PDBsum entry 1ai1

Complex (antibody/peptide)

PDB id: 1ai1

Contents

Protein chains

215 a.a. *

221 a.a. *

Ligands

HIS-ILE-GLY-PRO-GLY-ARG-AIB-PHE-TYR-THR

* Residue conservation analysis

PDB id:

1ai1

Name:

Complex (antibody/peptide)

Title:

HIV-1 v3 loop mimic

Structure:

Igg1-kappa 59.1 fab (light chain). Chain: l. Igg1-kappa 59.1 fab (heavy chain). Chain: h. Aib142. Chain: p. Engineered: yes. Other_details: the peptide is numbered according to the bh10 isolate

Source:

Mus musculus. House mouse. Organism_taxid: 10090. Strain: balb/c.

Biol. unit:

Trimer (from PQS)

Resolution:

2.80Å R-factor: 0.220

Authors:

J.B.Ghiara,I.A.Wilson

Key ref:

J.B.Ghiara et al. (1997). Structure-based design of a constrained peptide mimic of the HIV-1 V3 loop neutralization site. J Mol Biol, 266, 31-39. PubMed id: 9054968 DOI: 10.1006/jmbi.1996.0768

Date:

06-Nov-96 Release date: 15-May-97

Protein chain

P01837  (IGKC_MOUSE) -  Immunoglobulin kappa constant from Mus musculus

Seq: 107 a.a.

Struc: 215 a.a.

Protein chain

P01868  (IGHG1_MOUSE) -  Ig gamma-1 chain C region secreted form from Mus musculus

Seq: 324 a.a.

Struc: 221 a.a.

Enzyme reactions

• Enzyme class: Chains L, H: E.C.?

DOI no: 10.1006/jmbi.1996.0768

J Mol Biol 266:31-39 (1997)

PubMed id: 9054968

Structure-based design of a constrained peptide mimic of the HIV-1 V3 loop neutralization site.

J.B.Ghiara, D.C.Ferguson, A.C.Satterthwait, H.J.Dyson, I.A.Wilson.

ABSTRACT

Antigenic variation among different HIV-1 isolates has been a major problem in the development of an effective vaccine against AIDS. Peptide vaccines incorporating structural elements common to groups of viral isolates, such as the clade subtypes of HIV-1, hold promise; however, the design of such immunogens has been hampered by the lack of specific structural information on the viral proteins to be targeted. As part of a structure-based approach to this problem, we report the design and characterization of a conformationally restricted peptide analog (Aib142) of a highly conserved HIV-1 clade-B sequence from the third variable loop of the membrane glycoprotein gp120. The design strategy incorporates peptide conformational data derived from crystal structure analysis of an MN-isolate peptide (RP142) in complex with the Fab fragment (Fab59.1) of a broadly neutralizing antibody. The synthetic peptide (Aib142) replaces an alanine residue within the V3 loop epitope sequence GPGRAF by the conformationally restricted helicogenic alpha-aminoisobutyryl residue. As expected, the crystal structure of the Fab 59.1-Aib142 complex at 2.8 A resolution shows that the peptide interacts very similarly with the neutralizing antibody. Proton nuclear magnetic resonance (NMR) studies indicate that the free Aib142 peptide is indeed more ordered in solution with a conformational preference that corresponds to the X-ray structure of its Fab-bound form. Aib142 thus represents the first step in the design of conformationally constrained peptide analogs built to mimic biologically relevant structural forms of HIV-1 neutralization sites.

Selected figure(s)

Figure 1.
Figure 1. Stereo image of the F[o] − F[c] omit electron density for the bound peptide. The density contoured at 2.7σ is shown only around peptide residues Gly^P319-Pro-Gly-Arg-Aib-Phe^P324, which form the S-shaped double turn in the Fab combining site. When compared to the density for RP142 [Ghiara et al 1994], additional electron density is seen around the extra methyl group of the Aib residue. Peptide residue numbers are according to the BH10 isolate sequence [Ratner et al 1985], and are preceded by the letter P. Peptide Aib142 was synthesized on a methyl benzhydrylamine (MBHA) resin following standard procedure [Schnolzer et al 1992]. The Aib was coupled manually for 30 minutes followed by a second coupling for 60 minutes. Following treatment with anhydrous HF and purification by preparative HPLC, the peptide was characterized by analytical HPLC and electrospray mass spectrometry (observed 2891.6 (+/−0.6)Da; calculated average 2891.5 Da). Fab fragments were prepared by enzymatic digestion of monoclonal antibody 59.1 (IgG1) with pepsin followed by controlled reduction in the presence of cysteine. Crystals of Fab 59.1 in complex with peptide Aib142 were obtained by vapor diffusion in sitting drops (2 to 5 μl), that contained between 1.4 and 1.8 M mixed phosphate (NaH[2]PO[4] and K[2]HPO[4] pH range 5.0 to 6.75), with 18 mg/ml Fab at 22°C and a tenfold molar excess of Aib142. The crystals resemble the Fab59.1-RP142 crystals in morphology. Screenless precession photographs confirmed that the cell was orthorhombic, with similar cell dimensions to the Fab59.1-RP142 complex (a = 89.7 Å, b = 154.0 Å, and c = 121.9 Å [Ghiara et al 1994]). X-ray diffraction data were collected from a large crystal (0.8 mm × 0.4 mm × 0.3 mm) grown in 1.4 M mixed phosphate buffer, pH 6.5, on a MAR image plate mounted on a Siemens X-ray generator at a maximum power of 100 mA and 50 kV. Detector data were processed with MOSFLM [Leslie et al 1986]. The space group was confirmed to be C 222[1], with unit cell dimensions a = 89.9 Å, b = 154.4 Å, and c = 121.4 Å. The final data set consists of 50,028 total observations of 19,192 unique reflections and is 92% complete to 2.8 Å (90% complete in the 2.8 to 2.9 Å outer shell) with an R[sym](I) value of 8.0%. The Fab59.1-Aib142 structure was determined by molecular replacement. As the crystal cell dimensions for the two complexes (with RP142 and Aib142) are within 0.5 Å of each other and their respective peptides differ only at one residue, the refined Fab59.1-RP142 complex structure ([Ghiara et al 1994] Brookhaven Protein Data Bank, code 1ACY) was used as the initial starting model. After rigid body refinement using X-PLOR [Brunger 1992], the R-value was 0.24 for 8.0 to 4.0 Å data with F>2σ. The model was then refined with the positional and simulated annealing protocols in X-PLOR to an R-value of 0.27 for all data from 12.0 to 2.8 Å. The RP142 peptide was included in the model at this stage to prevent any side-chains from CDR loops from moving into the peptide electron density during simulated annealing. F[o] − F[c] omit maps, calculated by excluding the peptide, clearly demonstrated no significant difference for the Aib142 peptide conformation. One cycle of model building of the entire complex into 10% 2F[o] − F[c] omit maps was then undertaken with FRODO [Jones 1978 and Jones 1982]. The hydrogen atom on Ala of the peptide was replaced with a methyl group with Insight II, version 2.2 (Biosym Technologies, 1993), to incorporate the Aib residue into the model; the parameter and topology files used in X-PLOR [Engh and Huber 1991] were also modified to include specifications for the Aib residue. Two more cycles of model building and refinement resulted in the current structure, with an R-value of 0.22 for all data in the 12.0 to 2.8 Å range, with atomic B-factor refinement (the average overall B-value is 30 Å^2) and rms deviations from ideality for bond lengths and angles of 0.015 Å and 2.0°, respectively. Coordinates have been deposited in the Brookhaven Data Bank, code 1A T 1.

Figure 2.
Figure 2. (a) Conformation of the bound peptide. The three turns in the peptide are formed by residues Gly-Pro-Gly-Arg (type II), Gly-Arg-Aib-Phe (type III), and Arg-Aib-Phe-Tyr (type I). The intra-peptide hydrogen bonds are indicated by broken yellow lines. (b) Superimposition of the peptides RP142 (pink) and Aib142 (blue) as seen in the two independent Fab59.1-peptide complexes. The Fabs have been omitted from the Figure for clarity. The peptide conformations are very similar, with rms deviations of 0.34 Å for backbone atoms (N, C^α, C and O) and 0.64 Å when all non-hydrogen atoms of the peptides are considered.

The above figures are reprinted by permission from Elsevier: J Mol Biol (1997, 266, 31-39) copyright 1997.

Figures were selected by an automated process.