Icosahedral coat protein assemblies, by definition, are highly regular structures, presenting multiple copies of identical functionality on their external and internal surfaces. Dendrimers, polymers, metal nanoparticles, and cells all have some of these qualities, but viruses are unique in the size and regularity of their molecular structures. No other entities of comparable dimensions are available with structures known and controllable to atomic resolution. Such well-defined "polyvalency" is crucial to the biological function of viruses and of critical importance to our efforts to use them in many different applications.

Shown on the right are CPMV particles with inserted cysteine residues (red dots) providing attachment points in different patterns.

At the present time, research is ongoing in the following areas:

Chemical Fundamentals.   New methods of chemical derivatization of proteins and virions; enhancing particle stability; learning the chemistry of new viruses particles.

Polyvalent Presentation to Biology.  Attachment of biologically active compounds, from small molecules to proteins, to virus scaffolds and exploration of their impact on biological systems. Work includes the display of carbohydrate, peptide, and drug molecules, and characterizing interactions with cellular receptors and tissues.

Catalysis.  Arraying of organometallic and organic catalytic engines on virus particles, and use of the tunable virus structure to influence catalyst activity. Such systems are easy to separate from reaction products and provide high local concentrations of catalysts in otherwise dilute solutions, which creates opportunities for interesting reaction cascades and therapeutic applications.

Nanoparticle Assemblies and Materials Science.  Programmed self-assembly of virus particles in two and three dimensions; hybrid materials incorporating virus building blocks, metal and ceramic nanoparticles, and organic polymers. The goals here encompass aspects of nanotechnology, sensors, and molecular actuators.

Selection and Evolution.  Creation of virus libraries and selection for function, retaining the property of high expression yields.

Our collaborators, with whom we greatly enjoy working, include:

      John E. Johnson and Tianwei Lin (TSRI) - all aspects of virus production and structure
      Anette Schneemann (TSRI) - Flock House Virus
      Marianne Manchester (TSRI) - virus targeting to cells and tissues
      Sandra Schmid (TSRI) - receptor-mediated endocytosis
      Glen Nemerow (TSRI) - peptide-directed viral entry mechanisms
      Dorian McGavern (TSRI) - molecular mechanisms of immunology
     
Other laboratories. For other approaches toward the use of biological protein assemblies (including viruses) in chemistry, see the work of the following research groups: Trevor Douglas and Mark Young (Montana State University), Angela Belcher (M.I.T.), and Matt Francis (U.C. Berkeley).

"Click Chemistry". As defined by K.B. Sharpless, "click" chemistry involves the use of only the most reliable, general, high-yielding, and byproduct-free organic reactions for the construction of compounds with desired function. Its central hypothesis is that most, if not all, chemical functions can be attained by many different compounds among the nearly limitless possibilities of three-dimensional structural space. If you restrict yourself as much as possible to using only the best reactions, more diversity will be accessible, since your reactions work with a wider variety of pieces. One stands a better chance, therefore, of finding functional molecules this way, although they will not look like the compounds made by nature for the same purpose.

To appreciate the possibilities, synthetic chemists might ask themselves “what if every bond connection worked flawlessly, regardless of the structure of the connecting pieces?” Non-practitioners of the synthetic art, might wonder what would be possible if molecules of any desired shape or pattern of functional group display were available without much trouble. We are not yet close to such ideals, but the concept is powerful and enabling.

Click reactions tend to involve high-energy (“spring-loaded”) reagents with well-defined reaction coordinates, giving rise to selective bond-forming events of wide scope. Examples include the nucleophilic trapping of strained-ring electrophiles (epoxide, aziridines, aziridinium ions, episulfonium ions), certain forms of carbonyl reactivity (aldehydes + hydrazines or hydroxylamines, for example), and several types of cycloaddition reactions. The azide-alkyne 1,3-dipolar cycloaddition has emerged as the most important. Our collaborative efforts using click chemistry include three projects.

Synthesis of Drug Candidates: in situ Formation of Enzyme Inhibitors (with Profs. K. Barry Sharpless, Valery Fokin, and Hartmuth Kolb of TSRI). The 1,2,3-triazole-forming azide-alkyne cycloaddition involves a pair of groups which are uniquely reactive with each other and unreactive with nearly all other structures in biology. It thus provides a way to stitch pieces together in the presence of biomolecules and even living organisms. We have exploited this capability by asking an enzyme to assemble its own inhibitor in situ. Acetylcholinesterase has two known small-molecule binding sites at either end of a narrow gorge. When presented with pairs of binding pieces to which azide and alkyne were attached, one combination was assembled into a bivalent triazole in the enzyme binding pocket.

This compound, shown in the structure on the left (thanks to Dr. Flavio Grynszpan for the image), is a femtomolar inhibitor of the enzyme, binding more strongly than the best previously known noncovalent inhibitors (including snake venom toxin) by two orders of magnitude. More exciting still has been follow-up x-ray structural studies which show that such strong binding is accompanied by an unprecedented change in enzyme structure -- a true example of the induced fit model of enzyme inhibition.

These findings demonstrate that the "kinetic capture" of such a compound by azide-alkyne cycloaddition gave rise to a structure that could not have been anticipated by even the most sophisticated molecular modeling, since all such modeling is based on known structural forms of the enzyme.

We are currently using these techniques in collaboration with the laboratory of Prof. Palmer Taylor at UCSD in a search for selective small-molecule agonists and antagonists of acetylcholine binding proteins. We are also engaged in a collaborative search for inhibitors to HIV protease and its many clinically important mutants. In these cases, the protein targets are multi-subunit entities with bivalent or multivalent binding sites created at protein-protein interfaces. Such systems - ubiquitous in biochemistry - are natural targets for the click chemistry approach.

Materials Synthesis with "Click Chemistry" (with Profs. K. Barry Sharpless and Valery Fokin of TSRI).

The branch of chemistry most heavily dependent on reactions which meet the click chemistry standard is polymer synthesis. Indeed, the reliability of polymerization reactions and the rich functions of the products were among the original inspirations for the click chemistry concept. The identification of each new click reaction immediately enables the synthesis of novel materials.

Prof. Valery Fokin and coworkers of TSRI discovered in 2002 that Cu(I) accelerates the azide-alkyne cycloaddition to a remarkable degree in aqueous solutions, establishing the process as the cream of the click chemistry crop. We are employing it in dendrimer chemistry and template polymerization and intend to significantly expand this part of our research. As shown on the next page, the production of adhesives provides the best example thus far of a functional material made with this process.

Since copper metal provides Cu(I) ions which catalyze the reaction, and the product triazoles are known to bind well to metallic surfaces, mixtures of polyvalent azides and alkynes comprise adhesives which glue copper to itself or other metals in a cycloaddition curing process. The resulting materials are much stronger than commercial metal adhesives.

The picture (shown on the left) is a test of the strength of the bond formed between two copper plates, in which more than 30 killograms is supported by 50 milligrams of adhesive. The drum is filled with water, and additional load is provided by the weights stacked on the drum and chained to its sides. The inset at the top right shows a close-up view of the copper plates.

Tunable Electrophiles. In addition to their obvious importance to synthetic chemistry, electrophiles stable to water but reactive with protein side chains are highly useful as probes and potential covalent inhibitors of enzymatic function. We currently focus on two families of electrophiles with interesting reactivity.

Shown on the right (top) is pictured the class of compounds available from 1,5-cyclooctadiene which react with nucleophiles via anchimeric assistance, with efficiencies that place them in the "click" category. We are investigating both their physical organic chemistry and their application to the synthesis of functional compounds.

Shown on the right (bottom) is our recent discovery of a new and convenient access to formamidine ureas, which take on heteroatom nucleophiles at the formamidine carbon in exchange reactions. Interest here centers on their pharmacophoric properties and their controlled electrophilicities, which can be tuned by changing the steric and electronic nature of their substituents

 

 

We have two main projects in the development of new mass spectrometric techniques: (a) analysis of chiral compounds, and (b) porous silicon based desorption/ionization. Each bears its own clever acronym.

Mass Spectrometry Enantiomeric Excess Determination (MSEED). Combinatorial approaches to catalyst discovery are centrally dependent on efficient methods for determining catalytic activity and enantioselectivity. Responding to this challenge, we have developed kinetic resolution-based methods for the rapid determination of enantiomeric excess of many types of organic compounds using mass spectrometry.

                                                      

As shown in general form here, the technique employs enantiopure reagents (R-Y, S-Y) of different mass but designed to react as enantiomers with the chiral analyte of interest (R-X, S-X). The presence of even a slight degree of kinetic resolution (k₁ not equal to k₂) allows the determination of the ratio of analyte enantiomers to ±10% ee by measurement of the ratio of product masses from the probe reaction. For calibration, a sample of the racemic analyte and a sample of known enantiomeric excess are required.

The method is suited to initial screening of candidate catalysts, since neither chiral chromatography nor purification of reaction mixtures is required. Catalyst substrates do not need to be tagged with chromophores or other labels. Different substrates can be processed at the same time, since no analytical optimization is required and only the masses of interest are important - all other peaks in the mass spectrum can be ignored. We continue to develop these techniques, which are currently being extended to chiral epoxides, ketones, aldehydes, olefins, alkynes, and azides.

Desorption/Ionization on Silicon (DIOS). In 1999, Prof. Gary Siuzdak and Dr. Jillian Buriak of TSRI reported that the use of porous silicon (pSi) instead of a gold plate allowed the analysis of small molecules using MALDI (matrix-assisted laser desorption/ionization) instrumentation.