Crystal Structure of Keratin 1/10(C401A) 2B Heterodimer Demonstrates a Proclivity for the C-Terminus of Helix 2B to Form Higher Order Molecular Contacts

Abstract

We previously determined the crystal structure of the wild-type keratin 1/10 helix 2B heterodimer at 3.3 Å resolution. We proposed that the resolution of the diffraction data was limited due to the crystal packing effect from keratin 10 (K10) residue Cys401. Cys401^K10 formed a disulfide-linkage with Cys401 from another K1/10 heterodimer, creating an "X-shaped" structure and a loose crystal packing arrangement. We hypothesized that mutation of Cys401^K10 to alanine would eliminate the disulfide-linkage and improve crystal packing thereby increasing resolution of diffraction and enabling a more accurate side chain electron density map. Indeed, when a K10 Cys401Ala 2B mutant was paired with its native keratin 1 (K1) 2B heterodimer partner its x-ray crystal structure was determined at 2.07 Å resolution; the structure does not contain a disulfide linkage. Superposition of the K1/K10(Cys401Ala) 2B structure onto the wild-type K1/10 2B heterodimer structure had a root-mean-square-deviation of 1.88 Å; the variability in the atomic positions reflects the dynamic motion expected in this filamentous coiled-coil complex. The electrostatic, hydrophobic, and contour features of the molecular surface are similar to the lower resolution wild-type structure. We postulated that elimination of the disulfide linkage in the K1/K10(Cys401Ala) 2B structure could allow for the 2B heterodimers to bind/pack in the A₂₂ tetramer configuration associated with mature keratin intermediate filament assembly. Analysis of the crystal packing revealed a half-staggered anti-parallel tetrameric complex of 2B heterodimers; however, their register is not consistent with models of the A₂₂ mode of tetrameric alignment or prior biochemical cross-linking studies.

Keywords: assembly; disulfide bond; heterodimer; intermediate filament; keratin; structure.

Figure 1

Intermediate filaments and the structural biology of keratin 1/10 complex. (a) The six major types of intermediate filaments, with the area of each pie slice representative of the total number of IFs for that type. (b) Common structural organization of IFs using K1 and K10 as an example, with the head, tail, and four helical domains denoted. Protein Data Bank (PDB) ID codes for determined K1/K10 crystal structures are accentuated (orange). (c) Proposed paradigm for correlating genotype, structurotype, and phenotype for clinical diseases. An example using the K1^S233L mutation causing epidermolytic palmoplantar keratoderma (EPPK) is shown. L233 creates aberrant hydrophobicity (orange) on the molecular surface of K1^S233L/K10-1B (PDB ID 6E2J). (d) Crystal structure of the wild-type K1/K10-2B heterodimer (PDB ID 4ZRY) depicting that a K10^C401-mediated disulfide bond (yellow) occurs between K10 helices on separate heterodimers.

Figure 2

Knob-pocket mechanism in helix 1B is important for A₁₁ tetramer formation for multiple IF types. (a) For orientation, a schematic depicts a keratin A₁₁ tetramer (upper right) and shows the position of symmetric knob-pocket interactions in the 1B helical region of the Type II keratin (blue). Below, the crystal structures of K1/K10-1B (blue, PDB ID 6EC0), vimentin-1B (gold, PDB ID 3UF1), and lamin A/C-1B (green, PDB ID 6JLB) were superposed and the knob-pocket interaction presented. There is structural conservation of the 1B knob-pocket tetramer assembly mechanism for Type II, Type III, and Type V IFs. There are no structures to date of Type IV or VI IFs. (b) Multiple sequence alignment for the knob and pocket regions of select Type II, Type III, and Type V IFs. There is complete conservation of the key phenylalanine residue in the knob. There is also a high degree of conservation of the pocket residues between Type III and Type V IFs (red boxes), which can be compared to the yellow bars for keratin 1. The structural superposition in panel (a) demonstrates the pocket is structurally conserved despite there being a variation in how the pocket is formed between heterodimeric IFs (Type I/II) and homodimeric IFs (Types III and V). The fact that heterodimeric and homodimeric IFs both utilize the 1B knob-pocket mechanism reinforces its importance to IF assembly.

Figure 3

Light scattering demonstrates K1/K10^C401A-2B is a dimer in solution. (a) ^HisK10^C401A-2B bound untagged K1-2B in the bacterial lysate in a 1:1 stoichiometry and purified as a 1:1 complex using nickel affinity chromatography. After thrombin cleavage of the N-terminal His₆ tag, K10^C401A-2B migrated on the polyacrylamide gel at the same molecular weight (MW) as K1-2B, as expected. The final purified K1/K10^C401A-2B complex after gel filtration is shown (gel lane farthest right). (b) Wild-type K1/K10-2B heterocomplex was primarily a heterodimer in solution (Peak 2), but a small fraction of a tetrameric species was observed (Peak 1) by multi-angle light scattering (MALS). (c) MALS of K1/K10^C401A-2B produced a single heterodimer peak without evidence for a tetrameric species, indicating substitution of C401 with alanine abrogates disulfide-mediated tetramer formation.

Figure 4

Human K1/K10^C401A-2B heterodimer crystal structure. (a) Representative crystal of the K1/K10^C401A-2B complex. (b) Superposition of the K1/K10^C401A-2B heterodimer onto the wild-type K1/K10-2B heterodimer (PDB ID 4ZRY) had a root-mean-square deviation (RMSD) of 1.88 Å across all atom pairs. There is enhanced curvature at the C-terminus of K1-2B in the K1/K10^C401A-2B heterodimer (light blue) compared to the wild-type structure. (c) A zoomed in view of the superposed structures from (b) at K10 position 401 demonstrated electron density for an alanine, confirming the K10^C401A substitution. There is no structural evidence for a disulfide bond in the K1/K10^C401A-2B heterodimer.

Figure 5

Comparison of the crystal lattice packing for wild-type K1/K10-2B and K1/K10^C401A-2B structures. (a) The crystal lattice for the wild-type K1/K10-2B heterodimer (PDB ID 4ZRY) demonstrated an “X-shaped” structure due to a transdimer disulfide bond between K10^C401 residues on adjacent heterodimers. Two examples of the X-shaped tetrameric species are colored (K1-2B, dark blue; K10-2B, magenta). (b) The crystal lattice for K1/K10^C401A-2B did not contain a disulfide-linked X-shaped structure. It did contain an octameric assembly, which is colored (K1-2B, light blue; K10^C401A-2B, pink). (c) The octameric K1/K10^C401A-2B assembly from the crystal lattice is shown separately for clearer viewing; its contacts are mediated largely by the C-terminal aspect of helix 2B.

Figure 6

Residue interactions contributing to the K1/K10^C401A-2B octameric assembly. (a) Within the K1/K10^C401A-2B octameric assembly, one tetrameric interface is defined by half-staggered anti-parallel 2B heterodimers (top, green). The molecular surface for the half-staggered anti-parallel 2B tetramer is shown in pink (K10^C401A) and light blue (K1-2B). Specific amino acid interactions by the C-terminus of K1-2B that are critical for tetramer stabilization are depicted (lower half). (b) A second tetrameric interface within the K1/K10^C401A-2B octameric assembly occurs exclusively between the splayed C-terminal ends of two 2B heterodimers: hence, a “splayed C-terminal tetrameric interface” (top, green). Its molecular surface is shown in pink (K10^C401A) and light blue (K1-2B). Specific amino acid interactions by the C-termini of K1-2B and K10^C401A that are critical for tetramer stabilization are depicted, including important salt bridges (lower half). (c) Partial, aligned, amino acid sequences for the C-terminal half of helix 2B for K1 and K10^C401A. Residues participating in contacts < 5 Å in either the half-staggered or splayed tetrameric interface are colored light blue (K1) or pink (K10^C401A) on the sequence. There is a clustering of residues involved in important contacts at the C-terminus of K1-2B and the K10-2B region just distal to the C401A substitution site (bold red “A”).

Figure 7

Molecular surface analysis of K1/K10^C401A-2B half-staggered tetramer. (a) Within the K1/K10^C401A-2B octameric assembly, two half-staggered anti-parallel tetramers make a contact at a K10-K10’ interface. The K10 residues involved in this interface are depicted structurally (inset, pink) and mapped onto the amino acid sequence (left). One K10 interface (391-407) occurs around the C401A site (red “A”; “*” denotes the structural location of K10^C401A whose side chain faces away from the viewer). The other K10 interface (440-456) occurs at the C-terminus of K10-2B. (b) Molecular surface of the K1/K10^C401A-2B half-staggered anti-parallel tetramer colored by hydrophobic potential (orange is more hydrophobic). Surface-exposed hydrophobic residues are clustered at the N-terminus of the 2B heterodimer (inset, left). Other surface-exposed hydrophobic residues identified in the previous wild-type structure are buried in the symmetric C-terminal binding interface, which contains three adjacent cavities in the center of the tetramer (panel b, bottom). Two N-Octyl-β-D-glucoside molecules are bound in these cavities. (c) Molecular surface of the K1/K10^C401A-2B half-staggered anti-parallel tetramer colored for electrostatic potential (red, acidic; blue, basic). The symmetric C-terminal binding interface at the center of the half-staggered anti-parallel tetramer is highly acidic. (d) 107 bound waters (red spheres) were modeled in the K1/K10^C401A-2B heterodimer structure.

Figure 8

The K1/K10^C401A-2B structure does not capture the A₂₂ tetrameric alignment. (a) Schematic of the proposed A₂₂ tetramer alignment based on prior cross-linking and nearest neighbor analysis. The two K1/K10 heterodimers have their 2B regions in-phase. The knob (triangle) and pocket (partial circle) structural elements identified in the 1B region are depicted. (b) Six cross-linked mouse K1- and mouse K10-2B tryptic peptides were identified in prior biochemical cross-linking studies on mature K1/K10 IFs (left). The mouse peptide sequences are translated into their corresponding human sequences (right), demonstrating many of the lysines involved in the mouse keratin cross-links are not conserved in humans. (c) The human residues corresponding to the key lysine residues involved in mouse cross-links are mapped onto the K1/K10^C401A-2B half-staggered anti-parallel tetramer structure. (d) The amino acid sequence for seventeen residues at the C-terminus of helix 2B for K1/K10 and K5/K14 are illustrated. Five human skin diseases map to mutations in this “mutational hotspot” in keratins 1/10 and 5/14: epidermolytic ichthyosis (EI; formerly known as epidermolytic hyperkeratosis or bullous congenital ichthyosiform erythroderma), cyclic ichthyosis with epidermolytic hyperkeratosis (CIEH), epidermolytic palmoplantar keratoderma (EPPK), epidermolysis bullosa simplex (EBS, including Koebner, Weber-Cockayne, and Dowling-Meara types), and recessive epidermolysis bullosa simplex (REBS). Residues that are mutated in one of these skin diseases are enclosed in a red box. Colored lines connect the residue with the appropriate disease listed in the center based upon keratin mutations documented online in the Human Intermediate Filament Database.