Overview of Research
Our research is directed towards taking an interdisciplinary approach to generate effective therapeutics for autoimmunity and cancer. This involves a combination of antibody/protein engineering, fluorescence microscopy and in vivo studies in mice. These studies are funded by the National Institutes of Health, National Multiple Sclerosis Society and the Cancer Prevention and Research Institute of Texas (CPRIT). Broadly, the work can be described as follows:
Antibody Engineering for Tumor Targeting
We have ongoing projects in the area of tumor targeting. Several approaches, funded by the Cancer Prevention and Research Institute of Texas (CPRIT) are being used:
we are developing improved, engineered antibodies to target growth factor receptors such as HER2,
we are using state-of-the-art microscopy methods to understand how tumor targets traffic within cells, and how their trafficking is affected by (therapeutic) antibodies that recognize them,
as a collaborative, multi-investigator effort, we are generating new and improved antibodies or antibody-fusions to target exposed phosphatidylserine on tumor cells.
Our overall goal in this area is to use a combination of mechanistic studies and protein engineering to produce a new generation of biologics for the treatment of cancer.
Engineering Antibodies to Modulate Interactions with the MHC Class I-Related Receptor, FcRn
A major interest of our laboratory is to investigate the mechanisms that regulate the transport and concentrations of IgG at different sites in the body. This has significance to understanding how an effective humoral response develops, which in turn relates to multiple aspects of human health (immunodeficiency, pathogen resistance etc.). The successful use of diagnostic and therapeutic antibodies also depends, in part at least, on understanding the factors that regulate their distribution and persistence.
An area of interest in the laboratory is to develop engineered antibodies that are altered in their binding properties for FcRn with the goal of altering antibody dynamics in vivo. Earlier work in our laboratory involved the generation of antibodies that have longer in vivo persistence and transport better across cellular barriers. Such ‘half-life extended’ antibodies are being developed for use as therapeutics in the biopharma industry. A second class of engineered antibodies that we have developed is called Abdegs, for ‘antibodies that enhance IgG degradation’. We are currently investigating the efficacy of Abdegs in treating antibody-mediated autoimmunity in mouse models of arthritis and multiple sclerosis.
The investigation of approaches to induce T cell tolerance in a mouse model of multiple sclerosis is also an active area of research in the laboratory. These studies involve the engineering of antibody Fc fragment-T cell epitope fusion proteins which have different half-lives and intracellular trafficking properties with the aim of correlating these properties with the induction of T cell tolerance in mouse models of multiple sclerosis.
Using Engineered Mice to Map the Functional Sites of FcRn Function
FcRn is expressed in many different cell types in vivo, and for antibody engineering it is important to understand the site of functional activity of this receptor. We have therefore engineered a mouse strain in which FcRn can be knocked out in different cell types using Cre-loxp technology. This is allowing us to identify the body sites and cell types that are important for FcRn function. We are combining these ‘whole body’ studies with a variety of live cell fluorescence imaging methodologies to investigate how FcRn and its (engineered) IgG ligand traffic in cells (see below).
Image Analysis and Single Molecule Microscopy
An important component of our work relates to the development of methodology for image analysis for cellular microscopy. Special emphasis is being placed on the development of image analysis approaches for single molecule microscopy, which allows the properties of individual (protein) molecules to be studied. Due to the low signal to noise ratio that is characteristic of fluorescence microscopy and the quantum limited nature of the detection process, this area presents novel problems of both a theoretical and experimental nature.
A central component of this work has involved an investigation of the accuracy with which a single molecule can be localized using a fluorescence microscope. We have also derived a new resolution criterion for two point sources. This new resolution measure overcomes several deficiencies of classical criteria such as Rayleigh's criterion and has been validated in experimental single molecule studies. Ongoing efforts include the development of parameter estimation algorithms that are of importance for the tracking of fluorescently or quantum dot labeled single molecules in, for example, tubules and vesicles within cells. A fundamental aspect of our work has been to incorporate parameter estimation problems in fluorescence microscopy into a well founded analytical framework.
To be able to carry out advanced fluorescence microscopy experiments we have invested a substantial amount of effort into the building of high performance microscopy imaging stations. Each of the workstations is equipped with several laser lines and multiple cameras that allow the rapid, simultaneous imaging of different fluorophores in individually selectable combinations of widefield and total internal reflection excitation.
We are also actively continuing the development of our imaging approach in which different focal planes can be imaged simultaneously to build up three dimensional, dynamic images of cells. Combining this imaging approach with a multi-color labeling strategy of the cellular proteins allows us, for example, to simultaneously investigate processes on the cell surface using the high sensitivity of total internal reflection microscopy together with the intracellular events that correlate with the membrane events. This has allowed us for the first time to visualize the trafficking pathways from intracellular sorting endosomes to the plasma membrane (exocytosis) and from the plasma membrane to sorting endosomes (endocytosis).
Software development is another active component in our work. Novel microscopy technologies, such as fast and high sensitivity imaging detectors generate new challenges for software design, not least of which is the large amount of data that is being produced. For example, due to the lack of appropriate software, biologists often spend an extraordinary amount of time analyzing acquired imaging data. We are therefore developing software that allows for the efficient acquisition, processing and analysis of the acquired data. (See MIATool for our software package for the analysis of microscopy images and SPRTool for the data analysis of surface plasmon resonance experiments).
Systems Biology and Modeling
The question of the regulation of antibody (IgG) transport and dynamics in the body is not only of fundamental interest in immunology but also has important implications for the use of antibodies in therapy. From the point of view of systems biology this provides a fascinating subject. Information is available on very different time and size scales from studies of antibody dynamics in humans or laboratory animals, over knowledge of the intracellular trafficking pathways, to data related to the interaction dynamics between IgG molecules and the transport/salvage receptor FcRn.
It is possible to use protein engineering techniques to modulate the interactions between IgGs and FcRn. Experimentally, it has been shown that this influences IgG dynamics in vivo and in vitro. We have initiated a project in which we aim to model these effects to gain an understanding of how molecular interactions determine the intracellular trafficking behavior and in vivo dynamics of IgGs.
Biosensor Data Analysis
Additional research interests include the resolution of data analysis problems for the determination of molecular interaction constants using surface plasmon resonance (e.g. BIAcore instruments). We have addressed problems in data analysis that are encountered by the laboratory scientist, in addition to the study of more fundamental questions. The scientist needs to predetermine the experimental conditions and instrument settings such that the estimated interaction parameters have the required accuracy. This frequently has to be done in the context of limitations in protein availability, instrument time etc. Optimal experimental design therefore provides an important challenge in these studies. In a series of publications we have investigated this problem by calculating the Fisher information matrix/Cramer Rao lower bound for a number of model systems that represent typical biological interactions. Please also see SPRTool for our software package to analyze experimental SPR data.