Ramkumar Sabesan, Ph.D. - US grants

Project summary Photoreceptor death due to retinal degenerations is the leading cause of blindness in the developed world. Recent approaches to treat these blinding diseases include re-conferring the light sensitivity of the degenerated retina, either by genetic manipulation or stem-cell transplants of photoreceptors. The fundamental impediment to translating these treatments to the living human retina is the ability to visualize and manipulate their functional mechanisms at a cellular scale in vivo. Efforts to functionally probe photoreceptor physiology have relied mainly on bulk measures (electroretinograms or ERG) in vivo, while finer spatial scales are accessible only ex vivo via electrophysiology (patch clamp or suction-electrode recordings) and are thus unsuitable for clinical use. We propose a novel approach rooted in classical interferometry to record how photoreceptors interact with light on a cellular scale. The transduction of photons to electrical signals is well- characterized by a set of biochemical molecular events in cones and is also accompanied with a change in physical structure at nm/ms spatiotemporal resolution. These physical changes can be reliably encoded in the phase of the interferometric signal emanating from the photoreceptors when light of low coherence is used to image them. In Aim 1, we will validate a parallel, phase-resolved optical coherence tomography system(OCT) ? essentially a low-coherence interferometer ? to image the light-driven optophysiology activity in cones. We have implemented a free-space OCT system that is based on line-field, spectral domain operation. This allows an entire phase-stable B-scan to be obtained in a single camera snapshot allowing the probing of multiple cells in parallel. Further, with the use of high-speed aerial sensors, extremely fast volume rates are obtainable. We will first validate our approach by measuring the light-driven optical activity in zebrafish photoreceptors in vivo and compare them against Gnat2-/- mutants which do not exhibit the traditional phototransduction cascade. These measurements will categorize the light-driven optophysiological signal on the basis of those arising from light-opsin interaction vs. those arising due to phototransduction. In Aim 2, we will translate the OCT platform to living humans by combining it with an existing adaptive optics scanning laser ophthalmoscope (AOSLO), such that cone activity can be monitored by individually stimulating them in isolation. A power spectrum detailing the dependence of phase sensitivity on eye motion will be estimated by obtaining high fidelity motion traces from the AOSLO. The application of this technology will be validated in humans by probing a few hallmarks of the trichromatic cone mosaic - spectral topography & spectral sensitivity - and comparing them against retinal densitometry and single-unit recordings respectively. These multiple stages of validation will set the stage for a wider application of this technology to clinical and basic research thus potentially transforming mechanistic studies of retinal circuitry and its diseases.

Project Summary Vision is the result of billions of neurons working in tandem to extract ecologically meaningful information from the external world. In contrast, only a few million cone photoreceptors serve as the gatekeepers for vision, by initiating light absorption and converting it into a language that the rest of the visual system can decode. In the retina, cone photoreceptors are also among the most vulnerable to disease. If therapies aimed at their rescue are to evolve in the future and restore normal vision with all its exquisite features, the underlying neural substrates for vision need to be detailed on a cellular scale. The properties of the trichromatic cone mosaic pose few well recognized ambiguities for visual processing. The photoisomerization in one cone alone, for instance, can arise from numerous combinations of intensity and wavelength leaving the visual system to rely on comparisons in postreceptoral circuitry to divorce these two elementary aspects of physical stimuli. Postreceptoral pathways encode intensity and wavelength variations in the retinal image by comparing trichromatic cone signals in a local region of space, the mechanisms of which remain mysterious. This, yet unknown, spatiochromatic code initiated at the level of the cone mosaic is inherited by, and consequently constrains the information available to downstream neurons responsible for decoding chromatic and achromatic properties of the visual scene. The lack of information on the natural variation in cone topography within and between individuals is a source of ambiguity for models of spectral processing in downstream neurons. Furthermore, the general lack of tools to directly link the outputs of the cones and their ensuing circuits onto behavior has hindered progress in outlining the neural substrates for color appearance and detection. We have recently developed tools to a) efficiently map the topography of the cone mosaic with adaptive optics assisted densitometry and b) test the visual sensations elicited by targeted stimulation of the retina with help of cellular-scale eye tracking. With knowledge of the spectral organization in the central retina across a range of individuals, we will establish the genetic and developmental mechanisms shaping the adult human retina in Aim 1. By undertaking concomitant chromatic and luminance detection measurements in the same retinae with optical aberrations removed, we will outline the postreceptoral wiring strategies that dictate the underlying limits for these classical perceptual tasks. In Aim 2, we will map the output of individual and collection of cone cells of known spectral type onto perception. We will first test hypotheses that characterize the rules by which cone signals are integrated to mediate detection and appearance. Next, we will detail the spatial grain of color signaling across the central retina and test whether they fall in line with standard models of center-surround opponency. Together, this work will lay the foundation for computational models of visual processing, establish a new line of experiments to test model predictions linking physiology and perception; and eventually set the stage for a wider application of these tools to cellular-scale behavioral testing in retinal disease and their therapies.

Project Summary/Abstract The last few decades have seen major inroads into detailing the physiological mechanisms supporting vision as well as therapies aimed at rescue and repair of neurons affected by retinal diseases. For the continued evolution of treatments and their rapid translation to the clinic, it is essential to find a non-invasive, all-optical biomarker to monitor the efficacy of disease and potential therapeutic agents. To this end, we propose to develop the optoretinogram, or ORG, the optical analog to the electroretinogram (ERG) which is the current gold standard for retinal function assessment in humans. The ORG is rooted in classical interferometry and enables a highly sensitive assay of how neurons interact with light. Using this technique, our group has demonstrated the ability to visualize light-driven neural activity across a range of spatiotemporal resolution ? from single cells to a collection of neurons, and from µsecs to ms timescales. Here, we aim to expand the capabilities of the ORG and demonstrate its efficacy for basic science and clinical applications. The proposed technology is built upon a solid foundation of established approaches, and combines them in new and complementary ways to achieve an optimal combination of speed, resolution and sensitivity geared towards overcoming the key challenges faced with imaging cellular structure-function in humans. The core technologies are phase-resolved OCT, an eye-safe, interferometric method to measure nm-scale changes at ms time scales in vivo, adaptive optics (AO) to overcome ocular aberrations, increase the signal-to-noise and allow resolution down to single cells and real-time eye tracking to overcome eye motion and allow targeting, recording and averaging of responses from single and a collection of retinal neurons. These are implemented across three ORG platforms. At the University of Washington, we will refine the line-scan phase-resolved OCT with improvements in optical design and eye-tracking and use it to characterize the basic properties of phototransduction and inner retinal function in healthy human volunteers and patients with retinal degenerations. At Stanford University, we will develop a similar line-scan system for rodents, and together with transgenic models and pharmacology, determine the biophysical mechanisms that underlie the ORG and develop templates for human recordings. At UC Berkeley, we will push the envelope of speed and sensitivity by incorporating a real-time eye-tracking system to drive an AO-OCT interferometric probe, with the aim to measure the tiniest and briefest neuronal changes in the human retina. This bioengineering research partnership will benefit from complementary expertise, research direction and ORG implementation across the three sites, and the use of common approaches for image/data analysis, eye tracking and visual stimulation. Ultimately, the aggregate technology will facilitate a deeper mechanistic understanding of early visual processing and eye disease, and provide entirely new avenues for accelerating therapeutic interventions.