Journal Name
Dynamic Article Links ►
Cite this: DOI: 10.1039/c0xx00000x www.rsc.org/xxxxxx
ARTICLE TYPE
Chemical Communication, 2012, 48, 11886–11888.
5
Clicked Europium Dipicolinates Complexes for Proteins X–Ray Structural Determination.
10
15
20
This journal is © The Royal Society of Chemistry [year]
Journal Name, [year], [vol], 00–00 | 1
CREATED USING THE RSC COMMUNICATION TEMPLATE (VER. 3.0) - SEE WWW.RSC.ORG/ELECTRONICFILES FOR DETAILS
ARTICLE TYPE
www.rsc.org/xxxxxx | XXXXXXXX
Clicked Europium Dipicolinates Complexes for Proteins X–Ray Structural Determination. Romain Talona,b,c, Lionel Nautond,e, Jean–Louis Canetd, Richard Kahna,b,c , Eric Girard*a,b,c and Arnaud Gautier*d,e 5
10
15
20
25
30
35
40
45
50
55
Received (in XXX, XXX) Xth XXXXXXXXX 200X, Accepted Xth XXXXXXXXX 200X DOI: 10.1039/b000000x alkynes.9 Therefore, dpa were functionalized in two steps by propargylic, homo propargylic alcohols and a choline derivative to afford the ligands 4–6 in good yields (scheme 1, ESI).
Abstract. New trisdipicolinic acids lanthanide complexes are reported as phasing agent for X–ray crystallography of proteins. It is demonstrated that CuAAC modifications allow protein co– crystallization with low concentration in lanthanide complexes leading to an accurate structural determination. X–ray crystallography is of importance for structure determination, especially in the field of proteins. To obtain an atomic model of the protein of interest, it is necessary to determine both the amplitudes and the phases of the diffracted X– rays. In a diffraction experiment, the amplitude data are derived directly from measured intensities but phase information is lost and retrieving it is a complicated process.1 Therefore, methods such as multiple isomorphous replacement (MIR), single– wavelength anomalous diffraction (SAD) or multiwavelength anomalous diffraction (MAD) are required and involve the incorporation of heavy atom into the protein crystal.2 In this context, lanthanide ions are probably among the most appropriate elements due to their large anomalous signal.3 Lanthanides have been incorporated in proteins by exploiting native calcium– binding sites.4 However, the particular coordination sphere of lanthanide ions can generally not be fullfilled by proteins, limiting the use of lanthanide in protein crystallography. Various solutions, including mimics of calcium–binding sites5 or chelating agents either linked to the protein via a chemical modification6 or directly incorporated within the crystal, have been proposed.7 Therefore, regarding the latter method, several chelating agents have been proposed: dipicolinic acid 1 (dpa, scheme 1), HPDO3A, DO3A, DOTA, DTPABMA.7 In this context, some of us have described the properties of tris(dipicolinate)–lanthanide complexes [Ln(dpa)3]3- which are attractive for protein crystallography since their incorporation is controlled by electrostatic interactions (ie: guanidinium moieties of the arginine residues) and as their successful incorporation is easily detected through the luminescence of the europium or terbium complexes (antennae effect).7b,7c During the co– crystallization of [Ln(dpa)3]3- with hen egg–white lysozyme (HEWL), the former property led to a non native protein crystal form.7c However, these studies have been limited to the dipicolinic acid only. Therefore, we hypothesised that the structural variations on the pyridine dicarboxylate core would expand the applicability of dipicolinic acid complexes for protein crystallography. We thought that a variation of the lipophilicity, the adjunction of alcohol functions and/or the modulation of the global charge, will impact the binding mode of the complex which still keep the antennae effect and luminescence properties already observed. To achieve theses chemical modifications, Copper–catalyzed Azide–Alkyne Cycloaddition (CuAAC) affords an easy and fast entry.8 This is highlighted for the reactions of 4–azidodipicolinic ester 2 catalyzed by [Cu(SIMes)(4,7–dichloro–1,10–phenanthroline)Cl] with three 2 | Journal Name, [year], [vol], 00–00
R N3 O
OO
N OH OH 1: dpa
1) Alkynes, Cat. (5mol-%) 2) NaOH then HCl O
N OMe
85 to 95 %yield
2
O
4 R= CH2OH +
6 R= CH2N (Me)2CH2CH2OH O
R N N N
Eu N O O
Z
O
O 3 [Ln(dpa)3]
OH
5 R= CH2CH2OH
O O
N
O
N OH
O
N
O O
O
OMe
3
O
N N N
O
3-
O
O
N
O O
O O
7 R= CH2OH, z = -3 8 R= CH2CH2OH, z = -3 O 9 R= CH2N+(Me)2CH2CH2OH, z = 0
Eu N
N R
60
65
70
75
80
N N N
O O O
N O
N N R
Scheme 1. Structures of dpa and its CuAAC modified analogues.
Overall, the corresponding complexes possess two types of charges: -3 for 3, 7 and 8 and neutral for 9. Co–crystallization of the europium complexes were tested by using the hanging–drop method with two model proteins: HEWL and thaumatine from Thaumatococcus daniellii (TdTHAU). These proteins contain a positively charged binding site that was used to incorporate 3, but they differ in their hydrophobic nature: HEWL contains solvent– exposed hydrophobic tryptophane residues while TdTHAU is devoid of. Successful co–crystallization occurred with complexes 7–9 for HEWL and with 7, 8 for TdTHAU. For 7 and 8, X–ray exploitable crystals were obtained (complex concentration for HEWL: 2.7 mM, and for TdTHAU: 1.4 mM) at concentrations 18 to 35 fold lower than what is required with 3 (50–100 mM). This represents the lower concentration ever reported for the Ln(dpa)33- complexes. For the neutral complex 9, a constant formation of X–ray suitable crystals was observed over a concentration range from 2.7 to 12 mM.‡ The low concentrations required of 7–9 seem to indicate a strong affinity toward the proteins. It is noteworthy that HEWL co–crystallize with 7–9 in its native space group P43212 whereas 3 co–crystallize in the non native C2 space group.5b
This journal is © The Royal Society of Chemistry [year]
Table 1. Crystallization conditions. Complex HEWL 3[e] 7 8 9 TdTHAU 3[e] 7 8 9
[C][a]
Results[b]
Space group
n[c]
Res.
PDB Number
50–100 1.4–2.7 1.4–2.7 1.4–12
S S S S
C2 P 43212 P 43212 P 43212
5 2 1 1
1.54 1.35 1.51 1.21
2PC2 4BAD 4BAF 4BAP
50–100 1.4 1.4 1.4–12
S S S U
P 41212 P 41212 P 41212
1 1 1
1.46 1.30 1.20
2PE7 4BAL 4BAR
1) and the NεH of Arg73 (HEWL 2). The longest alcoholic tails in 8 allows the establishment of another H–bonding interaction with the NεH of Arg14 of HEWL 3.
[a]: mM. [b]: S successful, U unsuccessful. [c]: number of fixation sites. [d]: Resolution (Ǻ). [e]: taken from ref 7c.
5
10
For the different proteins, experimental anomalous–based phasing was performed as exemplified by complex 7 and HEWL (ESI). As observed for 3 the experimental phases led to an unambiguously interpretable electron density map as depicted in figure 1.7b,c Importantly, taking into account that a lower complex concentration in lanthanide conplexes was used (compared to 3), an excellent quality of the experimental phases was obtained in a straightforward manner leading to automatic model building, ended with more than 95 % of the final model built,.
40
45
50
55
15
20
25
30
35
Figure 2. Binding sites of 7 at the interface of HEWL 1 (brown) and HEWL 2 (blue).11
The complex 9 co–crystallized with HEWL into a single positively charged site with 62% of occupancy at the interface between two proteins (HEWL 1 and 2, Figure 3, Table 2). Its first ligand establishes parallel π–π interactions (3.5 Å) between the triazole and the phenol ring of Trp123 (HEWL 1) and electrostatic interactions between a dpa carboxylate function and the NζH2 of Lys33 (HEWL 1). For the second ligand, parallel hydrophobic π–π interaction (3.7 Å) between its triazole and the indol ring of Trp62 (HEWL 2) is found. The two carboxylates of the third ligand establish electrostatic interactions with both HEWL 1 and 2 via the NεH of Arg5 (HEWL 1) and NH1 of Arg73 (HEWL 2) respectively. The choline part is very flexible and can make direct or indirect interaction with HEWL 1 and 2, as observed with Glu7 and Lys1 of HEWL 1.
Figure 1. Experimental solvent flattened electron density map at 1.35 Å resolution of HEWL derivative crystal with complex 7 (contoured at 1.5 σ). Final refined model is superimposed.10
The binding mode of the complexes was elucidated in each of the refined structures. It is noteworthy that in all cases described below the complexes co–crystallize as a single enantiomer to fit within the space allowed by the chiral pocket of the crystallization site. Co–crystallization of complexes 7 and 8 with HEWL occured in a single binding site (70% and 55% of maximum occupancy for 7 and 8 respectively) located at the interface between two or three HEWLs (HEWL 1, 2 and 3), a second binding site was observed for compound 7, but only the highest occupied one allowed a precise description of the interactions (Figure 2, Table 2). For the first ligand, beside the classical arginine-based interactions (NζH2 of Lys33, HEWL 1) with the carboxylates moieties of the dpa, the triazole establishes perpendicular hydrophobic CH–π (d = 3.4 Ǻ) interactions with the phenol ring of tryptophan Trp123 of HEWL 1. The second ligand establishes a parallel hydrophobic π–π stacking interaction (d = 3.4 Ǻ) with the indol heterocycle of Trp62 of HEWL 2. The third ligand completes the binding through electrostatic interactions with the NεH of Arg5 (HEWL This journal is © The Royal Society of Chemistry [year]
Figure 3. Mode of interaction for 9 with HEWL.11
60
Complexes 7 and 8 co–crystallize with TdTHAU into a single binding site (48% and 31% of maximum occupancy respectively, Figure 4, Table 2). Complex 7 and 8 are located at the interface between two proteins (TdTHAU 1 and 2) and interact through electrostatic interactions between two dpa carboxylates of two ligands with the NεH of the respective Arg79 of TdTHAU 1 and 2.
Journal Name, [year], [vol], 00–00 | 3
e 35
Figure 4. Mode of interaction for 8 with TdTHAU.11 Table 2. Compiled interactions established (complexes 7–9) with HEWL and TdTHAU proteins. HEWL
Lig[a] L1
L2 L3
TdTHAU
5
10
15
20
L1 L2 L3
7[b] Tryp 123 (P 1) Lys 33 (P 1) Tryp 62 (P 2) Arg 5 (P 1) Arg 73 (P 2) Arg 79, (P 1) Arg 79, (P 2) N. I.[c]
8[b] Tryp 123 (P 1) Lys 33 (P 1) Tryp 62 (P 2) Arg 5 (P 1) Arg 73 (P 2) Arg 14 (P3) Arg 79, (P 2) Arg 79, (P 1) N. I.[c]
9[b] Tryp 123 (P 1) Lys 33 (P 1) Tryp 62 (P 2) Arg 5 (P 1) Arg 73 (P 2) N.C.[d] N.C. N.C.
[a]: Ligand., [b]: P = protein, [c]: N. I.: No direct interactions detected. [d]: N. C.: no crystallization.
Overall, while the fixation of the anionic trisdipicolinate complex 3 is essentially due to the electrostatic interactions with arginine residues, the chemical modifications of complexes 7–9 bring additional π–stacking interactions that allow co–crystallization in the native form at 18 to 35 fold lower concentrations. Whereas the two negatively charged complexes 7 and 8 are incorporated into both proteins, the neutral complex 9 presents a distinct picture. It is not incorporated into the positively charged site of TdTHAU but it is in HEWL which presents, in addition to the charges, solvent–exposed hydrophobic residues.
CNRS, UMR 6296, ICCF, F–63171 AUBIERE. Tel : +33 (0)4 73 40 76 46 ; Fax : +33 (0)4 73 40 77 17 ; E–mail : arnaud.gautier@ univ–bp clermont.fr. † Electronic Supplementary Information (ESI) available. 1 G. Taylor, Acta Crystallogr. D Biol. Crystallogr. 2003, 59, 1881– 1890. 2 (a) For MIR, see : D. W. Green, V. M. Ingram, and M. F. Perutz, Proc. R. Soc. Lond. A, 1954, 225, 287–307. (b) For MAD/SAD, see: W. A. Hendrickson, Acta Cryst, 1979, A35, 245–247. (c) For MAD/SAD, see: J. Karle, Int. J. Quantum Chem, 1980, 18, 357–367. 3 E. Girard, M. Stelter, J. Vicat, and R. Kahn, Acta Crystallogr. D Biol. Crystallogr., 2003, 59, 1914–1922. 4 R. Kahn, R. Fourme, R. Bosshard, M. Chiadmi, J. L. Risler, O. Dideberg, J. P. Wery, FEBS Lett. 1985, 179, 133–137. 5 N. R. Silvaggi, L. J. Martin, H. Schwalbe, B. Imperiali, K. N. Allen, J. Am. Chem. Soc. 2007, 129, 7114–7120. 6 M. D. Purdy, P. Ge, J. Chen, P. R. Selvin, M. C. Wiener, Acta Crystallogr D Biol Crystallogr. 2002, 58, 1111–1117. 7 For co–crystallization / soaking, using DO3A, DOTA, DOTMA and DTPA–BMA see (a) E. Girard, M. Stelter, P. L. Anelli, J. Vicat, and R. Kahn, Acta Crystallogr. D Biol. Crystallogr., 2003, 59, 1914– 1922.For dpa, see: (b): G. Pompidor, A. D’Aléo, J. Vicat, L. Toupet, N. Giraud, R. Kahn, and O. Maury, Angew. Chem. Int. Ed., 2008, 47, 3388–3391. (c) G. Pompidor, O. Maury, J. Vicat, R. Kahn Acta Crystallogr D Biol Crystallogr.2010, 66, 762–769. (d) R . Talon, R. Kahn, M. A. Durá, O. Maury, F. M. Vellieux, B. Franzetti, E. Girard J. Synchrotron Radiat. 2011, 18, 74–78. 8 (a) C. W. Tornøe and M. Meldal, in Peptides: The wave of the future, M. Lebl, R. A. Houghten, Eds.; Kluwer Academic Publishers, Dordrecht, 2001, pp. 263–264. (b) H. C. Kolb, M. G. Finn and K. B. Sharpless, Angew. Chem. Int. Ed., 2001, 40, 2004–2021. (c) M. Meldal and C. W. Tornoe Chem. Rev. 2008, 108, 2952–3015. 9 (a) M.–L. Teyssot, L. Nauton, J.–L. Canet, F. Cisnetti, A. Chevry and A. Gautier Eur. J. O. C., 2010, 3507–3515. (b) M.–L. Teyssot, A. Chevry, M. Traïkia, M. El–Ghozzi, D. Avignant and A. Gautier Chem. Eur. J., 2009, 15, 6322–6326. 10 Image generated by PyMOL. The PyMOL Molecular Graphics System, Version 1.5.0.4 Schrödinger, LLC. 11 Image generated by CHIMERA. E. F. Pettersen,T. D. Goddard, C. C. Huang, G. S. Couch, D. M. Greenblatt, E. C. Meng, T. E. Ferrin. J Comput Chem. 2004, 25, 1605–1612.
In conclusion, click chemistry modification affords a straightforward modulation of the dipicolinic core that can increase the affinity for the bending site using charge and hydrophobic interactions. We believe that such modular approach will increase the scope of the applicability of lanthanide dipicolinic acids in structural protein resolution.
Notes and references 25
30
a
CEA, DSV, Institut de Biologie Structurale (IBS), 41 rue Jules Horowitz, Grenoble F–38027, France. b CNRS, UMR 5075, Grenoble, France. c Université Joseph Fourier, Grenoble 1, F–38000, France. Tel : +33 (0)4 38 78 96 45 ; Fax : +33 (0)4 38 78 54 94 ; E–mail :
[email protected] d Clermont Université, Université Blaise Pascal, Institut de Chimie de Clermont-Ferrand, BP 10448, F–63000 CLERMONT-FERRAND Aubière, France.
4 | Journal Name, [year], [vol], 00–00
This journal is © The Royal Society of Chemistry [year]
