Scientific Report 2006

Chemistry

Chemistry, Biology, and Inflammatory Disease

P. Wentworth, Jr., J.Y. Chang, Y.P. Chen, J. Dambacher, L. Eltepu, R.K. Grover, J. Nieva, M. Puga, A. Shafton, B.D. Song, M.M.R. Peram, J. Rogel, S. Tripurenani, R. Troseth, H. Wang, A.D. Wentworth

Our research is interdisciplinary and involves bioorganic, biophysical, physical organic, synthetic, and analytical chemistry coupled with biochemical techniques, cell-based assays, and animal models. These diverse approaches are combined to facilitate a better understanding of and generate new therapeutic approaches to complex disease states. Ongoing projects include studies on atherosclerosis, neurodegeneration, ischemia-reperfusion injury, macular degeneration, cancer, inflammation, and infectious diseases.

The Antibody-Catalyzed Water Oxidation Pathway

Our discovery that all antibody molecules, regardless of source or antigenic specificity, can catalyze the reaction between singlet oxygen and water to give hydrogen peroxide is causing a revision of the axiom that antibodies are the classical adapter molecule of the immune system, linking recognition and killing of foreign pathogens. Both the chemical and biological aspects of this pathway are being explored extensively, and intriguing new insights into how this pathway may play a role in immune defense and inflammatory damage are emerging.

We recently showed that flavins, biologically relevant photoactive molecules, can act as efficient triggers for the antibody-catalyzed water oxidation pathway at physiologically relevant concentrations and a certain amount of visible light radiation. This observation indicates that immune defense and damage may be linked to an association of flavins, such as riboflavin or flavin mononucleotide, with immunoglobulins. The trigger would be antibody-antigen binding on cells exposed to visible light, such as the skin or retina.

In collaboration with I.A. Wilson, Department of Molecular Biology, we recently studied, at atomic resolution, an antibody, IgG^GAR, that has riboflavin bound in the complementarity-determining regions (Fig. 1). Of interest, the bound riboflavin is essentially photochemically inert. The x-ray structure of the IgG^GAR-flavin complex offers an insight into the lack of photochemical activity; the isoalloxazine ring is effectively sandwiched between several aromatic side chains. Binding of riboflavin occurs in a narrow cleft with the isoalloxazine ring stacked between parallel aromatic groups of tyrosine at position H33 (with the re face of riboflavin), phenylalanine at position H58, and tyrosine at position H100A (both associated with the si face of the flavin); the distances between the isoalloxazine ring and the respective aromatic rings from these 3 residues are about 3.2, 3.5. and 3.4 Å, respectively. Such π stacking is known to quench the excited state of riboflavin, a feature that a number of flavin-binding proteins have also evolved, presumably to protect themselves from the intrinsic photochemistry of the flavin moiety.

Fig. 1. Schematic presentation of riboflavin within the binding site of IgGGAR. Residues that form van der Waals interactions with riboflavin are indicated; those that participate in the hydrogen bonds with the riboflavin are shown in ball-and-stick representations. Hydrogen bonds are illustrated as dotted lines.

We are searching for the active site for the antibody-catalyzed water oxidation pathway within the antibody structure. We have cloned and expressed soluble individual domains of the murine Fab 4C6. Initial attempts to efficiently express wild-type domains were unsuccessful, but by mutating hydrophobic residues involved in interdomain interactions into soluble hydrophilic residues, we prepared folded domains. All the domains we cloned and expressed have been successfully purified to homogeneity from the periplasm of Escherichia coli transformed with plasmids encoding individual domains. In addition, to assess the critical nature of tryptophan residues in the photosensitization of antibody-bound triplet oxygen to singlet oxygen, we mutated 2 tryptophan residues within domain C_H1 to leucine. Interestingly, neither the single nor the double mutation affected the ability of ultraviolet-irradiated C_H1 to generate hydrogen peroxide.

Cholesterol Seco-Sterols and Inflammatory Disease

We recently discovered that the 5,6-seco-sterols atheronal-A and atheronal-B (Fig. 2), compounds of a class of cholesterol ozonolysis products, are present in human atherosclerotic plaques and plasma. In addition, we found that the atheronals are present in murine models of atherosclerosis, in a rabbit model of acute respiratory distress syndrome (studies done in collaboration with C. Cochrane, Department of Molecular and Experimental Medicine), and in brain tissue from patients with Lewy body dementia (studies done in collaboration with J. Kelly, Department of Chemistry).

Fig. 2. The cholesterol seco-sterols atheronal-A (top) and atheronal-B (bottom).

The atheronals have a range of biological properties that in combination would increase the number of macrophages at sites of vascular inflammation. When cultured macrophage cells were incubated with atheronal-A and atheronal-B complexed with low-density lipoprotein (LDL), marked upregulation of scavenger receptor class A, but not CD36, occurred, showing that cultured macrophages respond to complexes of atheronals and LDL in a manner highly analogous to the response to acetylated LDL. Both atheronal-A and atheronal-B induce chemotaxis of cultured macrophages in a dose-dependent manner. When complexed with LDL, atheronal-A, but not atheronal-B, induces a dose-dependent upregulation of the cell-surface adhesion molecule E-selectin on vascular endothelial cells. When complexed with LDL, atheronal-B, but not atheronal-A, induces cultured human monocytes to differentiate into macrophage cell lineage.

These in vitro data and the effects of cholesterol 5,6-seco-sterols on the formation of foam cells and the cytotoxic effects of macrophages indicate that the atheronals have biological effects that could lead in vivo to the recruitment, entrapment, dysfunction, and ultimate destruction of macrophages, the major leukocytes in inflammatory artery disease.

The ultimate goals of research for genetic and environmental factors that increase the propensity of a specific protein to misfold are the understanding and treatment of disease states as diverse as atherosclerosis, light-chain deposition disease, systemic amyloidosis, Alzheimer’s disease, and Parkinson’s disease. We showed that the inflammation-derived atheronal-A and atheronal-B trigger a deformation in the secondary structure of the normally folded protein apolipoprotein B-100 into a proamyloidogenic form.

In collaboration with Dr. Kelly, we extended this model and showed that these cholesterol seco-sterols also trigger the misfolding of amyloid β-peptide(1–40), leading to formation of fibrils similar to those observed in patients with Alzheimer’s disease. Interestingly, analysis of the structure-activity relationship revealed that among a panel of aldehydes, only atheronal-A, atheronal-B, and 4-hydroxynonenal trigger misfolding of amyloid β-peptide, suggesting that structural aspects of the aldehyde and not simple protein adduction were critical to this misfolding. More recently, using mutated synthetic sequences of amyloid β-peptide(1–40), we found that the accelerated aggregation of this protein only occurs when a particular lysine of the sequence is modified. We have also shown that atheronals and other lipid aldehydes accelerate the aggregation of several wild-type amyloidogenic proteins, including immunoglobulin light chains, mouse prions, and the tumor suppressor protein p53. The generality and specificity of this process suggest that inflammatory aldehydes and their posttranslational modification of amyloidogenic peptides may be the chemical link between the known associations of inflammation, oxidative damage, and various protein misfolding diseases.

Interaction Between Protozoan J-Binding Protein 1 and Glycosylated DNA

Current treatments of parasitic infections such as leishmaniasis, African trypanosomiasis, and American trypanosomiasis are limited in terms of their effectiveness, increasing drug resistance and the inherent toxic effects of the drugs. Thus, an elucidation of new parasite-specific biological targets for new therapeutic agents is needed. In this regard, the discovery that DNA from members of the order Kinetoplastida, but not from other eukaryotes, contains an unusual modified base, β-D-glucosyl(hydroxymethyl)uracil, called base J (compound 1a in Fig. 3) was a breakthrough. Extracts of several kinetoplastids contain a J-binding protein (JBP) that specifically binds to J-containing duplex DNA. JBP-1 is essential in Leishmania.

Fig. 3. Base J (1a) and base J analogs (1b–1g) synthesized and used as probes for studying JBP-1–J-DNA recognition. Inset, Image of base J in doubled-stranded DNA shows how glucose sits in the major groove.

As a drug target, JBP has merit. The protein shares little homology with other proteins in the Protein Data Bank, and it has a unique ligand, J-DNA containing telomeric stretches of double-stranded DNA, that does not occur in other eukaryotes. However, a preliminary high-throughput screen, focused on disrupting binding between JBP-1 and J-DNA, with a library of compounds consisting of all the major drug pharmacophoric groups revealed no compounds of interest.

In parallel, we have studied the nature of the molecular recognition that underpins JBP-1 recognition of glycosylated DNA. We synthesized a panel of modified J-containing bases (compounds 1b–1g in Fig. 3) and incorporated them into a 16-nucleotide telomeric stretch of double-stranded DNA. In collaboration with D. Millar, Department of Molecular Biology, we determined the dissociation constants of these analogs for JBP-1 and generated a ΔG of binding assessment of each hydroxyl around the glucosyl core. This analysis revealed that some loci around the glycan are essential for binding and some are not. This information is being translated into structure-based design of chemical libraries as inhibitors of JBP-1 binding.

Publications

Bielsch, J., Wang, Q.T., Bosco, D., Powers, E.T., Wentworth, P., Jr., Kelly, J. Inflammatory metabolite-initiated protein misfolding. Acc. Chem. Res., in press.

Bosco, D.A., Fowler, D.M., Zhang, Q., Nieva, J., Powers, E.T., Wentworth, P., Jr., Lerner, R.A., Kelly, J.W. Elevated levels of oxidized cholesterol metabolites in Lewy body disease accelerate α-synuclein fibrilization [published correction appears in Nat. Chem. Biol. 2:346, 2006]. Nat. Chem. Biol. 2:249, 2006.

Nieva, J., Kerwin, L., Wentworth, A.D., Lerner, R.A., Wentworth, P., Jr. Immunoglobulins can utilize riboflavin (vitamin B₂) to activate the antibody-catalyzed water oxidation pathway. Immunol. Lett. 103:33, 2006.

Rogers, C.J., Dickerson, T.J., Wentworth, P., Jr., Janda, K.D. A high-swelling reagent scaffold suitable for use in aqueous and organic solvents. Tetrahedron 61:12140, 2005.

Takeuchi, C., Galve, R., Nieva, J., Witter, D.P., Wentworth, A.D., Troseth, R.P., Lerner, R.A., Wentworth, P., Jr. Proatherogenic effects of the cholesterol ozonolysis products, atheronal-A and atheronal-B. Biochemistry 45:7162, 2006.

Toker, J.D., Tremblay, M., Yli-Kauhaluoma, J., Wentworth, A.D., Zhou, B., Wentworth, P., Jr., Janda, K.D. Exploring the scope of the 29G12 antibody catalyzed 1,3-dipolar cycloaddition reaction. J. Org. Chem. 70:7810, 2005.

Witter, D., Wentworth, P., Jr. The antibody-catalyzed water-oxidation pathway from discovery to an emerging role in health and disease. Antioxid. Redox Signal., in press.

Zhu, X., Wentworth, P., Jr., Kyle, R.A., Lerner, R.A., Wilson, I.A. Cofactor-containing antibodies: crystal structure of the original yellow antibody. Proc. Nat. Acad. Sci. U. S. A. 103:3581., 2006.