We have previously identified a 13-residue cyclic peptide, Compstatin, that binds to complement component C3 and inhibits complement activation. Herein, we describe the binding kinetics, structure-activity relationship, and biotransformation of Compstatin. Biomolecular interaction analysis using surface-plasmon resonance showed that Compstatin bound to native C3 and its fragments C3b and C3c, but not C3d. While binding of Compstatin to native C3 was biphasic, binding to C3b and C3c followed the 1:1 Langmuir binding model; the affinities of Compstatin for C3b and C3c were 22- and 74-fold lower, respectively, than that of native C3. Analysis of Compstatin analogs synthesized for structure-function studies indicated that 1) the 11-membered ring between disulfide-linked Cys2-Cys12 constitutes a minimal structure required for optimal activity; 2) retro-inverso isomerization results in loss of inhibitory activity; and 3) some residues of the type I β-turn segment also interact with C3. In vitro studies of Compstatin in human blood indicated that a major pathway of biotransformation was the removal of Ile1, which could be blocked by N-acetylation of the peptide. These findings indicate that acetylated Compstatin is stable against enzymatic degradation and that the type I β-turn segment is not only critical for preservation of the conformational stability, but also involved in intermolecular recognition.

Glycoprotein D (gD) is a structural component of the herpes simplex virus (HSV) envelope which is essential for virus entry into host cells. Chinese hamster ovary (CHO-K1) cells are one of the few cell types which are nonpermissive for the entry of many HSV strains. However, when these cells are transformed with the gene for the herpesvirus entry mediator (HVEM), the resulting cells, CHO-HVEM12, are permissive for many HSV strains, such as HSV-1(KOS). By virtue of its four cysteine-rich pseudorepeats, HVEM is a member of the tumor necrosis factor receptor superfamily of proteins. Recombinant forms of gD and HVEM, gD-1(306t) and HVEM(200t), respectively, were used to demonstrate a specific physical interaction between these two proteins. This interaction was dependent on native gD conformation but independent of its N-linked oligosaccharides, as expected from previous structure-function studies. Recombinant forms of gD derived from HSV-1(KOS)rid1 and HSV-1(ANG) did not bind to HVEM(200t), explaining the inability of these viruses to infect CHO-HVEM12 cells. A variant gD protein, gD-1(⌬290-299t), showed enhanced binding to HVEM(200t) relative to the binding of gD-1(306t). Competition studies showed that gD-1(⌬290-299t) and gD-1(306t) bound to the same region of HVEM(200t), suggesting that the differences in binding to HVEM are due to differences in affinity. These differences were also reflected in the ability of gD-1(⌬290-299t) but not gD-1(306t) to block HSV type 1 infection of CHO-HVEM12 cells. By gel filtration chromatography, the complex between gD-1(⌬290-299t) and HVEM(200t) had a molecular mass of 113 kDa and a molar ratio of 1:2. We conclude that HVEM interacts directly with gD, suggesting that HVEM is a receptor for virion gD and that the interaction between these proteins is a step in HSV entry into HVEM-expressing cells.

Human complement receptor type 2 (CR2, CD21) is a cell surface receptor that binds three distinct ligands (complement C3d, Epstein-Barr virus gp350/220, and the low-affinity IgE receptor CD23) via the N-terminal two of fifteen or sixteen short consensus/complement repeat (SCR) domains. Here, we report biophysical studies of the CR2 SCR 1-2 domain binding to its ligand C3dg. Two recombinant forms of CR2 containing the SCR 1-2 and SCR 1-15 domains were expressed in high yield in Pichia pastoris and baculovirus, respectively. Circular dichroism spectroscopy showed that CR2 SCR 1-2 receptor possessed a beta-sheet secondary structure with a melting temperature of 59 degrees C. Using surface plasmon resonance, kinetic parameters for the binding of either CR2 SCR 1-2 or the full-length SCR 1-15 form of CR2 showed that the affinity of binding to immobilized C3d is comparable for the SCR 1-15 compared to the SCR 1-2 form of CR2. Unexpectedly, both the association and dissociation rates for the SCR 1-15 form were slower than for the SCR 1-2 form. These data show that the SCR 1-2 domains account for the primary C3dg binding site of CR2 and that the additional SCR domains of full-length CR2 influence the ability of CR2 SCR 1-2 to interact with its ligand. Studies of the pH and ionic strength dependence of the interaction between SCR 1-2 and C3d by surface plasmon resonance showed that this is influenced by charged interactions, possibly involving the sole His residue in CR2 SCR 1-2. Sedimentation equilibrium studies of CR2 SCR 1-2 gave molecular weights of 17 000, in good agreement with its sequence-derived molecular weight to show that this was monomeric. Its sedimentation coefficient was determined to be 1.36 S. The complex with C3d gave molecular weights in 50 mM and 200 mM NaCl buffer that agreed closely with its sequence-derived molecular weight of 50 600 and showed that a 1:1 complex had been formed. Molecular graphics views of homology models for the separate CR2 SCR 1 and SCR 2 domains showed that both SCR domains exhibited a distribution of charged groups throughout its surface. The single His residue is located near a long eight-residue linker between the two SCR domains and may influence the linker conformation and the association of C3d and CR2 SCR 1-2 into their complex. Sedimentation modeling showed that the arrangement of the two SCR domains in CR2 SCR 1-2 is highly extended in solution.