Diamonds reveal neuronal secrets

The brain is arguably one of the most complex structures in the known universe.

Continued advances in our understanding of the brain and our ability to effectively treat a variety of neurological disorders have come from studying the brain’s neural microcircuits in increasing detail.

A class of techniques for examining neural circuits is referred to as voltage imaging. These techniques allow us to see the voltage generated by our brain’s firing neurons and tell us how neural networks develop, function, and change over time.

Today, voltage imaging of cultured neurons is performed with dense arrays of electrodes on which cells are grown (or cultured), or by applying light-emitting dyes that optically respond to voltage changes on the cell’s surface.

But the level of detail we can see with these techniques is limited.

The smallest electrodes cannot reliably distinguish individual neurons, about 20 millionths of a meter across, let alone the dense network of nanoscale connections that form between them, and no significant technological advances have been made in this area for over two decades.

In addition, each electrode requires its own cable connection and amplifier, severely limiting the number of electrodes that can be measured simultaneously.

Dyes can overcome these limitations by wirelessly imaging the voltage as light — meaning the complex electronics in a camera can be placed away from the cells.

The result is high resolution over large areas, capable of distinguishing each individual neuron in a large network. But again there are limitations, the voltage responses of prior art dyes are slow and unstable.

Our latest research, published in nature photonicsis exploring a new breed of high-speed, high-resolution, and scalable stress imaging platform designed to overcome these limitations – a diamond stress imaging microscope.

Developed by a team of physicists from the University of Melbourne and RMIT University, the device uses a diamond-based sensor that converts voltage signals on its surface directly into optical signals – meaning we can see electrical activity as it takes place.

The transformation exploits the properties of an atomic-level defect in diamond’s crystal structure known as a nitrogen vacancy (NV).

NV defects can be constructed by bombarding the diamond with a nitrogen ion beam using a special type of particle accelerator. Fabrication of the sensor starts with using this process to create a high-density, ultra-thin layer of NV defects near the diamond surface.

You can think of each NV defect as a bucket containing up to two electrons. When this bucket is empty, the NV defect is dark. With an electron, the NV defect emits orange light when illuminated by a laser—a property called fluorescence. With two electrons, the color of the fluorescence becomes red.

A previously discovered property of NV defects is that the number of electrons they contain – and the resulting fluorescence – can be controlled with a voltage. Unlike dyes, the voltage response of an NV defect is very fast and stable.

Our research aims to overcome the challenge of making this effect sensitive enough to neural imaging activity.

On the diamond’s surface, the crystal structure ends with a one atom thick layer of hydrogen and oxygen atoms. The NV defects closest to the surface are most sensitive to stress changes outside of the diamond, but they are also very sensitive to the atomic makeup of the surface layer.

Too much hydrogen and the NVs are so dark that the optical signals we are looking for cannot be seen. Too little hydrogen and the NVs are so bright that the small signals we are looking for are completely washed out.

So there is a “Goldilocks zone” for voltage imaging where the surface contains just the right amount of hydrogen.

To reach this zone, our team developed an electrochemical process for controlled hydrogen removal. This has enabled us to achieve voltage sensitivities that are two orders of magnitude better than previously reported.

We tested our sensor in salt water using a microscopic wire 10 times thinner than a human hair. By applying a current, the wire can create a small cloud of charge in the water above the diamond. The formation and subsequent diffusion of this charge cloud creates small stresses on the diamond surface.

By capturing these voltages through high-speed recording of NV fluorescence, we can determine the speed, sensitivity, and resolution of our diamond imaging chip.

We were able to further increase the sensitivity by structuring the surface of the diamond into ‘nanopillars’ – conical structures with the NV centers embedded in the tips. These pillars direct the light emitted by the NVs to the camera, dramatically increasing the amount of signal we can capture.

With the development of the Diamond Voltage Imaging microscope to detect neuronal activity, the next step is to record the activity of cultured neurons in vitro – these are experiments on cells grown outside of their normal biological context, also known as test tube or petri dish experiments .

What distinguishes this technology from existing state-of-the-art in vitro techniques is the combination of high spatial resolution (of the order of one-millionth of a meter or less), large spatial scale (a few millimeters in each direction – e.g. a network of mammalian neurons is quite large) and complete stability over time.

No other existing system can offer these three qualities simultaneously, and it is this combination that will allow our Melbourne-made technology to make a valuable contribution to the work of neuroscientists and neuropharmacologists worldwide.

Our system will support these researchers to develop both fundamental knowledge and the next generation of treatments for neurological and neurodegenerative diseases.