PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
 
J Mol Biol. 2009 December 11; 394(4): 613–620.
PMCID: PMC2971551

Segmented Helical Structure of the Neck Region of the Glycan-Binding Receptor DC-SIGNR

Abstract

Carbohydrate-recognition domains (CRDs) in the glycan-binding receptors DC-SIGN (dendritic-cell-specific intercellular adhesion molecule 1-grabbing nonintegrin; CD209) and DC-SIGNR (DC-SIGN-related receptor, also known as L-SIGN and variously designated CD209L and CD299) are projected from the membrane surface by extended neck domains containing multiple repeats of a largely conserved 23-amino-acid sequence motif. Crystals of a fragment of the neck domain of DC-SIGNR containing multiple repeats in which each molecule extends through multiple unit cells, such that the observed crystallographic asymmetric unit represents one repeat averaged over six repeats of the protein, have been obtained. The repeats are largely α-helical. Based on the structure and arrangement of the repeats in the crystal, the neck region can be described as a series of four-helix bundles connected by short, non-helical linkers. Combining the structure of the isolated neck domain with a previously determined overlapping structure of the distal end of the neck region with the CRDs attached provides a model of the almost-complete extracellular portion of the receptor. The results are consistent with previous characterization of the extended structure for the isolated neck region and the extracellular domain. The organization of the neck suggests how CRDs may be disposed differently in DC-SIGN compared with DC-SIGNR and in variant forms of DC-SIGNR assembled from polypeptides with different numbers of repeats in the neck domain.

Abbreviation used: CRD, carbohydrate-recognition domain
Keywords: DC-SIGNR, crystal structure, glycan-binding receptor, oligomerization, polymorphisms

DC-SIGN (dendritic-cell-specific intercellular adhesion molecule 1-grabbing nonintegrin; CD209) and DC-SIGNR (DC-SIGN-related receptor, also known as L-SIGN and variously designated CD209L and CD299) are glycan-binding receptors of the immune system.1 DC-SIGN is expressed on dendritic cells and some types of macrophages,2,3 while DC-SIGNR is expressed on sinusoidal endothelial cells and in the placenta.4 Similar to mannose-binding protein, the macrophage mannose receptor, and langerin, these receptors contain C-type carbohydrate-recognition domains (CRDs).5 Although each of these receptors binds mannose and related monosaccharides, they have differing specificities for oligosaccharides and for surfaces of microorganisms.1,6 Both DC-SIGN and DC-SIGNR bind high mannose oligosaccharides found on viral envelope glycoproteins and can participate in viral infection either directly by facilitating viral entry into a target cell or indirectly by presenting viruses to target cells.7–10 DC-SIGN also binds fucose-containing glycans, which can serve in recognition of parasites and in binding to T cells.11,12

The CRDs in DC-SIGN and DC-SIGNR tetramers are projected at the ends of extended neck regions formed from a series of conserved but not identical sequences of 23 amino acids.13 The neck portions of DC-SIGN and DC-SIGNR function as independent tetramerization domains.14 Circular dichroism analysis indicates that the necks contain extensive α-helical regions, and the sequences of the repeats show evidence of a heptad repeat, indicating a possible coiled-coil structure. However, each repeat also contains a conserved proline residue that would interrupt a helical structure, so the overall structure of the neck repeats has been unclear. Several lines of evidence suggest that the CRDs in DC-SIGN and DC-SIGNR are not held in a fixed position and that they can reorient to engage with ligands,15,16 but a more detailed understanding of the neck region is needed to explain its role in positioning the CRDs.

Common polymorphisms in the DC-SIGNR gene result in polypeptides with different numbers of neck repeats.17,18 Various genotypes at the DC-SIGNR locus result in DC-SIGNR homo- and hetero-oligomers with different polypeptide compositions in different individuals.19 Genetic studies indicate that these differences may correlate with susceptibility to infections.17,18 A better understanding of how length differences are accommodated in the structure of the DC-SIGNR neck would provide insight into their effects on disposition of ligand-binding sites and thus the ability of different forms of the receptor to interact with viruses and other pathogens.

Characterization and crystallographic analysis of a truncated neck domain

In order to examine the structure of the neck portion of DC-SIGNR, we investigated several truncated forms of the neck region as possible crystallization targets. In previous studies, the extracellular domain was truncated only at the C-terminus by replacing the CRD with a simple two-histidine tag, which allowed efficient binding of the tetrameric neck to chelated nickel columns through the resulting cluster of eight histidine residues.14 Further truncated forms were generated by removing the N-terminal non-repeat region or by removing this region and the first half of the first repeat unit (Fig. 1a). The latter truncation site was based on the presence of a protease-sensitive site immediately after Leu104 in the full-length neck, suggesting that the first part of the neck may be conformationally flexible when released from the membrane and in the absence of the normal N-glycosylation in this repeat.19 Gel filtration suggested that the purified truncated fragment is tetrameric, and differential scanning calorimetry demonstrated that it retains the stability of the full-length DC-SIGNR neck region, with a melting temperature of 81.7 °C compared with the measured value of 80.1 °C for the full-length neck (Fig. 1b and c). Thus, the N-terminal region that has been removed does not contribute significantly to the stability of the tetramer.

Fig. 1
(a) Summary of the organization of DC-SIGNR and the fragment expressed and crystallized along with the sequence of this fragment and the corresponding portion of DC-SIGN. Hydrophobic residues packing with one another in the four-α-helical bundles ...

Crystals of the truncated neck fragment were obtained following screening of a sparse matrix of crystallization conditions. After refining the crystallization conditions, a diffraction data set was measured (Table 1). The crystals are in space group P4212, with unit cell dimensions a = 34.2 Å, b = 34.2 Å, and c = 36.7 Å. The asymmetric unit of these crystals cannot contain the full protein used for crystallization. With only one repeat in the asymmetric unit, the calculated Matthews coefficient is 2.0, with a solvent content of 40%. Lower-symmetry space groups were considered, but both data statistics and the final solution indicated that the symmetry is in fact P4212. Several crystals were collected, washed with a solution similar to the mother liquor but with a higher concentration of polyethylene glycol, and analyzed by SDS-PAGE to confirm that the polypeptide was completely intact and not degraded during crystallization.

Table 1
Crystallographic data statistics

The structure was determined by multiwavelength anomalous dispersion phasing using crystals soaked in Pb(CH3COO)2 (Table 1). An anomalous-difference Patterson map showed strong peaks corresponding to a single heavy atom site. The site was found using the program Phenix22 and was refined in CNS23 to a figure of merit of 0.8. The phases were improved with density modification to give a map at 2.5-Å resolution that showed clear α-helices. A polyalanine model was built into the map using Coot,24 and rigid-body refinement against the remote data set was performed with CNS.

Within the unit cell, four molecules interact around the crystallographic 4-fold axis to generate a four-helix bundle (Fig. 2a). This bundle packs against a 2-fold-symmetry-related antiparallel four-helix bundle that also runs along the c axis (Fig. 2b). The four-helix bundles are translated by various numbers of repeats along the c axis throughout the crystal. This arrangement produces a unit cell with an asymmetric unit one repeat long that is an average of six nearly identical repeats.

Fig. 2
Structure of the 23-amino-acid repeat motif. (a) The 23-amino-acid model in the asymmetric unit shown with the final 2Fo − Fc electron density map (1.0σ contour). (b) Proposed packing of eight neck repeats in the crystals, ...

Simulated annealing of the partial polyalanine model with Phenix clarified the density for some missing residues. Side chains that are conserved among the six repeats were gradually added during several rounds of positional and isotropic temperature factor refinement alternating with manual model adjustment, increasing the resolution to 2.3 Å. This partial model was rigid-body refined against the native data set. Identical reflections were chosen as test set to calculate Rfree to be consistent with the heavy atom data. The model was further refined in Phenix to 2.2 Å. Completely conserved residues were given occupancy of 1.0, and side chains at the non-conserved positions were built with occupancy assigned based on their frequency in repeats 2 through 7 (Fig. 1a): proline with occupancy of 0.83 and serine with occupancy of 0.17 at position 1 as well as glutamate with occupancy of 0.17, glutamine with occupancy of 0.33, and arginine with occupancy of 0.5 at position 15, resulting in the final refined model described in Table 2.

Table 2
Crystallographic refinement statistics

The final model contains all 23 residues of the repeat and 21 water molecules. The electron density map did not show any features corresponding to repeat 8, which is the least conserved of the repeats. In addition, the 4-fold symmetry places the α-helices in close proximity, such that the large side chain of a phenylalanine residue at position 20 of repeat 8 (Phe261 in Fig. 1a) would not be accommodated. Both the lack of electron density for the nonconserved side chains in repeat 8 and the positions of the α-helices suggest that repeat 8 is disordered in the crystal or that diffraction from the nonconserved structure is obscured by the signal from the other six, almost identical, repeats. Thus, there may be gaps of one or more unit cells between successive tetramers along the c axis to accommodate a disordered repeat 8 and the extra residues at the N- and C-termini of the truncated protein.

Structure and linkage of neck repeats

The structure of the neck unit representing repeats 2 through 7 of the neck of DC-SIGNR is predominantly that of a four-helix bundle. The final model in the asymmetric unit is one repeat long (Fig. 2a) and, applying the 4-fold symmetry present in space group P4212, produces the four-α-helical bundle (Fig. 2c). The sequence of amino acids in the helical portions of the repeats follows the expected heptad pattern, with hydrophobic amino acids at the a and d positions (Fig. 1a). These hydrophobic residues are positioned toward the center of the bundle and pack against one another (Fig. 2c). There are also multiple hydrophobic, hydrogen-bond, and ionic interactions between side chains within each subunit and between subunits (Fig. 2d). The side chain of Lys17 is not well defined in the map, but it is possible that another rotamer of this residue would allow it to interact with Glu12 (Fig. 2d) to make the interaction proposed by Tabarani et al.25

Translation of the α-helical bundle produces a polymer of repeats in which residues Pro1 and Leu23 at the N- and C-termini are covalently linked to residues Leu23 and Pro1 from symmetry-related molecules (Fig. 2e). An α-helix of one repeat is connected to the α-helix of another repeat that is translated along the 4-fold axis and rotated 90° about the same axis. The side chain of Pro1 is positioned toward the center of the bundle, packing against the main chain of Pro1, Glu2 and Lys3, and the side chain of Leu6 from a symmetry-related monomer (Fig. 2e).

A model for the structure of the extended neck region was initially built by applying rotational symmetry operations and translations to the one repeat in the asymmetric unit, extending the four-helical bundle to represent repeats 2–8 (Fig. 3a). The neck is seen to adopt an elongated form, composed of α-helical repeats that form four-helical bundles that are separated by short kinks around the proline residues at position 1 of each repeat.

Fig. 3
Construction of a model of the full expressed fragment of the extracellular domain. (a) Seven four-helical bundles made by applying rotational and translational symmetry operations to the model in the asymmetric unit of the neck domain. (b) Top and side ...

Previous hydrodynamic studies were conducted on a fragment of DC-SIGN that included all of the repeats plus the less conserved 15-residue extension at the N-terminal end that links the repeat region to the membrane.14 Using the crystallographic data, we constructed an eight-repeat structure as a model for this fragment, lacking just 8 N-terminal residues and the C-terminal Gly–His–His tag. Calculation of the hydrodynamic properties of the modeled protein using HydroPro 7c26 gave values for the sedimentation coefficient of s20,w = 3.41 S and the diffusion coefficient of D20,w = 3.88 × 10− 7 cm2/s, in very close agreement with the measured values of s20,w = 3.37 S and D20,w = 3.80 × 10− 7 cm2/s. Thus, the model for the neck derived from the repeat structure in the crystal is consistent with the behavior of the isolated neck as a rigid, extended structure.

Structure of the extracellular domain of DC-SIGNR

A previously published structure of a fragment of DC-SIGNR [Protein Data Bank (PDB) ID 1XAR] comprises the C-terminal part of repeat 8 of the neck with the CRD in a tetrameric form (Fig. 3b). Comparison between the four-α-helical bundle of repeat 8 in this structure and the four-helical bundle seen here in the neck repeats reveals that the helices are positioned similarly up to residue Leu254 (extracellular domain numbering). Residues in positions 253 and 254 are present in both the tetramer-CRD and neck domain structures, and Cα superposition of the four copies of these two residues in the two structures gives an rmsd of 0.67 Å. In contrast, the positions of residues in the C-terminal portion of repeat 8, comprising residues 255–264, are slightly different in the tetramer-CRD structure and the neck repeats. The α-helices are splayed apart from one another, which breaks the 4-fold symmetry to form a dimer of dimers. This difference accommodates residue Phe261, corresponding to position 20 in repeat 8 (Figs. 1a and 3b) and places the CRDs in the orientations seen in the tetramer structure. Cα superposition of four copies of these eight residues in the two structures gives an rmsd of 1.53 Å.

A model of the almost-complete extracellular domain was generated by superimposing the Cα positions of the four copies of residue Leu254 from the tetramer-CRD onto those of the four-helix bundle of the neck. The final model comprises repeats 2–7 and the N-terminal portion of repeat 8 up to residue 253 taken from the neck structure as well as the C-terminal part of repeat 8 starting at residue 254 and the CRD as seen in the repeat 8–CRD tetramer structure (Fig. 3c). The side chains in the N-terminal portion of repeat 8 were changed to correspond to the natural sequence, with Asp243, Gln244, Gln247, and Gln249, at positions 2, 3, 6, and 8 of this repeat (Fig. 1a).

The structural model in Fig. 3c is consistent with other physical characterizations of the extracellular portion of DC-SIGNR. The model represents almost exactly the portion of the extracellular domain of DC-SIGNR that was previously characterized in hydrodynamic experiments, lacking only three N-terminal and three C-terminal residues.19 The program HydroPro 7c26 was used to predict values for the sedimentation coefficient of s20,w = 5.22 S and the diffusion coefficient of D20,w = 3.37 × 10− 7 cm2/s, which compare well with the measured values of s20,w = 5.36 S and D20,w = 3.45 × 10− 7 cm2/s. The fact that the measured values are slightly higher than the predicted values may reflect the fact that the hydrodynamic modeling program is based on a rigid molecule, while in fact the CRDs are not fully fixed in position. The presence of breaks in the α-helical structure, resulting in relatively short helices of 18 residues abutting non-helical regions at each end, means that there is a much larger end effect in measurements of circular dichroism than would be expected for a continuously helical neck.27 Thus, the segmented nature of the helices accounts for the low estimate of 40% helical content obtained using global fitting algorithms.14

Force–distance measurements have previously been used to study a version of the extracellular domain of DC-SIGN corresponding to the fragment of DC-SIGNR modeled in Fig. 3c with an N-terminal His6 tag for immobilization on a membrane surface.16 The results indicate that the protein extends 328 Å in the absence of interaction with ligand and that it can resist compression by an opposing membrane surface. The rigidity of the structure is consistent with the structure of the neck in DC-SIGNR, in which the four polypeptides interact extensively throughout their length, making the tetramer resistant to bending. However, when the CRDs engage with ligands on an apposed surface, there is a conformational change, probably due to reorientation of the CRDs, which results in a reduction of the overall length to approximately 280 Å. Parallel studies indicate that the extracellular domain of DC-SIGNR has similar dimensions and undergoes similar conformational changes (S. Menon, M. E. Taylor, K. Drickamer, and D. A. Leckband, unpublished results). The overall length of the model in Fig. 3c is 265–270 Å, which suggests that the disposition of the CRDs in the model resembles that of the structure engaged with ligand. A model for the DC-SIGNR tetramer with bound oligosaccharide ligand was obtained by superimposing the crystal structure of the CRD complexed with Man3GlcNAc26 (PDB ID 1K9J) onto the CRDs in the model in Fig. 3c and then superimposing the trimannose core of Man9GlcNAc2 onto the equivalent part of Man3GlcNAc2 (Fig. 3d). The resulting model shows that the reducing ends of the oligosaccharides would be oriented upward, consistent with the proposed arrangement when interacting with glycans on a membrane surface.

Modeling based on recent small-angle X-ray scattering measurements of the DC-SIGN extracellular domain is also roughly consistent with the overall length of the molecule, but the molecular envelope derived from the scattering data suggested an alternative rigid positioning of the CRDs in a more extended structure.25 Such an extended structure could be stabilized by anomalous inter-CRD disulfide bonds that are present when the protein from the bacterial expression system is prepared following the published procedure13 unless subjected to additional steps of purification (C. Tso, M. E. Taylor, and K. Drickamer, unpublished observations). Nevertheless, the surface force measurements of DC-SIGN and DC-SIGNR are consistent with the idea that there exists a more extended conformation of the extracellular domain that can change to the conformation seen in the crystals. In the light of these considerations, it would be interesting to assess how well the crystallographically observed neck structure would fit the scattering envelope.

Variations in DC-SIGN and DC-SIGNR structures

Although DC-SIGN and DC-SIGNR are very similar in amino acid sequence and polypeptide organization, they do display important differences in the neck regions as well as in the CRDs. Some of the differences in the neck region are in the heptad residues of the helical regions, particularly at position 6 in the repeat. The presence of glutamine at position 6 in repeats 6 and 7 in the neck of DC-SIGN, rather than leucine in DC-SIGNR, correlates with the higher stability of the DC-SIGNR neck in thermal denaturation studies. Chimeric necks have the stability properties of DC-SIGNR when they contain leucine at these positions but behave similarly to DC-SIGN when glutamine is present.14 A further key difference is in the final neck repeat, in which the phenylalanine residue at position 20 of the repeat, corresponding to residue 261 in the overall sequence of DC-SIGNR, is replaced by a valine residue in DC-SIGN. The smaller size of the valine side chain means that the helices of the final repeat could be packed similarly to those in the remainder of the neck and the 4-fold symmetry could extend all the way to the end of the molecule. This difference could shift the orientation of the CRDs slightly upward and change the relative disposition of the glycan-binding sites, and it might also explain the difference between the model of DC-SIGNR derived here and that of DC-SIGN in the work of Tabarani et al.25

The structure of the neck of DC-SIGN provides some important insights into how neck length polymorphisms in the DC-SIGNR gene affect the structure of the protein and how these might in turn affect interaction with ligands, such as viral membrane glycoproteins. First, the slight splaying apart of the C-terminal portion of the helical region in the final neck repeat places the CRDs somewhat to the side of rather than at the ends of the helices. As a result, in hetero-oligomers containing a mixture of polypeptides with different numbers of neck repeats, CRDs on the shorter polypeptides would project out from the side of the neck and would thus be accommodated adjacent to a further neck repeat in the longer polypeptide (Fig. 3e). Second, one of the most striking features of the model in Fig. 3e is the twisting of the neck, resulting from the pitch of the supercoiling in the segments of the four-helical bundle and the way that the proline linkers join these segments. As a consequence of this arrangement, the position of the C-terminal end of each repeat is rotated 90° from the previous repeat along the polypeptide moving away from the membrane surface. Therefore, in a hetero-oligomer, the position of a CRD on a shorter polypeptide would be rotated 90° around the neck as well as translated one repeat length closer to the membrane and would thus be below the CRD in an adjacent, longer repeat rather than next to it. The resulting difference in the orientation of the glycan-binding sites could dramatically affect the relative affinities of homo-oligomers and hetero-oligomers for surface glycoproteins on different enveloped viruses.

PDB accession number

Coordinates and structure factors have been deposited in the PDB with access number 3JQH.

Acknowledgements

This work was supported by U.S. National Institutes of Health grant GM50565 and Wellcome Trust grant 075565.

Notes

Edited by I. Wilson

References

1. van Kooyk Y., Rabinovich G.A. Protein–glycan interactions in the control of innate and adaptive immune responses. Nat. Immunol. 2008;9:593–601. [PubMed]
2. Geijtenbeek T.B.H., Torensma R., van Vliet S.J., van Duijnhoven G.C.F., Adema G.J., van Kooyk Y. Identification of DC-SIGN, a novel dendritic cell-specific ICAM-3 receptor that supports primary immune responses. Cell. 2000;100:575–585. [PubMed]
3. Tailleux L., Pham-Thi N., Bergeron-Lafaurie A., Herrmann J.-L., Charles P., Schwartz O. DC-SIGN induction in alveolar macrophages defines privileged target host cells for mycobacteria in patients with tuberculosis. PLoS Med. 2005;e381:2. [PMC free article] [PubMed]
4. Soilleux E.J., Morris L.S., Rushbrook S., Lee B., Coleman N. Expression of human immunodeficiency virus (HIV)-binding lectin DC-SIGNR: consequences for HIV infection and immunity. Hum. Pathol. 2002;33:652–659. [PubMed]
5. Weis W.I., Taylor M.E., Drickamer K. The C-type lectin superfamily in the immune system. Immunol. Rev. 1998;163:19–34. [PubMed]
6. Guo Y., Feinberg H., Conroy E., Mitchell D.A., Alvarez R., Blixt O. Structural basis for distinct ligand-binding and targeting properties of the receptors DC-SIGN and DC-SIGNR. Nat. Struct. Mol. Biol. 2004;11:591–598. [PubMed]
7. Alvarez C.P., Lasala F., Carrillo J., Muñiz O., Corbí A.L., Delgado R. C-type lectins DC-SIGN and L-SIGN mediate cellular entry by Ebola virus in cis and trans. J. Virol. 2002;76:6841–6844. [PMC free article] [PubMed]
8. Simmons G., Reeves J.D., Grogan C.C., Vandenberghe L.H., Baribaur F., Whitbeck J.C. DC-SIGN and DC-SIGNR bind Ebola glycoproteins and enhance infection of macrophages and endothelial cells. Virology. 2003;305:115–123. [PubMed]
9. Geijtenbeek T.B.H., Kwon S.S., Torensma R., van Vliet S.J., van Duijnhoven G.C.F., Middel J. DC-SIGN, a dendritic cell-specific HIV-1-binding protein that enhances trans-infection of T cells. Cell. 2000;100:587–597. [PubMed]
10. Pöhlmann S., Soilleux E.J., Baribaud F., Leslie G.J., Morris L.S., Trowsdale J. DC-SIGNR, a DC-SIGN homologue expressed in endothelial cells, binds to human and simian immunodeficiency viruses and activates infection in trans. Proc. Natl Acad. Sci. USA. 2002;98:2670–2675. [PubMed]
11. Appelmelk B.J., van Die I., van Vliet S.J., Vandenbroucke-Grauls M.J.E., Geijtenbeek T.B.H., van Kooyk Y. Carbohydrate profiling identifies new pathogens that interact with dendritic cell-specific ICAM-3-grabbing nonintegrin on dendritic cells. J. Immunol. 2003;170:1635–1639. [PubMed]
12. Bogoevska V., Horst A., Klampe B., Lucka L., Wagener C., Nollau P. CEACAM1, an adhesion molecule of human granulocytes, is fucosylated by fucosyltransferase IX and interacts with DC-SIGN of dendritic cells via Lewis x residues. Glycobiology. 2006;16:197–209. [PubMed]
13. Mitchell D.A., Fadden A.J., Drickamer K. A novel mechanism of carbohydrate recognition by the C-type lectins DC-SIGN and DC-SIGNR: subunit organisation and binding to multivalent ligands. J. Biol. Chem. 2001;276:28939–28945. [PubMed]
14. Yu Q.D., Oldring A.P., Powlesland A.S., Tso C.K., Yang C., Drickamer K. Autonomous tetramerization domains in the glycan-binding receptors DC-SIGN and DC-SIGNR. J. Mol. Biol. 2009;387:1075–1080. [PMC free article] [PubMed]
15. Feinberg H., Guo Y., Mitchell D.A., Drickamer K., Weis W.I. Extended neck regions stabilize tetramers of the receptors DC-SIGN and DC-SIGNR. J. Biol. Chem. 2005;280:1327–1335. [PubMed]
16. Menon S., Rosenberg K., Graham S.A., Ward E.M., Taylor M.E., Drickamer K. Binding site geometry and flexibility in DC-SIGN demonstrated with surface force measurements. Proc. Natl Acad. Sci. USA. 2009;106:11524–11529. [PubMed]
17. Khoo U.S., Chan K.Y., Chan V.S., Lin C.L. DC-SIGN and L-SIGN: the SIGNs for infection. J. Mol. Med. 2008;86:861–874. [PubMed]
18. Li H., Tang N.L., Chan P.K., Wang C.Y., Hui D.S., Luk C. Polymorphisms in the C-type lectin genes cluster in chromosome 19 and predisposition to severe acute respiratory syndrome coronavirus (SARS-CoV) infection. J. Med. Genet. 2008;45:752–758. [PubMed]
19. Guo Y., Atkinson C.E., Taylor M.E., Drickamer K. All but the shortest polymorphic forms of the viral receptor DC-SIGNR assemble into stable homo- and hetero-tetramers. J. Biol. Chem. 2006;281:16794–16798. [PubMed]
20. Project C.C. The CCP4 suite: programs for protein crystallography. Acta Crystallogr., Sect. D: Biol. Crystallogr. 1994;50:760–763. [PubMed]
21. Otwinowski Z., Minor W. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 1997;276:307–326.
22. Adams P.D., Gopal K., Grosse-Kunstleve R.W., Hung L.-W., Ioerger T.R., McCoy A.J. Recent developments in the PHENIX software for automated crystallographic structure determination. J. Synchrotron Radiat. 2004;11:53–55. [PubMed]
23. Brunger A.T., Adams P.D., Clore G.M., Delano W.L., Gros P., Grosse-Kunstleve R.W. Crystallography and NMR System (CNS): a new software system for macromolecular structure determination. Acta Crystallogr., Sect. D: Biol. Crystallogr. 1998;54:905–921. [PubMed]
24. Emsley P., Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2004;60:2126–2132. [PubMed]
25. Tabarani G., Thèpaut M., Stroebel D., Ebel C., Vivès C., Vachette P. DC-SIGN neck domain is a pH-sensor controlling oligomerization: SAXS and hydrodynamic studies of extracellular domain. J. Biol. Chem. 2009;284:21229–21240. [PubMed]
26. Garcia de la Torre J., Huertas M.L., Carrasco B. Calculation of hydrodynamic properties of globular proteins from their atomic-level structure. Biophys. J. 2000;78:719–730. [PubMed]
27. Engel M., Williams R.W., Erickson B.W. Designed coiled-coil proteins: synthesis and spectroscopy of two 78-residue α-helical dimers. Biochemistry. 1991;30:3161–3169. [PubMed]