Chemical Biology of Cellular Protein-Protein and Lipid-Protein Interactions: Application of State-of-Art Chemical, Analytical, Physical, and Computational Tools to Studying Function and Regulation of Proteins in Living Cells

 

Regulation of cellular processes, such as cell signaling, is mediated by the myriads of molecular interactions. Many cellular proteins form dynamic protein-protein interaction networks and protein complexes on or near cell membranes. Cellular protein-protein interactions are tightly regulated in a spatially and temporally specific manner and accumulating evidence shows that membrane lipids play important roles in the regulation of spatiotemporal dynamics and activities of cellular proteins (Figure 1). In general, lipid-protein interactions are highly dynamic and specific processes that are mediated by a wide variety of bioactive lipid molecules that recruit a large number of cellular proteins to various membrane locations. Because it has been well established that misregulation of lipid-mediated protein localization and activation leads to many human diseases, including cancer, diabetes, autoimmune diseases, and inflammatory disorders, it is important to determine the mechanisms of cellular lipid-protein interactions and lipid-mediated protein-protein interactions. We take a multidisciplinary approach, encompassing chemical, analytical, physical, computational, and biological methods, to studying the mechanisms of cellular protein-protein and protein-lipid interactions both on a single-molecule level and on a systems biology level. Specifically, we develop new tools that allow direct and quantitative measurements of complex lipid-protein and protein-protein interactions as well as protein activation in living cells, and design and evaluate new therapeutics and strategies to treat diverse human diseases caused by dysfunction of these processes.

 

 

 

Figure 1. Different models of membrane targeting of signaling proteins and signaling complexes. (A) Membrane recruitment driven primarily by protein-protein interactions. These proteins interact with a receptor or a membrane-associated protein through modular protein-interaction domains, such as the SH2 domain. (B) Membrane recruitment requiring the cooperativity of weak protein-protein and lipid-protein interactions. These proteins, such as Ste5, weakly bind anionic lipids, such as phosphatidylserine or PtdIns(4,5)P₂, through either surface cationic residues or a modular lipid-binding domain. (C) Membrane targeting driven mainly by lipid-protein interactions. These proteins (such as Akt) contain one or more modular domains, such as the PH domain, that binds a signaling lipid, such as PtdIns(3,4,5)P₃, with high affinity. 

 

(1) Bioinformatics-Based Identification and Characterization of Lipid Binding Domains

 

Lipid binding domains are growing members of structural modules that are specialized in stereospecific lipid recognition (Figure 2). Various cellular proteins that are involved in lipid-mediated processes, such as cell signaling and vesicle trafficking, and implicated in human diseases contain a single copy or multiple copies of lipid binding domains which play important roles in membrane translocation and activation of these proteins. We are identifying and characterizing a wide range of lipid binding domains with novel lipid specificities and unique membrane binding modes and determining how they direct the cellular localization and function of the cellular proteins they reside in. Our current research interests include:

 

1. Determination of structures and membrane binding mechanisms of lipid binding domains (C1, C2, PH, ENTH, ANTH, FYVE, PX, BAR, Tubby, and PDZ domains; see Figure 2) by various biophysical techniques, including X-ray crystallography, surface plasmon resonance, lipid monolayer penetration, X-ray reflectivity and EPR analyses, to understand the basis of their cellular actions.

 

2. Bioinformatics-based prediction and identification of new classes of lipid binding domains by our newly developed high-throughput screening.

 

3. Determination of cellular membrane binding mechanisms of these lipid binding domains using various state-of-art fluorescence microscopy techniques.

 

4. Design and evaluation of small molecules that specifically and potently inhibit the membrane binding of lipid binding domains as novel therapeutic agents.

 

 

Figure 2. Proposed membrane binding modes of representative C1, C2, FYVE, PH, PX, ANTH, ENTH, and BAR domains. They show different degrees of membrane penetration, which is important for their cellular function and regulation. ENTH and BAR domains are also known to induce membrane curvature and deformation.

 

(2) Development of new molecular sensors for lipid quantification

 

Various lipid molecules regulate a wide range of cellular processes by serving as site-specific signals that recruit their effector proteins from the cytosol or interacting with membrane proteins. One of unsolved questions in cell regulation is how a single signal, such as a signaling lipid, can specifically and differentially control multiple target proteins. Since the cellular location and concentration of signaling lipids, which are tightly regulated by a series of proteins, appear to be a critical factor for this type of complex cellular regulation, we are developing various lipid sensors and analytical methods for real-time quantitative determination of different lipid concentrations in living cells in a spatially specific manner. These methods should allow quantification of multiple bioactive lipid molecules and should provide new insight into how spatiotemporal dynamics of lipid molecules regulate complex downstream cellular processes. We are currently preparing sensors for diacylglycerol, phosphoinositides and sphingolipids, and developing analytical and computational methods for data acquisition and analysis.

 

(3) Mechanism of membrane deformation by lipid binding domains

 

Mammalian cells constantly exchange materials with the surrounding through endocytosis and exocytosis, and these vesicle trafficking require a large array of proteins that interact with membranes and membrane-associated proteins. Many proteins involved in vesicle trafficking contain lipid binding domains, including ENTH and BAR domains, which not only bind but also deform cell membranes (Figure 2). We are studying the mechanisms by which various ENTH and BAR domains induce membrane deformation, including bilayer tubulation and vesiculation, by means of various real-time fluorescence measurements. These studies should aid in understanding the mechanisms of cellular vesicle trafficking and might also lay the foundation of designing new classes of lipid-based nanomaterials. 

 

(4) Mechanistic studies of cell signaling proteins in living cells

 

There are many pharmacologically important signaling proteins whose regulation depends critically on membrane interactions. They include phospholipases, lipid kinases, lipid phosphatases, and other lipid-dependent enzymes. Since their cellular activities are typically regulated by complex lipid-protein and protein-protein interactions, full understanding of their regulatory mechanisms can only be achieved by studying them in the cellular context. We are therefore developing real-time cellular activity assays for these signaling proteins, which should allow us not only to investigate their cellular regulatory mechanisms but also to screen a library of small molecules in living cells. Our current research interests include (but not limited to):

 

1. Determination of mechanisms by which phospholipases A₂, 5-lipoxygenases, and other enzymes work in concert to mediate the biosynthesis of potent inflammatory mediators, prostaglandins and leukotrienes, and development of new types of anti-inflammatory drugs for asthma and rheumatoid arthritis.

 

2. Determination of mechanisms of membrane binding and activation of multiple protein kinase C (PKC) isoforms that are involved in the myriads of cellular processes and implicated in many human diseases. We are particularly interested in understanding how PKCq, which is uniquely involved in immune T cell activation and is thus a potential drug target for autoimmune diseases, is specifically targeted to the so-called immunological synapse and gets activated. We are applying both conventional and single molecule-based methods to cellular localization and activation studies of PKCq. 

 

3. Determination of regulatory mechanisms of sphingosine kinase 1 that produces a potent mitogen, sphingosine-1-phosphate and is thus an attractive target for anticancer drug development.