Structure of human collapsin response mediator protein 1: a possible role of its C-terminal tail

Abstract

Collapsin response mediator protein 1 (CRMP-1) is the first identified member of the CRMP family and is crucial for both the mediation of neuronal differentiation and in suppressing the invasion of lung cancer. The crystal structure of full-length human CRMP-1 was determined at a resolution of 3 Å. Human CRMP-1 comprises a tetrameric assembly; its overall structure is similar to that of mouse CRMP-1, but the measured electron density of the C-terminal residues 488-496 show a randomly coiled link that connects the protomers to each other, within which residues 497-572 are proteolytically susceptible in vivo. Deletion of residues 472-572 by thrombin in vitro not only releases a randomly coiled tail but also transduces observable structural changes of CRMP-1, as revealed by analytical size-exclusive chromatography and circular dichroism spectra. These results indicate a possible alternative role in CRMP dynamics and function.

Keywords: circular dichoism; collapsin response mediator protein 1; crystal structure; lung cancer suppressor; non-small-cell lung cancer.

Figure 1

Crystallization and SDS–PAGE analysis of purified human CRMP-1. (a) Morphology of CRMP-1 crystals. Left, the crystals after crystallization (2 d, 293 K). Right, the crystals after incubation (20 d). (b) Analysis (SDS–PAGE) of CRMP-1 proteins that had not (lane 1) or had (lane 2) been incubated with thrombin protease and a sample taken from a crystallization well containing high-quality CRMP-1 crystals after incubation (20 d at 293 K; lane 3). Samples (3 µg) were characterized using a Tris–glycine gel (12%) and stained with Coomassie Blue. (c) SDS–PAGE analysis of the purified C-terminal tail (residues 472–572) of CRMP-1. Samples were separated with a Tris–tricine (20%) peptide-separation gel and were detected with silver staining.

Figure 2

A representative view of the final 2F _o − F _c electron-density map (blue, resolution 3 Å, contoured at the 1.0σ level), highlighting the fit of residues Arg467–Pro475 to the maps. The nonconsensus amino acid (depicted in red) in this region is Ala473 of human CRMP-1 (phenylalanine in mouse CRMP-1).

Figure 3

Crystal structure of human CRMP-1. (a) Ribbon diagram of a monomer. The N-terminal small domain is shown in blue and the α/β-barrel domain is shown in green/yellow. The long link and α-helix 19 are shown in violet. The view in the right panel has been rotated −90° with respect to that in the left panel. (b) A CRMP-1 tetramer. A chain, green; B chain, yellow; C chain, cyan; D chain, pink. Side chains significantly contributing to interfaces I and II are depicted as blue and red sticks, respectively. The view in the right panel has been rotated −90° with respect to that in the left panel.

Figure 4

Residues that are potentially important in determining interfaces in CRMP. The crystal structures of mouse CRMP-1 (yellow; PDB entry 1kcx; Deo et al., 2004 ▸), CRMP-2 (green; PDB entry 2gse; Stenmark et al., 2007 ▸), CRMP-4 (cyan; PDB entry 4bkn; Structural Genomics Consortium, unpublished work) and CRMP-5 (blue; PDB entry 4b91; Ponnusamy & Lohkamp, 2013 ▸) are superimposed onto human CRMP-1 (orange; PDB entry 4b3z). Both interfaces I (a) and II (b) contain conserved and nonconserved residues. The interacting residues are labelled for human CRMP-1; the corresponding residues of mouse CRMP-1, CRMP-2, CRMP-4 and CRMP-5 are given in parentheses.

Figure 5

Links located on the surface of the CRMP-1 tetramers. (a) A CRMP-1 tetramer shown as a surface diagram. In the tetramer structure, individual monomers are coloured green (A chain), yellow (B chain), cyan (C chain) and pink (D chain). The residues of the linkers (from Val488 to Arg496) are highlighted in dark pink (D chain to B chain), dark green (A chain to C chain), dark blue (C chain to A chain) and orange (B chain to D chain). The view in the right panel has been rotated 180° with respect to that in the left panel. (b) Detailed diagram of a portion of the link from the D chain to the B chain. Residues contributing to the cross-linking (Val488–Arg496 of the D chain and Glu221, Tyr290, Ser292 and Lys293 of the B chain) are depicted in stick representation and fitted to the 2F _o − F _c electron-density map (blue, resolution 3 Å, contoured at the 0.7σ level). Hydrogen bonds are shown as dashed green lines.

Figure 6

Oligomerization and characterization of the secondary structure of full-length and thrombin-cleaved CRMP-1. (a) Analytical gel-filtration chromatography of full-length (black line) and thrombin-cleaved (red line) CRMP-1 was performed as described in the Materials and methods ; the profiles are superimposed based on the ‘inject’ signal. The elution volumes corresponding to the molecular masses of the protein markers is marked with blue arrows for comparison. V _o and V _t denote the void and total volumes of the column, respectively. Under these operating conditions, the apparent predominant species of full-length and thrombin-cleaved CRMP-1 are tetramers. (b) CD spectra of CRMP-1 (black line) and thrombin-cleaved CRMP-1 (red line) were recorded in the 260–200 nm wavelength region. In the inset in (b), CD spectra of the C-terminal tail of CRMP-1 (amino acids 472–572; grey line), the α-helix-rich protein myoglobin (blue line) and the β-strand-rich protein concanavalin A (green line) (each protein was dissolved in 10 mM Tris–HCl pH 7.4, 150 mMNaF) are shown in mean residue ellipticity units; the buffer baseline was subtracted.