The University of California at San Diego is awarded a grant to develop methods for performing functional magnetic resonance imaging (FMRI) experiments in living sharks in order to study their brains as a complete neural network and provide information on how vertebrate neural systems have both structurally and functionally evolved. Understanding this organization and its relationship to shark behavior and ecology has major implications for our understanding of the evolution of vertebrate nervous systems, and the relationship between form and function in the vertebrate brain. In recent years, great advances in human neuroscience have been made using FMRI, where localized changes in brain activity are detected within a sequence of MR images, thus allowing brain activation as a function of time to be measured. In conjunction with advanced methods for non-invasive, high-resolution MRI for quantitating form in chondrichthyan brains, the goal of this project is to develop FMRI capabilities that facilitate the study of the relationship between form and function in these fishes. However, while FMRI in humans is now routine, performing it in sharks requires the development and refinement of both software and hardware to collect reliable, high quality anatomical and functional MRI data in a living partially submerged aquatic specimen. Hence our key objective is to develop these technological advances, and demonstrate the capability of performing FMRI experiments in the shark nervous system.
This project will provide a unique set of methods for studying the relationships between form and function in the shark nervous system, as well as other aquatic model systems, and thus has broad implications for many researchers for whom the link between morphology and function is of great importance, but difficult to explore simultaneously in a well-controlled experimental platform. This project also promotes interdisciplinary collaboration and training in imaging, computation, and data analysis for the morphology, physiology, and marine biology communities. The outcomes of this award including methods, equipment design, and data analysis software will be disseminated to the research community through the Digital Fish Library website (www.digitalfishlibrary.org).
Sharks remain one of the most fascinating groups of animals on Earth and an evolutionary marvel that have developed a highly complex and adaptable brain capable of astounding visual, electrosensory, and olfactory sensitivity. Although years of research into their evolutionary biology has continued to uncover unique structural and behavioral capabilities, little is actually known about their brain function, and in particular about their ability to integrate a wide range of sensory information for performing their spectacular behavioral sensitivity. What has been missing is an experimental method for the in-vivo assessment of spatial and temporal variations in brain activity. Functional magnetic resonance imaging (fMRI) describes a set of in vivo imaging techniques that are now common tools in the imaging of brain activity in humans and animal research models (e.g. non-human primates, rodents, cats, and birds), with these techniques providing insights into the principles of brain activity, pathophysiology of brain functions, as well as brain functional organization and plasticity. The most common fMRI method currently used is blood oxygen level-dependent (BOLD) contrast imaging, which reflects hemodynamic changes in response to brain activation triggered by periodically imposed events. FMRI thus offers the possibility of assessing brain activity in-vivo, but this method has never been attempted in a living shark, in large part due to the significant experimental challenges associated with imaging a live fish in an MRI scanner under reasonably normal conditions. Our goal was to demonstrate that fMRI methods can be extended to the imaging of brain activity in sharks responding to sensory stimulus presentations. The application of fMRI to studies of marine animals raises a number of issues not present in the more established application in humans. For example, in order to stabilize respiration and sense organ function there is the need to maintain a more complex physical environment for these animals, for instance, there is a physiological need to immerse them in seawater while they are inside the scanner. This is not only technically challenging, it can also introduce signal-to-noise issues in the image data if noise is not adequately managed. Perhaps more significantly, fMRI of ‘cold-blooded’ species with less familiar physiology than more typically imaged ‘warm-blooded’ subjects, means that we currently have a very limited understanding of the relationships between brain hemodynamic physiology, brain function, and the measured BOLD signal in our species. The primary goal of this project therefore was the development of a technical methodology for performing fMRI in living sharks that allows non-invasive assessment of brain function in these fish. This was successfully achieved, and involved: 1) The design and construction of experimental equipment that facilitates brain activation in restrained sharks within a human MRI scanner; 2) A shark sensory stimulation system; 3) An optimized BOLD fMRI image acquisition protocol which achieves high resolution, undistorted images of semi-submerged sharks possessing novel anatomy and physiology; and 4) A novel data post-processing methodology for detecting both task-based and resting state FMRI data. This project enabled significant progress towards our ultimate goal of demonstrating functional MRI activity in a living shark, as we were able to demonstrate solid experimental evidence of an FMRI detected brain activation. While irrefutable evidence of the veracity of this data remains inconclusive owing to the low numbers of positive experimental replicates, studies are ongoing to confirm and validate our initial findings. Another major outcome of this project eventuated from the technical difficulties we encountered in acquiring sufficient FMRI data in the presence of external stimuli, i.e. the standard "task based" FMRI paradigm. This led us to consider and also perform FMRI experiments in the absence of external stimuli, so called "resting state" FMRI. This resulted in one of our most important discoveries: a general theoretical and computational formulation for the analysis of space-time data, with a specific application to resting state FMRI data. This discovery has profound implications for the broad and ever expanding range of resting state FMRI experiments now ongoing in human neuroscience experiments.
