
Sugar 2
Figure 1. A. Structures of rigid-rod spacer bis-galactosides 3u3, 5u3, and monovalent reference GalTZ. B. Structures of peptide dendrimers GalAG2, GalAG1, and monovalent reference GalAG0. C. Model of binding mode between multivalent galactosides and lectin LecA proposed in this study. The labeling Sugar 1-2-3-4 refers to the numbering of galactosides in the MD data in Figure 4.
Herein we report a structural and functional study of LecA complexes with 3u3, 5u3, GalAG1 and GalAG2. Binding affinities of the glycopeptide dendrimers were measured by ITC (isothermal titration calorimetry) under the same conditions as 3u3 and 5u3, and the aggregation behaviour of LecA in the presence of the different ligands was tested by analytical ultracentrifugation. X-ray crystal structures of LecA complexes with 3u3 and 5u3 were obtained at 1.19 Å and 1.82 Å resolution. The binding mode of GalAG1 and GalAG2 was deduced from molecular modeling starting with a partial structure of the GalAG1.LecA complex at 1.40 Å. Taken together the data suggest that all of these ligands bind LecA in solution in a chelate binding mode involving two adjacent galactose binding sites of the LecA tetramer (Figure 1C). Only the tetravalent ligand GalAG2 induced significant lectin aggregation by utilizing its four available galactose units for
ACS Paragon Plus Environment
ACS Chemical Biology
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
4
binding to two different LecA tetramers. The much weaker biofilm inhibitory effects seen with mono- and divalent LecA ligands suggest that the lectin cross-linking ability of GalAG2 is critical for biofilm inhibition activity.
Results and Discussion Lectin Binding Divalent galactosides 3u3 and 5u3 were prepared by a convergent solution phase synthesis, and glycopeptide dendrimers GalAG1 and GalAG2 and the corresponding tripeptide GalAG0 by solidphase peptide synthesis as described previously.4, 27, 33 The binding affinities of the three glycopeptides with LecA were measured by ITC under the same conditions as 3u3 and 5u3 and the monovalent reference compound GalTZ (Table 1 and Supplementary Figure S1). All three galactosylated peptides had a stronger binding enthalpy per galactose unit (~15 kcal/mol) than the triazole galactose compounds (~7 kcal/mol) due to the additional CH-π interaction of their phenyl group with His50. On the other hand the triazole based ligands 3u3, 5u3 and GalTZ showed a much lower entropic penalty of binding than the conformationally quite flexible galactosylated peptides. The tetravalent dendrimer GalAG2 showed the tightest LecA binding (KD = 2.5 nM) driven by a strongly negative binding enthalpy compensating an unfavourable binding entropy, while 3u3 was the tightest binding of the divalent ligands (KD = 28 nM) , with approximately 3-fold stronger binding than GalAG1 (KD = 83 nM) and 4-fold stronger than 5u3 (KD = 130 nM). Analytical ultracentrifugation (AUC) showed that the tetravalent GalAG2 induced LecA aggregation by the detection of up to 53 % of LecA appearing as dimer in the AUC c(s) distribution, while the three divalent ligands 3u3, 5u3 and GalAG1 did not induce significant dimer formation (Table 1, Figure 2). Note that there was no reduction in the absorbance signal with any of the ligands indicating that higher oligomers or aggregates were not formed.
ACS Paragon Plus Environment
Page 4 of 19
Page 5 of 19
Table 1. ITC and ultracentrifugation data. na
KD, nMb
% bound Gal
AUC, % dimer formed at 0.5/1.0/2.0 Eq.c 0.65 ± 0.02 100 (ref) 2960 ± 50 -17.8 ± 0.3 10.2 -7.5 n.d. GalAG0 0.302 ± 0.003 93 83 ± 12 -29 ± 0.5 19.3 -9.7 2.3/1.0/2.0 GalAG1 0.136 ± 0.001 80 2.5 ± 0.1 -69 ± 1.1 57 -11.7 33/53/26 GalAG2 0.92 ± 0.05 22000 ± 5000 -8.3 ± 0.7 2.0 -6.3 n.d. GalTZ 0.55 ± 0.001 28 ± 3 -11.6 ± 0.05 1.3 -10.3 2.0/1.1/2.2 3u3 0.47 ± 0.005 130 ± 300 -11.5 ± 0.17 2.1 -9.4 0.0/0.0/3.8 5u3 a) stoichiometry value. b) Thermodynamic parameters and dissociation constant Kd reported as an average of two independent runs from ITC in Tris 20 mM, pH 7.5, 100 mM NaCl, 100 µM CaCl2, 25˚C. Concentrations (Ligand/LecA) are the following: GalAG0 (0.5 mM/0.0516 mM), GalAG1 (0.25 mM/0.0486 mM), GalAG2 (0.03 mM/0.018 mM). Data for GalTZ, 3u3 and 5u3 from ref. 4. c) % dimer formed in sedimentation velocity analytical ultracentrifugation (AUC) using 40 µM LecA and the indicated number of equivalent galactosyl groups per LecA binding site, see also Figure 2 and supporting information. 1.1 % of dimer is detected with LecA alone.
A 3u3
∆H, kcal/mol
LecA LecA
LecA LecA
6.0
4.0
3.0
2.0
2.0
1.0
1.0
0.0 2.5 5.0 7.5 10.0 Sedimentation coefficient / S
1.01.0 Dimer Eq. 1
4.0 2.0eq Dimer 1 2.0 Eq.
3.0
0.0
2.5 5.0 7.5 10.0 Sedimentation coefficient / S
2.0eq Dimer 1 2.0 Eq.
3.0
2.0
2.0
1.0
1.0
0.0
0.0 0.0
Eq. 1 0.50.5 Dimer
1.01.0 Dimer Eq. 1
4.0 2.0eq Dimer 1 2.0 Eq.
c(s) 280 nm
3.0
LecA LecA
5.0
Eq. 1 0.50.5 Dimer
1.01.0 Dimer Eq. 1
5.0
c (s) 280 nm
4.0
5.0
Eq. 1 0.50.5 Dimer
1.01.0 Dimer Eq. 1 2.0eq Dimer 1 2.0 Eq.
6.0
LecA LecA
Eq. 1 0.50.5 Dimer
5.0
D GalAG2
6.0
7.0
6.0
∆G, kcal/mol
C GalAG1
B 5u3
7.0
-T∆S, kcal/mol
c(s) 280 nm
Ligands
c (s) 280 nm
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
5
ACS Chemical Biology
0.0 0.0
2.5 5.0 7.5 10.0 Sedimentation coefficient / S
0.0
2.5 5.0 7.5 10.0 Sedimentation coefficient / S
Figure 2. Analytical ultracentrifugation of lectin LecA in the presence of di- and tetra-valent ligands.
X-ray crystallography Direct co-crystallization yielded a well-diffracting crystal for the 5u3.LecA complex only, while crystals of the 3u3.LecA and GalAG1.LecA complexes were obtained by soaking of galactose containing LecA crystals. Datasets at 1.19 Å, 1.82 Å respectively 1.40 Å were obtained for each of the complexes (Supplementary Table S1). GalAG2 failed to give any crystals with LecA under a variety of conditions due to the induced precipitation. The structure of the 3u3.LecA complex indicates a chelate binding mode of 3u3 with the lectin (Figure 3A). The electron density shows the bound β-galactosyl groups and the adjacent methyl-triazole-glucoside part of the linker but is partly discontinuous for the central glucose unit, ACS Paragon Plus Environment
ACS Chemical Biology
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
6
nevertheless a suitable model for the bound ligand can be obtained. While the triazole next to both galactosyl groups do not engage in the CH-π interaction with His50 seen with other aromatic galactosides, the adjacent glucose units participate in a hydrogen bond network connecting its C(2)OH and C(3)-OH to His50, Tyr36 and Gln40 via four structural water molecules. The central part of the linker consisting of a glucose unit flanked by two adjacent triazole rings is not well visible in the structure. The central glucose unit could be fitted in the electron density map with an all equatorial chair conformation, with hydrogen bond contacts between the C(2)-OH and C(3)-OH and Gln40 on each LecA monomer. The difficulty to resolve the central glucose unit might result from the asymmetric nature of 3u3 implying that the electron density might reflect a mixture of two possible binding orientations. The structure of the 5u3.LecA complex is well resolved and features the divalent galactoside bridging the galactose binding sites of two different lectin tetramers in the crystal lattice (Figure 3B). Binding is asymmetric with one end of the ligand stretched out with its first glucose unit engaging in hydrogen bond contacts with Thr39 on the lectin surface via a glucose C(3)-OH to threonine backbone amide nitrogen interaction (3.0 Å) and a glucose C(2)-OH to threonine OH interaction (3.2 Å). The other end of 5u3 makes a U-turn in solution and has no contact with the protein except for the bound galactose unit. In the structure of the GalAG1.LecA complex electron density is visible only for the terminal galactosylated dipeptide (Figure 3C). The phenyl group of the galactoside interacts with the C(ε)-H of His50 at a distance of 2.15 Å indicating a strong CH-π interaction similar to that observed previously with various other phenyl galactoside complexes of LecA.28 The pair of GalAG1 terminal galactosylated dipeptides occupying adjacent galactose binding sites of the same LecA tetramer point in opposite directions on the protein surface. With this orientation of the terminal galactosyl-dipeptide arm GalAG1 cannot bridge the two adjacent LecA binding sites, implying that the crystal contains one GalAG1 molecule for each galactose binding site.
ACS Paragon Plus Environment
Page 6 of 19
7
ACS Chemical Biology
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
8
stable during MD, with an average O4-O4 distance of 31.5 Å very close to the crystallographic distance of 30.9 Å. The distance fluctuation corresponded to that observed for the distance separating the two calcium ions in all the different MD simulations and reflected relative movements of the LecA monomers, showing that the LecA dimer interaction is somewhat flexible during MD. In contrast to the crystal structure where structural water molecules were seen in between the linker and the protein surface, the glucose-triazole linker of 3u3 was very close to the protein surface during MD (Figure 4D). MD simulation of free 5u3 showed a rather mobile ligand sometimes adopting a U-turn conformation comparable to the conformation in the cross-linked binding mode of the crystal, with a broader range of O4-O4 distances than for 3u3 (Figure 4A). Nevertheless a chelate bound model of 5u3 was assembled and was similarly stable during MD with a similar O4-O4 distance of 32.5 Å. The stability of this chelate bound complex during MD leads us to propose that the cross-linking binding mode observed by crystallography might only occur in the crystal (Figure 4A, 4E). Indeed, only a chelate binding mode interpretation of 5u3 can explain the much stronger LecA binding of this divalent ligand compared to its monovalent analog GalTZ, and the absence of lectin aggregation in ultracentrifugation discussed above. It should be noted that 5u3 was significantly more mobile than 3u3 during MD in both the free and the LecA bound states, which might explain its weaker LecA binding. In the case of GalAG1 an MD simulation with free ligand was first carried out for 10 ns at 300K, which showed that the distance between the two galactose units could adopt a broad range of values below and above the optimal distance of 31 Å necessary for LecA binding (Figure 4A). A chelate type GalAG1.LecA complex was built from a conformer with appropriate spacing and relative position from the free MD. This chelate complex remained stable upon further MD for 15 ns, showing that this binding geometry was indeed possible (Figure 4F). In the chelate bound conformation the phenyl ring of the galactoside in GalAG1 was aligned with the shortest path between the two galactose binding sites similarly to the first triazole ring in 3u3. This orientation ACS Paragon Plus Environment
Page 8 of 19
Page 9 of 19
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
ACS Chemical Biology
9
was compatible with establishing favourable CH-π interactions with His50 as seen in the GalAG1.LecA crystal structure, although the CH-π interaction itself did not occur in MD since it is not parametrized in the Amber force-field. The key difference between the chelate bound model of GalAG1 and its monovalent binding mode observed by crystallography resided in the orientation of the peptide chain, which was incompatible with chelate binding to LecA in the crystal, suggesting that chelate binding might not be optimal with regards to secondary interactions with the peptidic linker. A chelate binding mode is nevertheless the best model to explain the much stronger binding of GalAG1 to LecA compared to GalAG0 and the lack of lectin aggregation by GalAG1. One may speculate that chelate binding has a strongly positive effect on binding affinity which more than compensates the non-optimal orientation of the peptide chain imposed when switching from monovalent to chelate binding. Alternatively the GalAG1.LecA complex might exist in equilibrium between monovalent bound states on adjacent LecA binding sites and a chelate binding mode representing a higher energy intermediate in the rebinding process. A model of the GalAG2.LecA complex was constructed starting from the last frame of the MD simulation of the GalAG1.LecA complex by extending the structure with the core tripeptide and an additional divalent GalAG1 dendrimer. This complex was stable during MD simulation at 300 K over 15 ns with an O4-O4 distance of 29.5 Å separating sugar 1 and sugar 2 (Supplementary Figure S2, see Figure 1 for sugar numbering). As for GalAG1 the CH-π interaction with His50 was not preserved but the orientation of the phenyl groups remained favorable for this interaction to take place. Within the GalAG2.LecA complex the two remaining galactosyl groups of the tetravalent ligand (Sugars 3 and 4) moved freely in solution covering a very broad range of O4-O4 distances also covering the value of 31 Å suitable for binding to a second LecA tetramer. An alternate model of the GalAG2.LecA complex using Sugar 1 and Sugar 4 separated by a longer linker was also obtained and found stable during MD, however the geometry of the phenyl galactoside was clearly less optimal than for chelation via the shorter G1 linker.
ACS Paragon Plus Environment
ACS Chemical Biology
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
10
In view of the higher entropy costs that would be associated with chelation via the longer linker, the simpler, GalAG1 derived model seemed more probable and was therefore selected to construct a larger, 2:1 LecA-GalAG2 complex as observed in AUC and representing the association of LecA with GalAG2 presumably leading to precipitation during crystallization attempts. Note that the binding stoichiometry determined by ITC shows that all four galactose units engage simultaneously in LecA binding, which is only possible if the dendrimer binds two different LecA tetramers in such a complex. This complex was simulated at 300 K over 10 ns and was stable during that period, with an average O4-O4 distance of 30.5 Å and 31.5 Å separating the two pairs of galactose residues, suggesting that this association is indeed possible without imposing unusual constraints on the dendrimer (Figure 4B/C/G). As for the other complexes the orientation of the phenyl groups remained favourable for the CH-π interaction with His50 to take place. The very strong binding of GalAG2 (KD = 2.5 nM), which is the tightest LecA ligand reported to date in terms of ITC parameters compared to other multivalent galactosides,3, 4, 9 probably results from the optimal utilization of all four galactosyl groups in a geometry compatible with establishing four His50 CH-π interactions.
ACS Paragon Plus Environment
Page 10 of 19
Page 11 of 19
A
B
3000
C 3000
3000
GalAG2-1-2 free GalAG2-1-2 bound
3u3 free 2500
2500
3u3 bound
2000
GalAG2-1-3 free GalAG2-1-3 bound GalAG2-1-4 free GalAG2-1-4 bound
1500
X-ray distance
5u3 free 2000
no. of frames
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
11
ACS Chemical Biology
5u3 bound GalAG1 free
1500
GalAG1 bound X-ray distance
1000 500 0
2500 2000 1500
1000
1000
500
500 0
0
0
10
20
30
d(O4 – O4), Å
40
GalAG2-2-3 free GalAG2-2-3 bound GalAG2-2-4 free GalAG2-2-4 bound GalAG2-3-4 free GalAG2-3-4 bound X-ray distance
0
20
40
60
d(O4 – O4), Å
0
20
40
60
d(O4 – O4), Å
D 3u3 F GalAG1
E 5u3
G GalAG2
Figure 4. MD simulation of galactoside - LecA complexes. A-C. Frequency histograms for the distance separating the galactose C(4)-OH oxygen atoms in pairs of galactosyl groups. Numbering of galactosyl groups in GalAG2 according to Figure 1. D-G. Images of the last MD frame of the corresponding complexes showing the ligands in yellow stick models and the electrostatic potential map of the LecA surface (+4.8 kcal = blue, -4.8 kcal = red). Each ligand was simulated for 10 ns in the free state and 15 ns in the LecA bound state using Amber12 (see methods for MD details).
ACS Paragon Plus Environment
ACS Chemical Biology
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
12
Biofilm inhibition The biofilm inhibition activity of the various ligands was determined to investigate a possible relationship between binding mode and biological activity. A modified version of the method described by Diggle et al. was employed.36 Pseudomonas aeruginosa biofilms were induced by growing the bacteria under limiting conditions in polystyrene microtiter plates, removing the culture medium, resuspending the attached biofilm and quantifying live cells using the WST-8 assay, which detects the cellular redox biochemical activity.32 Biofilm inhibition was measured by carrying out the incubation in the presence of various concentrations of ligands. The tetravalent C-fucosyl containing dendrimer FD2 was used as a positive control.37 As observed previously, dendrimer GalAG2 inhibited biofilm formation at a concentration of 20 µM to a level comparable to that of FD2, while the weaker binding divalent dendrimer GalAG1 only showed significant inhibition at 450 µM. Under the same conditions, both of the rigid rod spacer divalent galactosides 3u3 and 5u3 only showed very weak inhibition at 225 µM (Figure 5). These data suggest that tetravalency is an important requirement for inhibiting P. aeruginosa biofilms in galactoside type ligands of LecA. Considering that GalAG2 and 3u3 have comparable binding affinities to LecA, lectin aggregation rather than tight LecA binding might be essential to induce biofilm inhibition. It should be noted that a divalent galactosylated ligand similar to GalAG1 was recently reported to block cell invasion by P. aeruginosa,9 suggesting that the requirements for multivalent galactosides for inhibition of biofilms or of cell invasion might be quite different.
ACS Paragon Plus Environment
Page 12 of 19
Page 13 of 19
13
B
A 1.5
1.5
0.5
Abs. at 450 nm
Abs. at 450 nm
1
C
1
0.5
0
0
D 1.5
1
0.5
0
1
Abs. at 450 nm
Abs. at 450 nm
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
ACS Chemical Biology
0.5
0
Figure 5. Inhibition of P. aeruginosa biofilms by A: 3u3, B: 5u3, C: GalAG1 and D: GalAG2. Biofilms were grown on polystyrene microtiter plates inoculated with PA01 for 24 hours at 37°C in the presence of the indicated amount of compound, after which viable biofilms were stained (see methods for details). Wells incubated without compound (w/o) or with 27 µM glycopeptide dendrimer FD2 (positive control, (Fuc-α-CH2CO-Lys-Pro-Leu)4(Lys-Phe-Lys-Ile)2LysHis-Ile-NH2) were used as the corresponding controls. The net absorbance over background (growth media without bacteria) is given. All measurements were performed in triplicates. Data are mean ± SD.
Methods Procedures for ITC, ultracentrifugation, X-ray data collection and refinement statistics, modeling and computational methods, and biofilm inhibition are described in the supporting information. Accession codes. Coordinates and structure factors of refined LecA complexed with 3u3, 3u5, GalAG1 are available from the Protein Data Bank with accession codes 4YWA, 4YW7 and 4YW6 respectively.
ACS Paragon Plus Environment
ACS Chemical Biology
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
14
Conclusion The structural and functional studies presented above illuminate the binding modes of divalent and tetravalent galactosides to LecA. The chelate binding mode of the high affinity rigid-rod spacer divalent galactoside 3u3, which had been postulated on the basis of molecular modeling as the best hypothesis to explain its unusually strong binding, was confirmed by X-ray crystallography. The slightly longer digalactoside 5u3 gave an entirely well-resolved structure in its LecA complex, revealing an unexpected cross-linked binding mode. However the lack of lectin aggregation and the much tighter binding of 5u3 compared to monovalent galactosides suggested a chelate binding mode in solution, which was supported by MD simulation and by analytical ultracentrifugation. A similar situation occurred with the GalAG1.LecA complex, whose crystal structure only showed the terminal galactosylated dipeptide in a monovalent binding mode. Nevertheless a chelate binding mode explaining the absence of aggregation and the much stronger binding of GalAG1 compared to its monovalent analog seems more probable in solution and indeed was shown to be possible by MD simulation. A model of a chelate bound tetravalent peptide dendrimer GalAG2 was readily obtained starting with the GalAG1.LecA structure explaining the very tight binding of GalAG2 to LecA (KD = 2.5 nM) and the aggregation of LecA as observed by analytical ultracentrifugation. The much weaker biofilm inhibitory effect by divalent galactosides, including the tight binding 3u3, compared to GalAG2 suggests that biofilm inhibition requires the ability for lectin cross-linking as observed with GalAG2, and that multivalent glycoclusters such as dendrimers offer a unique opportunity for controlling P. aeruginosa biofilms. Nevertheless, a more optimal divalent ligand design might be possible capitalizing on the observed crystallographic structure of 3u3 and could lead to even tighter LecA binding divalent ligands, which might eventually also turn out as biofilm inhibitors. Further optimization of divalent and monovalent LecA ligands is currently underway.
ACS Paragon Plus Environment
Page 14 of 19
Page 15 of 19
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
ACS Chemical Biology
15
Acknowledgements. This work was supported financially by the University of Berne, the Swiss National Science Foundation, the COST Action CM1102 MultiGlycoNano, the Dutch Technology Foundation STW, applied science division of NWO, the Technology Program of the Ministry of Economic Affairs, and Labex ARCANE (ANR-11-LABX-003). We thank the staff at the Swiss Light Source, Beamline X06DA (PXIII), Villigen, Switzerland, for support during data collection. ASSOCIATED CONTENT Supporting Information Available. Methods and data for Ultracentrifugation, ITC integrated titration curves, X-ray data collection and refinement statistics, Details of computational methods used in this study. This material is available free of charge via the Internet at http://pubs.acs.org. AUTHOR INFORMATION Corresponding Author * Jean-Louis Reymond; Fax: (+41) 31 631 80 57 E-mail:
[email protected] Author Contributions: RV performed X-ray crystallography, MD simulations, interpreted data and wrote the paper, XJ performed MD simulations, GM and MB synthesized peptide dendrimers and performed biofilm inhibition measurements, FP, OF and AP synthesized divalent galactosides, TRB and DMETW performed AUC studies, JK performed modelling, EG performed ITC, AI supervised ITC and interpreted data, AS supervised and performed X-ray crystallography and interpreted data, TD supervised the study and interpreted data, RP designed and supervised the study, JLR designed and supervised the study, interpreted data and wrote the paper.
ACS Paragon Plus Environment
ACS Chemical Biology
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
16
References 1.
Cioci, G., Mitchell, E. P., Gautier, C., Wimmerova, M., Sudakevitz, D., Perez, S., GilboaGarber, N., and Imberty, A. (2003) Structural basis of calcium and galactose recognition by the lectin PA-IL of Pseudomonas aeruginosa, FEBS Lett 555, 297-301.
2.
Wagner, V. E., and Iglewski, B. H. (2008) P. aeruginosa Biofilms in CF Infection, Clin. Rev. Allergy Immunol. 35, 124-134.
3.
Cecioni, S., Lalor, R., Blanchard, B., Praly, J. P., Imberty, A., Matthews, S. E., and Vidal, S. (2009) Achieving high affinity towards a bacterial lectin through multivalent topological isomers of calix[4]arene glycoconjugates, Chem. Eur. J. 15, 13232-13240.
4.
Pertici, F., de Mol, N. J., Kemmink, J., and Pieters, R. J. (2013) Optimizing divalent inhibitors of Pseudomonas aeruginosa lectin LecA by using a rigid spacer, Chem. Eur. J. 19, 16923-16927.
5.
Reymond, J. L., Bergmann, M., and Darbre, T. (2013) Glycopeptide dendrimers as Pseudomonas aeruginosa biofilm inhibitors, Chem. Soc. Rev. 42, 4814-4822.
6.
Bernardi, A., Jimenez-Barbero, J., Casnati, A., De Castro, C., Darbre, T., Fieschi, F., Finne, J., Funken, H., Jaeger, K. E., Lahmann, M., Lindhorst, T. K., Marradi, M., Messner, P., Molinaro, A., Murphy, P. V., Nativi, C., Oscarson, S., Penades, S., Peri, F., Pieters, R. J., Renaudet, O., Reymond, J. L., Richichi, B., Rojo, J., Sansone, F., Schaffer, C., Turnbull, W. B., Velasco-Torrijos, T., Vidal, S., Vincent, S., Wennekes, T., Zuilhof, H., and Imberty, A. (2013) Multivalent glycoconjugates as anti-pathogenic agents, Chem. Soc. Rev. 42, 47094727.
7.
Boukerb, A. M., Rousset, A., Galanos, N., Mear, J. B., Thepaut, M., Grandjean, T., Gillon, E., Cecioni, S., Abderrahmen, C., Faure, K., Redelberger, D., Kipnis, E., Dessein, R., Havet, S., Darblade, B., Matthews, S. E., de Bentzmann, S., Guery, B., Cournoyer, B., Imberty, A., and Vidal, S. (2014) Antiadhesive properties of glycoclusters against Pseudomonas aeruginosa lung infection, J. Med. Chem. 57, 10275-10289.
8.
Smadhi, M., de Bentzmann, S., Imberty, A., Gingras, M., Abderrahim, R., and Goekjian, P. G. (2014) Expeditive synthesis of trithiotriazine-cored glycoclusters and inhibition of Pseudomonas aeruginosa biofilm formation, Beilstein J. Org. Chem. 10, 1981-1990.
9.
Novoa, A., Eierhoff, T., Topin, J., Varrot, A., Barluenga, S., Imberty, A., Romer, W., and Winssinger, N. (2014) A LecA ligand identified from a galactoside-conjugate array inhibits host cell invasion by Pseudomonas aeruginosa, Angew. Chem., Int. Ed. 53, 8885-8889.
10.
Sharma, R., Kottari, N., Chabre, Y. M., Abbassi, L., Shiao, T. C., and Roy, R. (2014) A highly versatile convergent/divergent "onion peel" synthetic strategy toward potent multivalent glycodendrimers, Chem. Commun. 50, 13300-13303.
11.
Kitov, P. I., Sadowska, J. M., Mulvey, G., Armstrong, G. D., Ling, H., Pannu, N. S., Read, R. J., and Bundle, D. R. (2000) Shiga-like toxins are neutralized by tailored multivalent carbohydrate ligands, Nature 403, 669-672.
12.
Kiessling, L. L., Gestwicki, J. E., and Strong, L. E. (2006) Synthetic multivalent ligands as probes of signal transduction, Angew. Chem., Int. Ed. Engl. 45, 2348-2368.
ACS Paragon Plus Environment
Page 16 of 19
Page 17 of 19
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
ACS Chemical Biology
17
13.
Wang, S. K., Liang, P. H., Astronomo, R. D., Hsu, T. L., Hsieh, S. L., Burton, D. R., and Wong, C. H. (2008) Targeting the carbohydrates on HIV-1: Interaction of oligomannose dendrons with human monoclonal antibody 2G12 and DC-SIGN, Proc. Natl. Acad. Sci. U. S. A. 105, 3690-3695.
14.
Martinez-Avila, O., Hijazi, K., Marradi, M., Clavel, C., Campion, C., Kelly, C., and Penades, S. (2009) Gold manno-glyconanoparticles: multivalent systems to block HIV-1 gp120 binding to the lectin DC-SIGN, Chem. Eur. J. 15, 9874-9888.
15.
Lindhorst, T. K., Bruegge, K., Fuchs, A., and Sperling, O. (2010) A bivalent glycopeptide to target two putative carbohydrate binding sites on FimH, Beilstein J. Org. Chem. 6, 801-809.
16.
Schwefel, D., Maierhofer, C., Beck, J. G., Seeberger, S., Diederichs, K., Moller, H. M., Welte, W., and Wittmann, V. (2010) Structural basis of multivalent binding to wheat germ agglutinin, J. Am. Chem. Soc. 132, 8704-8719.
17.
Sansone, F., and Casnati, A. (2013) Multivalent glycocalixarenes for recognition of biological macromolecules: glycocalyx mimics capable of multitasking, Chem. Soc. Rev. 42, 4623-4639.
18.
Martinez, A., Ortiz Mellet, C., and Garcia Fernandez, J. M. (2013) Cyclodextrin-based multivalent glycodisplays: covalent and supramolecular conjugates to assess carbohydrateprotein interactions, Chem. Soc. Rev. 42, 4746-4773.
19.
Murphy, P. V., Andre, S., and Gabius, H. J. (2013) The third dimension of reading the sugar code by lectins: design of glycoclusters with cyclic scaffolds as tools with the aim to define correlations between spatial presentation and activity, Molecules 18, 4026-4053.
20.
Chabre, Y. M., and Roy, R. (2013) Multivalent glycoconjugate syntheses and applications using aromatic scaffolds, Chem. Soc. Rev. 42, 4657-4708.
21.
Gingras, M., Chabre, Y. M., Roy, M., and Roy, R. (2013) How do multivalent glycodendrimers benefit from sulfur chemistry?, Chem. Soc. Rev. 42, 4823-4841.
22.
Dane, E. L., Ballok, A. E., O'Toole, G. A., and Grinstaff, M. W. (2014) Synthesis of Bioinspired Carbohydrate Amphiphiles that Promote and Inhibit Biofilms, Chem. Sci. 5, 551-557.
23.
Weber, T., Chandrasekaran, V., Stamer, I., Thygesen, M. B., Terfort, A., and Lindhorst, T. K. (2014) Switching of bacterial adhesion to a glycosylated surface by reversible reorientation of the carbohydrate ligand, Angew. Chem., Int. Ed. Engl. 53, 14583-14586.
24.
Zhang, S., Moussodia, R. O., Murzeau, C., Sun, H. J., Klein, M. L., Vertesy, S., Andre, S., Roy, R., Gabius, H. J., and Percec, V. (2015) Dissecting molecular aspects of cell interactions using glycodendrimersomes with programmable glycan presentation and engineered human lectins, Angew. Chem., Int. Ed. Engl. 54, 4036-4040.
25.
Lindhorst, T. K., and Dubber, M. (2015) Octopus glycosides: multivalent molecular platforms for testing carbohydrate recognition and bacterial adhesion, Carbohydr. Res. 403, 90-97.
26.
Cecioni, S., Imberty, A., and Vidal, S. (2015) Glycomimetics versus multivalent glycoconjugates for the design of high affinity lectin ligands, Chem. Rev. 115, 525-561.
ACS Paragon Plus Environment
ACS Chemical Biology
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
18
27.
Kadam, R. U., Bergmann, M., Hurley, M., Garg, D., Cacciarini, M., Swiderska, M. A., Nativi, C., Sattler, M., Smyth, A. R., Williams, P., Camara, M., Stocker, A., Darbre, T., and Reymond, J. L. (2011) A glycopeptide dendrimer inhibitor of the galactose-specific lectin LecA and of Pseudomonas aeruginosa biofilms, Angew. Chem., Int. Ed. Engl. 50, 1063110635.
28.
Kadam, R. U., Garg, D., Schwartz, J., Visini, R., Sattler, M., Stocker, A., Darbre, T., and Reymond, J. L. (2013) CH-pi "T-Shape" Interaction with Histidine Explains Binding of Aromatic Galactosides to Pseudomonas aeruginosa lectin LecA, ACS Chem. Biol. 8, 19251930.
29.
Garber, N., Guempel, U., Belz, A., Gilboa-Garber, N., and Doyle, R. J. (1992) On the specificity of the D-galactose-binding lectin (PA-I) of Pseudomonas aeruginosa and its strong binding to hydrophobic derivatives of D-galactose and thiogalactose, Biochim. Biophys. Acta 1116, 331-333.
30.
Chen, C. P., Song, S. C., Gilboa-Garber, N., Chang, K. S., and Wu, A. M. (1998) Studies on the binding site of the galactose-specific agglutinin PA-IL from Pseudomonas aeruginosa, Glycobiology 8, 7-16.
31.
Rodrigue, J., Ganne, G., Blanchard, B., Saucier, C., Giguere, D., Shiao, T. C., Varrot, A., Imberty, A., and Roy, R. (2013) Aromatic thioglycoside inhibitors against the virulence factor LecA from Pseudomonas aeruginosa, Org. Biomol. Chem. 11, 6906-6918.
32.
Kadam, R. U., Bergmann, M., Garg, D., Gabrieli, G., Stocker, A., Darbre, T., and Reymond, J.-L. (2013) Structure-Based Optimization of the Terminal Tripeptide in Glycopeptide Dendrimer Inhibitors of Pseudomonas aeruginosa Biofilms Targeting LecA, Chem. Eur. J. 18, 17054-17063.
33.
Pertici, F., and Pieters, R. J. (2012) Potent divalent inhibitors with rigid glucose click spacers for Pseudomonas aeruginosa lectin LecA, Chem. Commun. 48, 4008-4810.
34.
Pertici, F., Varga, N., van Duijn, A., Rey-Carrizo, M., Bernardi, A., and Pieters, R. J. (2013) Efficient synthesis of phenylene-ethynylene rods and their use as rigid spacers in divalent inhibitors, Beilstein J. Org. Chem. 9, 215-222.
35.
Case, D. A., Cheatham, T. E., Darden, T., Gohlke, H., Luo, R., Merz, K. M., Onufriev, A., Simmerling, C., Wang, B., and Woods, R. J. (2005) The Amber biomolecular simulation programs, J Comput Chem. 26, 1668-1688.
36.
Diggle, S. P., Stacey, R. E., Dodd, C., Camara, M., Williams, P., and Winzer, K. (2006) The galactophilic lectin, LecA, contributes to biofilm development in Pseudomonas aeruginosa, Environ. Microbiol. 8, 1095-1104.
37.
Johansson, E. M., Crusz, S. A., Kolomiets, E., Buts, L., Kadam, R. U., Cacciarini, M., Bartels, K. M., Diggle, S. P., Camara, M., Williams, P., Loris, R., Nativi, C., Rosenau, F., Jaeger, K. E., Darbre, T., and Reymond, J. L. (2008) Inhibition and dispersion of Pseudomonas aeruginosa biofilms by glycopeptide dendrimers targeting the fucose-specific lectin LecB, Chem. Biol. 15, 1249-1257.
ACS Paragon Plus Environment
Page 18 of 19
Page 19 of 19
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
ACS Chemical Biology
LecA
LecA GalAG2.LecA
ACS Paragon Plus Environment