Retina neural circuitry seen with particle detector technology

Using particle physics techniques, high energy physics researchers have recently provided new insight into neural circuits inside the retina. After uncovering a new type of retinal cell and mapping how the retina deals with colours, the team from Santa Cruz (US), Krakow and Glasgow is now turning its attention to more complex issues such as how the retina gets wired up and how the brain deals with the signals it receives from the retina. All this using technology derived from high-density, multistrip silicon detectors…

Seen from the point of view of a particle physicist, eyes are image detectors that can gather many different types of data: light and dark, different colours, motion, etc. In particular, the retina, a thin tissue that lines the back of the eye, is a biological pixel detector that detects light and converts it to electrical signals that travel through the optic nerve to the brain. Neurobiologists know that many different cell types are involved in these processes, but they don’t know exactly how many there are, what they all do, or how they interact.

Alan Litke, an experimental particle physicist within the ATLAS collaboration, has been leading a team of physicists and engineers from the high energy physics community which aims to contribute answers to these questions. In collaboration with neuroscientists, this interdisciplinary team has worked to adapt the technology of particle detectors used in high energy physics to study the functioning of the retina. The main challenge is to understand how the retina processes and encodes the information about the outside visual world that it sends to the brain.

The experimental method used by the team involves placing a thin slice of retinal tissue in a chamber filled with a special solution that can keep the tissue alive for several hours. Images generated by a computer are focused on the retina’s photoreceptors that detect the visual stimulus, convert it into electrical signals, and then send it through a network of interconnected neurons for further processing. The output electrical signals from this neural network are then detected by an array of microelectrodes implanted on a glass slide. “The inspiration for our instrumentation techniques came directly from those that were used to build the microstrip silicon detectors in the ATLAS tracker”, explains Alan Litke. “For our first studies, we used arrays of 61 electrodes spaced at 60 µm; however, the cells we were looking for at the time represented such a small fraction of our sample that they were not identifiable in a statistically significant way. Therefore, we decided to design a new array with 512 electrodes, which allowed us to discover a new cell type in 2007”.

For the present study (whose results are published in the 7 October, 2010 issue of Nature), the team had to further improve the array, producing a high-density 519-electrode array with 30 µm electrode spacing, which allowed it to achieve finer spatial resolution and an even higher efficiency. “In collaboration with neurobiologists from the Salk Institute, we succeeded in describing the neural circuits at the resolution of individual neurons and the neural code used by the retina to relay colour information to the brain”, says Litke. “The very high granularity of the array and the possibility of simultaneous recording of signals from hundreds of the retina’s output cells were instrumental in the correct identification of a complete local population of the output cells involved in colour perception”. Technically, this required a miniaturisation process that people at CERN were familiar with. The multichannel integrated circuits to read out the electrical signals were designed by ATLAS member Wladyslaw Dabrowski and his group from the AGH University of Science and Technology in Krakow, and the high-density electrode arrays were developed by Keith Mathieson and Deborah Gunning from the Particle Physics Experiment group at the University of Glasgow. Ex-high energy physicist Alexander Sher, from the University of California, Santa Cruz, was one of the primary authors of the Nature paper.

Given the novel cross-disciplinary aspect of this research, Litke was confronted with the difficulty of obtaining funding to carry out the work. “When we started our studies, one of the core challenges was to get funds because practices in high-energy physics and biology are different. Coming from the world of high energy physics, we had no pilot neurobiology data to justify and validate our methods”, explains Litke. “Our proposal reviewers either did not believe we could develop the technology or did not believe that recording the signals from hundreds of neurons simultaneously was that interesting.” He is hopeful that the situation has now improved as the team has started to produce successful evidence of its approach.

The new technology opens the way to a wide range of possible biomedical applications, including the development of better methods for retinal prosthesis, and the treatment of retinal disease in people suffering from diabetes, thus potentially reducing the risk of blindness. Indeed, in some cases, diabetes can lead to the leakage of blood from tiny blood vessels into the eye. This can be treated with laser surgery. The multielectrode array technique can be used to evaluate and improve the effectiveness of different laser treatments.

So far, Litke and his colleagues have focussed on retinal processing, but they expect their results to be just the tip of the iceberg and that there is still a lot more to discover in this field. “These initial steps in understanding how visual information is processed by the retina do not exhaust the possible applications of our multielectrode array technology”, Litke asserts. “In the future, we plan to apply this technology to other exciting domains: how the complex but precise neural circuitry of the retina develops, and what the brain does with the data it receives from the retina”.


by CERN Bulletin