Cryptic association of B7-2 molecules and its implication for clustering

Abstract

T-cell co-stimulation through CD28/CTLA4:B7-1/B7-2 axis is one of the extensively studied pathways that resulted in the discovery of several FDA-approved drugs for autoimmunity and cancer. However, many aspects of the signaling mechanism remain elusive, including oligomeric association and clustering of B7-2 on the cell surface. Here, we describe the structure of the IgV domain of B7-2 and its cryptic association into 1D arrays that appear to represent the pre-signaling state of B7-2 on the cell membrane. Super-resolution microscopy experiments on heterologous cells expressing B7-2 and B7-1 suggest, B7-2 form relatively elongated and larger clusters compared to B7-1. The sequence and structural comparison of other B7 family members, B7-1:CTLA4 and B7-2:CTLA-4 complex structures, support our view that the observed B7-2 1D zipper array is physiologically important. This observed 1D zipper-like array also provides an explanation for its clustering, and upright orientation on the cell surface, and avoidance of spurious signaling.

Keywords: B7-2 structure; immune checkpoint blockade; immune receptors; protein clustering.

FIGURE 1

(a) Cartoon representation of four B7‐2 molecules in the asymmetric unit labelled as chains A, B, C, and D. Chains AB and CD forms two similar dimers. (b) Dimer interface is mediated by three close hydrogen bonds (shown as red dotted lines) between A strands of two protomers of the dimer. The continuous beta sheet formed by the strands A, G, F, C, C′, C″ of both the protomers of the dimer is shown in cartoon representation while the surface view of rest of the molecule is made transparent. (c) Cartoon representation of chain A and chain B is shown in VIBGYOR colours. The N‐terminus of both chains is in violet and the C‐terminus of both chains is shown in red. Note that the C‐terminus of both the associating chains is facing the same direction and are placed close together. The same orientation is maintained with respect to chains C and D in the asymmetric unit

FIGURE 2

Dimer like association observed in hB7‐2 structure is similar to that of mouse CD244(2B4) and human CD58 structures. (a) mouse CD244 (PDB entry: 2PTU) structure shown in cartoon representation. (b) human CD58 (PDB entry:1QA9) structure shown in cartoon representation. (c) structures of CD244 and CD58 superposed on B7‐2(PDB entry: 5YXK; structure reported in this study) and represented in ribbons. The colours in all the three figures are maintained in VIBGYOR where the N‐terminus is shown in violet and the C‐terminus is shown in red. Note that all three structures are oriented in a similar fashion and the C‐termini of both protomers in all the structures face the same direction in a biologically relevant orientation

FIGURE 3

Comparison of complete ectodomain structures of hB7‐1 (1DR9‐green) and hB7‐2 modelled with I‐TASSER (blue). (a) Ectodomain structure with both IgV and IgC domains of hB7‐1 superposed on ectodomain structure of hB7‐2 modelled with ITASSER. The superposition reveals over‐all similar domain architecture of hB7‐1 and hB7‐2. The enlarged view of superposed interdomain core of hB7‐1 and hB7‐2 shows similar hydrophobic interactions in both the structures. (b) Sequence logos representing conserved hydrophobic residues in interdomain core in hB7‐1 and (c) in B7‐2

FIGURE 4

F/Y‐P cis‐peptide in interdomain hydrophobic core is a signature of B7 family members. (a) Sequence logo representing conserved hydrophobic residues in the interdomain core of human B7‐family members. The logo was generated from multiple sequence alignment of human B7‐family members (hB7‐1, hB7‐2, hB7‐H1, hB7‐H2, hB7‐H3, hB7‐H4, hB7‐DC, and hB7‐H6). cis‐peptide forming proline183 which is 100% conserved and the adjacent aromatic residue F/Y182 are highlighted in rectangular box. (b) MSA of human B7‐family members. Note that, despite the over‐all similarity among the sequences being very less, the cis‐peptide proline is 100% conserved besides the IgSF characteristic canonical disulfide forming cysteines. 100% conserved residues are highlighted in red. (c) Superposition of interdomain hydrophobic core of the available structures of B7‐family members—hB7‐1 (PDB entry:1DR9; green), hB7‐H1(PD‐L1; PDB entry: 3BIS; orange), mB7‐DC (PD‐L2; PDB entry:3BP5; red), and hB7‐H6(PDB entry:3PV7; blue). F/Y‐P cis peptide is highlighted with a red‐dotted circle. Note that the proline of cis‐peptide restricts the adjacent aromatic ring to face the hydrophobic core

FIGURE 5

B7‐2 IgV forms a physiologically relevant 1D molecular zipper in the crystal: one‐dimensional zipper like molecular array running parallel to the crystallographic a‐axis Note that much of the surface of B7‐2 is buried while its CTLA‐4/CD28 binding surface, highlighted in light gray, is exposed for all the molecules of the array. The C‐termini of all the molecules (shown in red) is facing one direction as if emerging from membrane of a single cell

FIGURE 6

Super‐resolution microscopy to observe clustering of B7‐2: A and B represents the gallery of nanodomains for B7‐2 and B7‐1, respectively. Top panel represents the nanodomains having shape factor 0.75 to 1 and the panel below represents the nanodomains having shape factor <0.75. C, D, and E indicate the principal axis, shape factor and average intensity for the nanodomains of B7‐2 and B7‐1, respectively. n = 4,661 nanodomains for B7‐2 and n = 9918 nanodomains for B7‐1. Data points represent interquartile range‐Median (IQR25% to 75%). Statistical test is Mann–Whitney test; ***p < 0.001 and ****p < 0.0001. Scale bar at A and B is 0.5 μm