Brain signals were read by lighting up the genetically encoded activity sensors in the nerve cells.
Prof. Hausser and colleagues engineered nerve cells in the brains of mice to read and write brain signals.
They engineered the same nerve cells to express light-sensitive proteins and to write brain signals that can be activated with flashes of light.
Brain activity was observed and controlled in mice by combining the techniques.
"Just as we combine specific words into sentences that elicit a reply from someone we talk to, we used light to activate specific combinations of nerve cells in the intact brain and record how the other cells respond," he added.
The team found a way to activate several brain cells at the same time. Using a holographic technique, they split a beam of light into smaller beamlets that they directed to individually selected brain cells.
They tested the approach on a group of cortex brain cells that respond to touch. When they activated the chosen neurons with the beamlets, they saw flashes of activity not only in the activated neurons but also in hundreds of their neighbors.
They activated selected brain cells in different patterns and measured how the circuit responded - demonstrating how the technique offered a way to "interrogate" the chosen brain circuit.
The experiments were repeated in the same group of neurons with the same mice over days and weeks, allowing them to have an extended "conversation" with the brain circuit.
The team hopes eventually to crack the "neural code" of sensory perception - the language our brain cells use to tell each other about the information our senses gather from our environment.
Dr. John Isaac, Head of Neuroscience and Mental Health at the Wellcome Trust - a sponsor of the study said,
"This new approach helps us understand how complex behavior is produced by the nervous system.
The work is a step towards realizing one of the ultimate challenges of modern science: understanding how the brain processes information to produce appropriate actions."
The insights gained could also extend beyond understanding neural code - they could reveal how brain activity responds in disorders like autism and dementia.