A major goal of our research is to explore the relationship between cerebral water trafficking and brain function. Despite the importance of water to all organisms, there is little understanding of how it specifically affects the brain. Recent studies have suggested that water trafficking, especially as it relates to extra-cellular space (ECS) plays an important role in neuronal activation and sensitivity. (Glia 2006, 54: 358, Nat Med 2005, 11: 973) One of the key arbiters of ECS is the water specific channel Aquaporin 4 (AQP4), and hence it may be expected to play a significant role brain function.
AQP4 was first identified from lung tissue, (J. Biol. Chem. 1994, 269, 5497) and subsequently found to be distributed throughout the body and central nervous system. (Proc. Natl. Acad. Sci. USA 1994, 91, 13052) This protein is highly expressed along the perivascular and subpial end-feet of glial astrocytes, and is now widely accepted as being the principle water transporter in the mammalian brain. In addition to evidence suggesting this protein is important to normal brain function, it also appears to play an adverse role in several medical conditions, particularly those that involve cerebral inflammation. However, the specific nature of those roles, including a detailed understanding of AQP4's significance, have not been identified, and are the focus of this research project.
Two of our key technologies, Positron Emission Tomography (PET) and Magnetic Resonance Imaging (MRI) will be highly useful for studying AQP4 function in vivo. Both methods are non-invasive and can be used to identify real-time structural localization and physiological changes, respectively. However, to be successful, both methods first require the identification of AQP4 ligands with at least an intermediate level of protein affinity and selectivity.
Natural or previously identified ligands would typically be used as starting points for designing those to be used in the PET and MRI studies. However, no natural ligands have been identified for AQP4, and initially no other ligands were known. Therefore, we chose to follow an iterative approach to identifying suitable molecules based on a combination of computational simulation, in vitro bioassay and medicinal chemistry methods. A brief description of these methods is as follows:
Sometimes referred to as in silico screening or virtual docking, computational simulation allows the structural and electronic complementarity between organic molecules and a protein binding site to be explored. The Fujiyoshi laboratory at the University of Kyoto recently reported the solid-state structure of rat AQP4 (J. Molec. Bio. 2006, 355, 628), which in turn provided the atomic coordinates necessary to construct a virtual model of this protein. These studies serve two significant purposes: the first is to prioritize molecules for synthesis or in vitro analysis; and the second is to understand how AQP4 modulators interact with the protein.
In vitro bioassays are used to judge how well an organic molecule is able to interact with a protein target. Functional assays judge the interaction of a specific molecule by monitoring ligand dependent changes of a protein specific function, while binding assays determine the direct affinity of a molecule without regard to functional changes. Several types of functional assays but no binding assays have been proposed for measuring such ligand dependent changes to aquaporin water transport rates. Our laboratory has primarily focused on using changes to the swelling rate of AQP4 expressing Xenopus laevis oocytes under hypotonic stress as a means of assessing an individual molecule's ability to act as a modulator. While this assay method appears to be quite useful, ligands that bind away from the water channel or do not significantly alter AQP4's water transport rate will appear as false negatives. Such ligands may also be useful; therefore, we are in the process of developing an AQP4 compatible binding assay. Functional and binding assays are generally complementary and will allow us to better understand how specific molecules are binding to the protein, and how to optimize those interactions.
Medicinal chemistry is generally comprised of two components: the synthesis of organic molecules with biological activity; and optimization of that biological activity. The first aspect is relatively straight forward, molecules suggested as being useful by computational or biological studies can either be synthesized in-house or purchased through a commercial supplier. The second aspect is conceptually somewhat more difficult, by systematically changing the structure of known active compounds, the molecular features most important to biological activity can be identified and enhanced, while those unimportant ones can be eliminated. Medicinal chemistry plays an essential role in this study as the bridge between theoretical computational studies and practical In vitro biological assays.
Using the methods described above, we have identified a number of compounds active against AQP4. While most of these early compounds were known drugs, several new compounds were also uncovered in those studies. Preliminary reports on our initial descriptions have now been published. (Bioorg Med Chem Lett 2007, 17: 1270, Bioorg Med Chem 2008, Bioorg Med Chem 2008) Subsequent reports will focus on more detailed aspects on those compounds AQP4 inhibitory properties, as well as our efforts to design more optimal ligands.
Center for Integrated Human Brain Science