Engineering Biologically Active Materials
Summary: Milan Mrksich’s laboratory operates at the interface of materials, biology and chemistry to create model surfaces for the discovery and study of biological principles. Current programs address the development of high throughput methods for studying protein acetylation and glycosylation and the use of extracellular matrix mimics for identifying ligand-receptor interactions in cell adhesion.
My group's interests overlap chemistry, biology, and engineering, with an emphasis on the design and synthesis of biologically active materials and their application to relevant problems in the biological and medical sciences. Much of our work uses self-assembled monolayers of alkanethiolates on gold to prepare model surfaces that are structurally defined yet complex, in that they present combinations of ligands in spatially organized patterns. Our work is directed toward three broad themes.
1. We pioneered the design of "dynamic substrates" that present ligands whose activities can be switched on and off in response to electrical or optical signals, particularly for studies that address the responses of adherent cells to changes in the extracellular matrix (ECM). These ECM mimics have led the way to the discovery of novel ligands that mediate cell adhesion.
2. We have also developed robust surface chemistries for preparing biochip arrays that are compatible with new analytical methods for analyzing the arrays. For example, we pioneered the SAMDI method, which uses mass spectrometry to analyze the arrays, and we have extended this method to the first label-free approach for high-throughput screening, to the functional annotation of recently sequenced genes, and toward an understanding of the networks that regulate protein acetylation.
3. A recent program is creating defined systems for exploring biochemical reactions to understand the role that localization of enzymes and substrates plays in controlling reaction networks.
Mimics of the Extracellular Matrix
Most cells are adherent and must attach to and spread on a protein matrix in order to survive, proliferate, and control signaling processes. This ECM comprises several glycoproteins and glycosoaminoglycans that present a variety of ligands that interact with cell-surface receptors to mediate adhesion and intracellular signaling pathways. The integrin receptors interact with peptide ligands in the matrix and are central to maintaining cell adhesion and coordinating mechanical properties of the cell. The identification of new ligands from the matrix and the cell-surface receptors with which they interact, followed by understanding the signaling events that are triggered by these interactions, remains difficult.
We have advanced the use of self-assembled monolayers to create ECM mimics. These surfaces allow excellent control over the composition and densities of peptide and protein ligands and are also designed to prevent nonspecific adsorption of proteins, giving excellent control over the ligand-receptor interactions that operate between the cell and substrate. This approach has led to several findings of new ECM ligands for cell-surface receptors. For example, we found that the platelet receptor, in addition to binding the canonical RGD ligand, binds peptides having the arginine residue replaced by hydrophobic residues. This finding points to a new strategy for the development of antithrombotic agents, since the AGD peptide can selectively inhibit the aggregation of platelets by fibrinogen without disrupting the adhesion of endothelial cells.
In another example, we used peptide arrays to investigate the basis for α8β1-dependent adhesion to the protein nephronectin and found that the FEI tripeptide is a selective ligand for the α8 integrins and that this ligand binds synergistically with the RGD motif.
We have also used monolayers to characterize the influence of matrix ligands on cellular activities. In one example, we showed that monolayers presenting a high-affinity ligand promote osteogenesis of a mesenchymal stem cell culture, but monolayers having a low density of ligand promote an adipogenesis program. This switch stems from the ability of high-affinity ligands to sustain greater forces applied by the actomysoin cytoskeleton. In related work, we have shown how the shapes of cells can be engineered—using monolayers that are patterned with adhesive ligands—to promote contractility, and therefore osteogenesis, in adherent cells.
These examples show how molecularly defined substrates can be used to identify ligand-receptor interactions that mediate adhesion and to understand the roles these interactions play in regulating cell behavior.
The development of oligonucleotide arrays and the ability to globally profile gene expression in cells have transformed the study of biological processes. The success of DNA arrays has motivated considerable work over the past decade to develop arrays comprising other classes of biomolecules, including peptides, proteins, oligosaccharides, and small molecules. These applications have, however, proven significantly more challenging and still await a robust technology for profiling biochemical activities. One challenge stems from the difficulty in immobilizing molecules and ensuring that they are all active, present at a uniform density across the array, and not confounded by nonspecific interactions of the sample with the biochip. We have developed self-assembled monolayers as a platform for preparing arrays that address these challenges and have shown that they permit quantitative assays of biochemical activity.
A more significant limitation stems from the challenges in detecting the products of enzyme-mediated reactions of the immobilized molecules. The conventional approaches use fluorescent or radioisotopic labels—sometimes in conjunction with antibodies—to observe the products of a biochemical reaction. The labels can interfere with the enzyme activity and give false-positive or -negative results, can lead to substantial development times for new assays, and can prevent the identification of unanticipated activities.
We have contributed to the characterization of biochip arrays by developing a "label-free" approach. We found that the monolayers could be characterized directly with matrix-assisted laser desorption-ionization mass spectrometry, in a technique that is now known as SAMDI mass spectrometry. When the monolayers are irradiated with the laser, it is the bond between the sulfur atom and the gold substrate that is cleaved, releasing the full alkanethiolate (or the analogous disulfide) into the gas phase, where its mass can then be detected. This SAMDI technique provides a straightforward measure of a broad range of enzyme activities—including kinase, protease, methyltransferase, ligase, and others—using a single format that relies on detection of the reaction products according to their change in mass. By avoiding the use of labels, SAMDI is compatible with enzyme substrates that have a greater relevance to endogenous substrates, is applicable to enzyme activities that are otherwise difficult to measure in label-dependent formats, and can identify unanticipated activities.
We have applied the SAMDI method to functionally annotate recently sequenced genomes. For example, we have identified identified 80 putative glycosyltransferases, which were expressed and tested using 7 donors and an array presenting 25 acceptors. This profiling experiment identified several new glycosyltransferases, including one that catalyzes a linkage for which enzymes haven't been available. We have also used SAMDI to analyze the activity of lysine deacetylases on arrays presenting 400 peptide substrates. We have profiled most of the deacetylases and have used the arrays to identify peptides that are selective substrates. These substrates have enabled the profiling of deacetylases in cell lysates and nuclear extracts and are the first example of directly measuring the endogeneous enzyme activities.
Spatiotemporally Engineered Biochemistry
The cell has an elaborate structure that confines molecules, enzymes, and organelles to nonuniform distributions. This situation contrasts with biochemical studies of enzyme and protein activity, where all components of the reaction mixture are present at equilibrium in the sample, and raises the question of the relevance of in vitro studies of enzyme specificity in understanding signaling processes in the cell. In the cell, for example, a kinase enzyme will have no activity for an otherwise active substrate if the two partners are localized to distinct regions of the cell and the kinase may efficiently turn over a poor substrate if the two are colocalized.
We are developing defined in vitro model systems to investigate the extents to which colocalization of enzyme and substrate can alter apparent enzyme specificities. This program is important to providing a framework for interpreting in vitro and in vivo studies of enzyme activity and therefore is fundamental to cell biology. Our first studies are employing monolayers that are patterned with peptide substrates for a kinase and with phosphopeptides that bind to the SH2 adaptor domain of the kinase. In this way, the enzyme can be localized to the monolayer, where the substrate is presented at a higher effective concentration, and the relative kcat/KM constants can be determined with and without colocalization. We are also applying these concepts to understand the roles of the KDAC and KAT enzymes in modifying the histone substrates.
This research is also supported by the National Institutes of Health, the National Science Foundation, and the Department of Defense.
As of May 30, 2012