About TSRI
Research & Faculty
News & Publications
Scientific Calendars
Scripps Florida
PhD Program
Campus Services
Work at TSRI
TSRI in the Community
Giving to TSRI
Site Map & Search

The Skaggs Institute
for Chemical Biology

Scientific Report 2008

X-ray Crystallographic Studies of Therapeutically Important Macromolecular Targets

I.A. Wilson, M.A. Adams, C.H. Bell, R.M.F. Cardoso, S. Connelly, B.J. Droese, D.C. Ekiert, M.-A. Elsliger, Z. Fulton, B.W. Han, M. Hong, M.J. Jimenez-Dalmaroni, R.N. Kirchdoerfer, R. Pejchal, G.P. Porter, A. Schiefner, R.L. Stanfield, R.S. Stefanko, J.A. Vanhnasy, R. Xu, X. Xu, S.I. Yoon, X. Zhu

HIV Type 1 Vaccine Design

HIV type 1 (HIV-1) continues to constitute a major worldwide health threat, with approximately 2.1 million HIV-related deaths and 2.5 million new HIV-1 infections in 2007. Currently, more than 20 anti-HIV drugs approved by the Food and Drug Administration are on the market. Although these drugs can be effective at lowering the levels of circulating virus, they cannot completely eliminate the virus, are expensive, and must be taken daily for life. Clearly, an effective vaccine against HIV-1 is needed to control the rampant spread of this devastating pandemic. An effective vaccine likely must elicit a vigorous antibody response to block or neutralize viral infection. However, in many studies of patients infected with HIV-1, only a handful of potent, broadly neutralizing antibodies have been discovered that recognize a large percentage of the circulating viral strains. Our goal has been to understand how these rare antibodies are able to combat the virus.

We are elucidating the 3-dimensional structures of these rare antibodies in complexes with the antibodies' viral epitopes from the envelope proteins gp120 and gp41. The antibodies under study include b12, which recognizes a highly conserved, but deeply recessed pocket on gp120 that is the receptor-binding site for CD4; 2G12, which binds to a mannose-rich carbohydrate cluster on the normally immunologically silent face of gp120; several antibodies that recognize the hypervariable V3 region of gp120; and 4E10 and Z13e1, which interact with overlapping epitopes on gp41, just proximal to its membrane-spanning domain.

Our original structural studies of antibody 2G12 in complex with mannose sugars have led to design of many nonnatural carbohydrates, peptides, and small-molecule mimics for testing as potential immunogens. Recently, in collaboration with B. Davis, University of Oxford, Oxford, England, we determined a 2.8-Å crystal structure for 2G12 in complex with a novel, nonself mimic of the D1 arm of Man4/Man9GlcNac2, the type of carbohydrate commonly found on the gp120 silent face. This nonnatural mannose variant contains a C-6 methyl substitution of the mannose at the terminus of the D1 arm and inhibits binding of 2G12 to gp120 better than does the D1 arm itself. This compound is the first nonself D1 arm derivative to demonstrate inhibition of 2G12/gp120 binding better than that of the natural D1 arm. Preliminary diffraction data have also been collected from crystals of 2G12 in complex with a C-6 methyl monosaccharide compound. Further optimization of crystallization conditions for both complexes to obtain higher resolution diffraction data are under way.

We are currently refining the crystal structure for Fab Z13e1, a neutralizing antibody that recognizes an epitope in the gp41 membrane proximal external region that overlaps the Fab 4E10 epitope (Fig. 1). Z13e1 has been evolved to have higher affinity than the parent antibody Z13; the higher affinity is due to 5 mutations in complementarity-determining region L3 (residues L90–L97). Although Z13e1 and 4E10 recognize similar epitopes, the antibody-bound conformations for the epitopes differ markedly, perhaps giving some insight to conformational changes that may occur during viral entry into cells. The differences in the key contact residues also correlate with the breadth and neutralization potency of 4E10 vs Z13e1.
Fig. 1. Structure of Fab Z13e1 in complex with its epitope peptide on the HIV-1 gp41 membrane proximal external region. The Fab is shown in a solid surface representation, with the light and heavy chains in cyan and blue, respectively. The peptide, shown in a ball-and-stick representation, binds primarily to the heavy chain of the antibody and adopts a different conformation than do the corresponding residues bound to broadly neutralizing antibody 4E10. This epitope occurs in gp41 just before the protein enters the membrane, so Z13e1 may have to contact the membrane to bind its epitope.

Our studies on HIV are done in collaboration with D.R. Burton, M. Zwick, R. Pantophlet, P.E. Dawson, and C.-H. Wong, Scripps Research; B. Davis, University of Oxford; L. Cavacini and J.K. Scott, Simon Fraser University, Burnaby, British Columbia; J. Moore, Weill Medical College of Cornell University, New York, New York;
H. Katinger, R. Kunert, and G. Stiegler, University für Bodenkultur, Vienna, Austria; R. Wyatt and P. Kwong, Vaccine Research Center, National Institutes of Health, Bethesda, Maryland; W. Olson and K. Kang, Progenics Pharmaceuticals, Inc., Tarrytown, New York; the National Institutes of Health; and the Neutralizing Antibody Consortium of the International AIDS Vaccine Initiative.

Influenza Virus Glycoproteins

The 1918 influenza pandemic, which was responsible for more than 20 million deaths worldwide, and the more recent "bird flu," with its even higher mortality rate (about 60% of patients in whom it is diagnosed), are constant reminders of the potential devastation that could ensue if a new influenza pandemic were to occur. To aid in the design of a vaccine to protect against such highly virulent strains of influenza, we are carrying out structural and functional studies of envelope proteins of influenza virus in complex with neutralizing antibodies to the virus.

All known antibodies that neutralize influenza virus recognize the hemagglutinin viral envelope protein. In general, antibodies to hemagglutinin generally recognize highly variable epitopes at the membrane distal end of the hemagglutinin trimer. However, a small proportion of the host repertoire of antibodies is directed against other sites on hemagglutinin, including several antibodies that bind on the side of the hemagglutinin trimer and recognize highly invariant epitopes.

Several of these unusual antibodies neutralize the hemagglutinin of different strains and subtypes of influenza virus, both in vivo and in vitro. To gain insight into the mechanism of virus neutralization and the nature of the epitopes recognized, we are investigating several of these broadly neutralizing antibodies in complex with hemagglutinins that represent different pandemic strains and subtypes (H1, H2, H3) of human influenza virus as well as avian H5N1 influenza viruses. Understanding how these more broadly neutralizing antibodies interfere with viral entry and subsequent infection, as well as the nature of the highly conserved epitopes, will provide insights into the functional and antigenic constraints on the hemagglutinin of influenza virus.

Currently, we are working with 2 Fabs, H5M11 and H5M9, that neutralize the H5N1 avian influenza virus A/Goose/Guangdong/1/96 and more recent avian strains that have infected humans and with antibodies to the influenza virus that were isolated from survivors of the 1918 pandemic by J. Crowe, Vanderbilt University, Nashville, Tennessee. The structure of Fab H5M11 has been determined, and we are working toward crystallization of other Fabs in complex with hemagglutinins from the H5N1 and 1918 H1N1 viruses.

Hemagglutinin facilitates cell fusion through interactions with host membranes. Although crystal structures of hemagglutinin ectodomains have been extensively studied, little is known about the conformation or function of the membrane-interacting regions. We are working toward the determination of crystal structures of the full-length hemagglutinin in its states before and after fusion. In collaborative research with G. Tobin, Biological Mimetics, Inc., Frederick, Maryland, crystallization of the full-length hemagglutinin from A/Wyoming/3/03 (H3 subtype) and bacterial expression of the postfusion form of the protein are under way. We also propose to isolate and structurally analyze the hemagglutinin from the pandemic A/Japan/305/57 (H2 subtype) virus. These studies will advance our understanding of the mechanism of hemagglutinin-induced fusion and provide novel targets for design of fusion inhibitors.

The crystal structure of the neuraminidase of the 1918 H1N1 virus has been determined to 1.45 Å. A large cavity in the active site in the neuraminidase offers new opportunities for structure-based drug design. Crystal structures of the 1918 neuraminidase in complex with the antiviral drugs oseltamivir (Tamiflu) and zanamivir (Relenza) show that the loop bordering the cavity is extremely flexible in binding substrates, a characteristic that may indicate that the 1918 neuraminidase can bind more chemically diverse ligands than can neuraminidases from some other subtypes of the virus. This high-resolution structural information is being used for rational design of inhibitors against influenza virus.

Additional collaborators in the influenza research include our colleagues in the flu consortium funded by the National Institute of Allergy and Infectious Diseases; scientists at Crucell, Leiden, the Netherlands; J. Crowe, Vanderbilt University; A. Lanzavecchia, Institute for Research in Biomedicine, Bellinzona, Switzerland; and X. Che, Southern Medical University China, Guangzhou, China.

The Innate Immune Response Against Microbial Pathogens

Toll-like receptors (TLRs) are glycoproteins that are essential for innate immune recognition of microbial pathogens. The TLR extracellular domains are horseshoe-shaped molecules consisting of leucine-rich repeat domains that begin and end with an N- and a C-terminal cap domain. Recently, we have been studying the TLR4, which plays an essential role in recognition and signaling of bacterial lipopolysaccharide. Among the TLR family members, TLR4 is unique in requiring another molecule, myeloid differentiation protein-2 (MD-2), for its function. MD-2 directly binds lipopolysaccharide and induces dimerization and activation of TLR4 for signaling.

To provide insights into the structural mechanism used by lipopolysaccharide to activate TLR4, we have expressed the extracellular domain of TLR4 (sTLR4) in complex with MD-2 in a baculovirus expression system. Biophysical studies suggest that purified 1:1 complexes of sTLR4–MD-2 homodimerize to form 2:2 complexes in the presence of lipopolysaccharide. X-ray crystallographic studies are being carried out to reveal the structural architecture of the sTLR4–MD-2 assembly induced by lipopolysaccharide.

Jawless fish, such as the lamprey, do not have immune receptors, such as antibodies, T-cell receptors, or MHC molecules, yet the fish still have an adaptive immune response to antigen. Recently, it was shown that cell-surface molecules, termed variable lymphocyte receptors (VLRs) are responsible for the adaptive immune response in jawless fish. These receptors resemble the mammalian innate system TLRs, with an overall horseshoe shape made up of a variable number of different leucine-rich repeat domains. In collaboration with M. Cooper, Emory University, Atlanta, Georgia, we recently determined the first crystal structure of a VLR-antigen complex, RBC36, in complex with the H trisaccharide derived from the H antigen of human type O erythrocytes (Fig. 2). This structure reveals for the first time the location and nature of the VLR antigen-binding site.
Fig. 2. Structure of the lamprey VLR red blood cell in complex with its H trisaccharide antigen. Left, The VLR is shown as a solid surface, with the antigen depicted in a ball-and-stick representation. Right, The same view of the structure but highlighting the secondary structural elements of the VLR. The antigen binds to the concave side of the horseshoe-shaped VLR and is cradled on one side by a large fingerlike insert in the C-terminal leucine-rich repeat.

The CD1 family of innate receptors consists of MHC class I–like, antigen-presenting molecules that present lipids, glycolipids, and lipopeptides to effector T cells. The receptors are expressed on antigen-presenting cells and are involved in host defense and in immunoregulatory functions. Glycolipids presented by CD1d are capable of stimulating natural killer T cells. Natural killer
T cells are of clinical interest because when stimulated by CD1, they rapidly secrete a number of cytokines that either promote or suppress different immune responses. One of the most potent agonists for natural killer T cells is α-galactosylceramide. On the basis of our structural studies during the past 5 years, a series of glycolipids have been synthesized by scientists in the laboratory of C.-H. Wong, Skaggs Institute. These new ligands, which have phenyl ring substitutions in the fatty acid part of α-galactosylceramide, are more potent than the native ceramide and have an altered efficacy in T-cell assays. We have now determined the structures for 3 of the most stimulating glycolipids in complex with CD1d. Our analysis revealed that only minor structural changes occur in the A′ pocket (Fig. 3). The phenyl-ring derivatives have better packing than the native α-galactosylceramide and thereby increase the overall stability of the CD1d-ligand complexes.
Fig. 3. Crystal structure of CD1d with designed agonists. The ligand-presenting α1α2 platform of mouse CD1d is shown in gray, overlaid by the transparent molecular surface, with the bound superimposed ligands C6Ph (red), C8Ph (yellow), C8PhF (green), and the short α-galactosylceramide agonist PBS-25 (blue). The phenyl substitutions protrude deeply into the A′ pocket (see inset on right for view of ligands only). The 2 main A′ and F′ pockets in CD1d for ligand binding are indicated. A spacer lipid (likely palmitic acid) also partially fills the A′ pocket when the glycolipid ligand itself does not fully occupy the pocket. The Cα positions of the ligands show an overall root mean square displacement of 0.42 Å.


Astronomo, R.D., Lee, H.K., Scanlan, C.N., Pantophlet, R., Huang, C.Y., Wilson, I.A., Blixt, O., Dwek, R.A., Wong, C.H., Burton, D.R. A glycoconjugate antigen based on the recognition motif of a broadly neutralizing human immunodeficiency virus antibody, 2G12, is immunogenic but elicits antibodies unable to bind to the self glycans of gp120. J. Virol. 82:6359, 2008.

Bell, C.H., Pantophlet, R., Schiefner, A., Cavacini, L.A., Stanfield, R.L., Burton, D.R., Wilson, I.A. Structure of antibody F425-B4e8 in complex with a V3 peptide reveals a new binding mode for HIV-1 neutralization. J. Mol. Biol. 375:969, 2008.

Burton, D.R., Wilson, I.A. Immunology: square-dancing antibodies. Science 317:1507, 2007.

Debler, E.W., Kaufmann, G.F., Meijler, M.M., Heine, A., Mee, J.M., Pljevaljcic, G., Di Bilio, A.J., Schultz, P.G., Millar, D.P., Janda, K.D., Wilson, I.A., Gray, H.B., Lerner, R.A. Deeply inverted electron-hole recombination in a luminescent antibody-stilbene complex. Science 319:1232, 2008.

Debler, E.W., Müller, R., Hilvert, D., Wilson, I.A. Conformational isomerism can limit antibody catalysis. J. Biol. Chem. 283:16554, 2008.

Demartino, J.K., Hwang, I., Connelly, S., Wilson, I.A., Boger, D.L. Asymmetric synthesis of inhibitors of glycinamide ribonucleotide transformylase. J. Med. Chem. 51:5441, 2008.

Dhillon, A.K., Stanfield, R.L., Gorny, M.K., Williams, C., Zolla-Pazner, S., Wilson, I.A. Structure determination of an anti-HIV-1 Fab 447-52D-peptide complex from an epitaxially twinned data set. Acta Crystallogr. D Biol. Crystallogr. 64:792, 2008.

Huang, C.C., Lam, S.N., Acharya, P., Tang, M., Xiang, S.H., Hussan, S.S., Stanfield, R.L., Robinson, J., Sodroski, J., Wilson, I.A., Wyatt, R., Bewley, C.A., Kwong, P.D. Structures of the CCR5 N terminus and of a tyrosine-sulfated antibody with HIV-1 gp120 and CD4. Science 317:1930, 2007.

Johnson, S.M., Connelly, S., Wilson, I.A., Kelly, J.W. Biochemical and structural evaluation of highly selective 2-arylbenzoxazole-based transthyretin amyloidogenesis inhibitors. J. Med. Chem. 51:260, 2008.

Menendez, A., Calarese, D.A., Stanfield, R.L., Chow, K.C., Scanlan, C.N., Kunert, R., Katinger, H., Burton, D.R., Wilson, I.A., Scott, J.K. A peptide inhibitor of HIV-1 neutralizing antibody 2G12 is not a structural mimic of the natural carbohydrate epitope on gp120. FASEB J. 22:1380, 2008.

Relloso, M., Cheng, T.Y., Im, J.S., Parisini, E., Roura-Mir, C., DeBono, C., Zajonc, D.M., Murga, L.F., Ondrechen, M.J., Wilson, I.A., Porcelli, S.A., Moody, D.B. pH-dependent interdomain tethers of CD1b regulate its antigen capture. Immunity 28:774, 2008.

Stevens, J., Blixt, O., Chen, L.M., Donis, R.O., Paulson, J.C., Wilson, I.A. Recent avian H5N1 viruses exhibit increased propensity for acquiring human receptor specificity. J. Mol. Biol. 381:1382, 2008.

Verdino, P., Aldag, C., Hilvert, D., Wilson, I.A. Closely related antibody receptors exploit fundamentally different strategies for steroid recognition. Proc. Natl. Acad. Sci. U. S. A. 105:11725, 2008.

Wei, C.J., Xu, L., Kong, W.P., Shi, W., Canis, K., Stevens, J., Yang, Z.Y., Dell, A., Haslam, S.M., Wilson, I.A., Nabel, G.J. Comparative efficacy of neutralizing antibodies elicited by recombinant hemagglutinin proteins from avian H5N1 influenza virus. J. Virol. 82:6200, 2008.

Zajonc, D.M., Savage, P.B., Bendelac, A., Wilson, I.A., Teyton, L. Crystal structures of mouse CD1d-iGb3 complex and its cognate Vα14 T cell receptor suggest a model for dual recognition of foreign and self glycolipids. J. Mol. Biol. 377:1104, 2008.

Zajonc, D.M., Wilson, I.A. Architecture of CD1 proteins. Curr. Top. Microbiol. Immunol. 314:27, 2007.

Zhu, X., Xu, X., Wilson, I.A. Structure determination of the 1918 H1N1 neuraminidase from a crystal with lattice-translocation defects. Acta Crystallogr. D Biol. Crystallogr. 64:843, 2008.


Ian A. Wilson, D.Phil.

Wilson Web Site