Structures of the spectrin-ankyrin interaction binding domains

Abstract

As key components of the erythrocyte membrane skeleton, spectrin and ankyrin specifically interact to tether the spectrin cytoskeleton to the cell membrane. The structure of the spectrin binding domain of ankyrin and the ankyrin binding domain of spectrin have been solved to elucidate the structural basis for ankyrin-spectrin recognition. The structure of repeats 14 and 15 of spectrin shows that these repeats are similar to all other spectrin repeats. One feature that could account for the preference of ankyrin for these repeats is the presence of a conserved, negatively charged patch on one side of repeat 14. The structure of the ankyrin ZU5 domain shows a novel structure containing a beta core. The structure reveals that the canonical ZU5 consensus sequence is likely to be missing an important region that codes for a beta strand that forms part of the core of the domain. In addition, a positively charged region is suggestive of a binding surface for the negatively charged spectrin repeat 14. Previously reported mutants of ankyrin that map to this region lie mostly on the surface of the protein, although at least one is likely to be part of the core.

Figure 1

Schematic diagrams of erythrocyte β-spectrin and ankyrin. The schematic diagrams show the domain organization of the 2 molecules. (A) The intact human erythroid β-spectrin (HEβ) molecule consists mainly of tandem repeats of 3-helix bundles, each bundle known as a spectrin repeats (numbered from N to C terminus). The ankyrin binding domain comprises repeats 14 and 15 (light blue).^,, Additional binding domains include 2 N-terminal calponin homology (CH) domains responsible for actin binding (orange), spectrin repeats 1 and 2, which mediate α/β dimer formation (red), and a tetramerization domain consisting of a 2-helix bundle (ie, an incomplete repeat) near the C-terminus (green). (B) Human erythroid ankyrin (Ankyrin R) consists of an N-terminal membrane protein binding domain composed of ankyrin repeats (yellow), a central spectrin binding domain that harbors a ZU5-containing subdomain identified as the minimal binding domain for β-spectrin (ZU5-ANK) (red/pink), and a C-terminal regulatory domain that modulates the affinities of the other domains and contains a Death Domain (blue).

Figure 2

Structure of the the ZU5-ANK β-spectrin binding domain. (A) Stereo diagram showing the C_α trace of ZU5-ANK. Residues are numbered corresponding to their location in the intact ankyrinR molecule. In the diagram, adjacent C_α atoms are joined by a line. (B) Ribbon stereo diagram of the ZU5-ANK subdomain in the same orientation as in panel A. The overall structure is compact and well-folded with a β-sheet–rich core and several loops. β strands are shown as arrows, helical regions as coils, and regions with no well-defined secondary structure as lines. lines. Rotating views of several representations of the molecule are shown in Video S2 (available on the Blood website; see the Supplemental Materials link at the top of the online article).

Figure 3

ZU5-ANK displays a conserved, positively charged surface patch. (A) The electrostatic map (left) of ZU5-ANK subdomain of ankyrin displays a positively charged patch on its surface attributable to the presence of several positively charged amino acids. The basic residues involved are shown in the ribbon diagram (right) shown in the same orientation as the electrostatic surface. Only the side chains of the positively charged residues are shown (ball-and-stick). The middle panel shows the electrostatic map rotated 180° from the left panel and serves to illustrate that only one surface shows the positively charged patch. The molecular surface of the molecule is shown with the equipotential electrostatic surface mapped onto it at ± 15 K_bT/e_c. (B) Sequence alignment of 3 major human ankyrin isoforms (R, B, and G) indicates that many of the charged residues, boxed in blue, are highly conserved at the sequence level and are likely to form part of similar charged surfaces. The secondary structure elements in the structure are shown above the sequence (helices are shown as rectangles, sheets as arrows, and loops as dashed lines). The black box identifies the canonical ZU5 region.

Figure 4

Structure of human β-spectrin repeats 14 and 15, the ankyrin binding domain. Stereo diagram of β-spectrin repeats 14 and 15 corresponding to the ankyrin binding domain. The structure displays a canonical spectrin fold consisting of 3-helix bundles connected by a helical linker. Repeat 14 is shown in blue and repeat 15, which contains some disordered regions distal to repeat 14, is in red. Helices are marked as A, B, or C, corresponding to the first, second, or third helix of each bundle, respectively. Rotating views of several representations of the molecule are shown in Video S1.

Figure 5

The ankyrin binding domain of human β-spectrin (repeats 14-15) displays a conserved, negatively charged surface patch. (A) A group of acidic residues in repeat 14 form a negatively charged patch. The ribbon diagram of repeats 14 and 15 (left) shows the acidic residues involved in forming this patch (circled). Only the side chains of the negatively charged residues are shown (ball-and-stick). The electrostatic surface of the protein shows the presence of the patch (second diagram from left with circle) in the same orientation as the ribbon diagram. The third diagram from the left shows the electrostatic surface of repeats 14 and 15 rotated approximately 180° with respect to the ribbon diagram. As can be seen, this side of the protein does not show any large charged regions. The structures of other known spectrin repeats do not display such a localized charge distribution; the electrostatic map of β-spectrin repeats 8 and 9 (right) is shown as a representative. The molecular surface of the molecule is shown with the equipotential electrostatic surface mapped onto it at ± 15 K_bT/e_c. (B) Sequence alignment of β-spectrin from several isoforms and species. The alignment shows that the acidic residues on helices A and C are well-conserved, although, in general, the entirety of β-spectrin repeats 14 and 15 are well-conserved. The negatively charged amino acids forming the patch are boxed in red and the position of the helices is shown by cylinders above the sequence.

Figure 6

The structure of the ZU5 domain of ankyrin helps understand general features of this domain. (A) The ribbon diagram shows the ankyrin ZU5 domain (ZU5-ANK). The region corresponding to the canonical ZU5 region is shown in blue, whereas the last 55 amino acids forming this domain are shown in red. The diagram illustrates that 2 of the β strands in the β-sheet core come from residues lying outside the ZU5 consensus region, suggesting that the ZU5 consensus may extend beyond the canonical boundaries. (B) The ribbon diagram shows the functional mutants of ankyrin that map to the ZU5-ANK structure. The mutants are mostly found on the surface of a single face of the molecule. One of the mutations was mapped from an ankyrin variant implicated in hereditary spherocytosis and corresponds to L1046P, whereas other are functional mutants (DAR920AAA, A945P, ESY1026AAA, EE1044AA, KR1049AA) mapped from mutagenesis studies in other ankyrin isoforms that impair spectrin-binding activity in cell culture. Only the side chains of these mutants are shown (ball-and-stick). (C) Sequence alignment of human ankyrin R and 4 other ZU5 domain-containing proteins: Caenorhabditis elegans UNC-44, human netrin receptor, C elegans UNC-5, and human zona occludens-1. In correspondence with the color coding in panel A, the ZU5 domain is boxed blue whereas approximately 50 residues C-terminal to each protein's ZU5 consensus region (when available) are boxed in orange. Secondary structure assignments based on the structure of ZU5-ANK are provided above the sequence alignment as in Figure 3B. Residues printed in red indicate the locations of the functional mutants depicted in panel B. Individually boxed residues denote sites of universal (blue) or high (green) conservation among the presented sequences.