by Mike May

In neurobiology, crucial mechanisms work at various scales, from molecular interactions between two cells to information traveling through networks of many neurons. Using a variety of imaging techniques, scientists can already probe the molecular interactions between neurons, and various approaches to recording give some idea of how information moves through networks. To really figure out how things work, though, scientists would like to simultaneously explore both the subcellular and network levels. A system called PhotoMEA can do just that.

In 2004, Giancarlo Ferrigno a professor in the bioengineering department at the Politecnico di Milano in Italy suggested to his graduate student, Diego Ghezzi, that optical technology could be used to explore neural networks at various spatial scales. So, Ghezzi considered ways to stimulate and detect neuronal signals with light. He knew that a microelectrode array often abbreviated as MEA could stimulate and record electrically from many spots. In addition, voltage-sensitive dyes can track network activity with light. Ghezzi wanted to combine those ideas, but also improve the spatial resolution. Developing PhotoMEA depended on a team of scientists from the bioengineering department (including Alessandra Pedrocchi, Sara Mantero, and Giancarlo Ferrigno), professor of physics Giulio Cerullo, professor of nuclear engineering Alberto Fazzi, and Andrea Menegon and Flavia Valtora, both of the experimental neuropharmacology unit at the San Raffaele Scientific Institute, which is also in Milan.

Ghezzi decided to build an optical stimulating and recording array to study networks of neurons in culture. He says, “Our idea was to exchange the electrodes with optical devices, to stimulate using light and to detect activity with light.” He used waveguides to control where the stimulating light would go, but light alone will not turn on a neuron. So Ghezzi turned to caged compounds molecules that are inactive because they are connected to other chemicals, the so-called cages. Those connections, though, can be broken with a brief flash of near-ultraviolet light. So by adding a caged neurotransmitter such as the caged glutamate that Ghezzi has used so far to a culture of neurons and then directing light to a specific neuron, it can be turned on.

That light-stimulating system gives Ghezzi fine control over where he stimulates a network of neurons. With a typical microelectrode array, says Ghezzi, stimulating current can spread over a couple hundred microns. Optical pulses, on the other hand, can be limited to 10 microns or less. “This is very fine spatial control,” Ghezzi says.

That turns on the neurons, but how does Ghezzi record the results? For that step, he uses a voltage-sensitive, fluorescent dye, Di-4-ANEPPS. Such dyes change their transmission frequency according to the voltage that they experience, such as a neuron’s membrane potential. So, as that potential varies, the dye changes color. Ghezzi uses Di-4-ANEPPS because it’s so common, but virtually any voltage-sensitive dye would work.

Usually, scientists record voltage-sensitive dyes with a CCD camera, but Ghezzi wanted something faster. So he selected a CMOS camera, a semiconductor-based device, that can record up to 2,000 frames per second with a large sensor area up to one-million pixels. So instead of using a CCD camera and a microscope, Ghezzi will incorporate the CMOS-chip detector into a substrate for a culture of neurons.

So far, Ghezzi can show that his idea works in principle. He already built an experimental set-up that works through stimulating waveguides and caged compounds, and it also records neural activity with a voltage-sensitive dye and a CMOS detector. Nonetheless, lots of work remains. Ghezzi wants to build a commercial product from his experimental device. In fact, Ghezzi and his colleagues are awaiting the patent on this technology.

Today, Ghezzi focuses on integrating this technology into a refined package. “We have optical waveguides in glass,” he says, “and have used that to stimulate neurons.” To put that operation on a chip, he has to make some fundamental decisions, such as what diameter the waveguides should be. For that, he’s still deciding. His CMOS camera also works, but incorporating that into the chip-based product will be a significant challenge, says Ghezzi. He plans to make the sensing side of PhotoMEA provide high spatial resolution, leaving just 5 microns between each photodiode of the integrated detector. “This will provide subneuron resolution,” Ghezzi says.

To get from his mock-up to a product will probably take a few years maybe longer. Still, Ghezzi hopes to be well on his way when he completes his Ph.D. in 2008. He says, “I think we will have a complete prototype by then maybe something with 5 to 10 waveguides and maybe hundreds of detectors.” With such a device, Ghezzi can stimulate precisely and record across a network’s expanse. In this culture of neurons from a rat’s hippocampus, caged glutamate was turned on with a flash of light, and Di-4-ANEPPS a voltage-sensitive dye revealed the subsequent activity in this network.

Looking Backward For Diego Ghezzi, the PhotoMEA already provides a powerful tool to study back-propagating action potentials. In general, neurobiologists think of action potentials moving down a neuron, from the neuron’s cell body to the synaptic terminals, but action potentials can also go the other way. Some action potentials move back into the dendrites, the cell-body projections, which can modulate an entire neuron’s activity. Diego Ghezzi “Our idea for PhotoMEA,” says Ghezzi, “is to study the effect of a single compartment of the neuron on the behavior of a neural net.” So, for example, he wants to study how action potentials moving into the dendrites impact that neuron and others around it. He can also add toxins or potential drugs to the culture, and then see how the network’s activity changes.