Fast multiphoton microscope can record firing of neurons in 3-D
David L. Shenkenberg
Neurons receive, process and transmit electrical impulses that underlie how humans think and feel. Classically, electrodes have been used to record these impulses, but even microelectrodes are too big to fit on neuronal dendrites, the source of the signals. Unable to analyze dendrites, neuroscientists have placed the electrodes on the neuron’s cell body to record the electrical signal.
Multiphoton microscopy with activity-dependent molecular optical indicators can monitor the electrical signals from dendrites. Since the invention of multiphoton microscopy in 1989, the technique also has enabled the deepest and least invasive optical imaging of the structure and function of the brain. Now a handful of research groups around the world are racing to develop multiphoton microscopy into a fast three-dimensional imaging technique.
Although a couple of 3-D imaging systems have been reported in the past two years, neither could operate fast enough to capture the firing of neurons in three dimensions. Now members of the Peter Saggau laboratory at Baylor College of Medicine in Houston have developed the first multiphoton microscope that can capture 3-D images of a calcium wave in a single neuron.
To direct the scanning laser beam, the Saggau system uses two pairs of acousto-optic deflectors (AODs) (Figure 1). Acoustic waves in these aptly named devices deflect the beam in a direction that depends on the acoustic frequency.
Figure 1. Acousto-optic deflectors move the scanning laser beam in a patented multiphoton microscopy system. PMT = photomultiplier tube. Images reprinted with permission of Nature Neuroscience.
The configuration of the AODs in the Saggau system enables the beam to hop anywhere. This ability allows the beam to trace any shape in a single pass, which is important because biological components have irregular shapes. In contrast, conventional microscopes scan in straight lines, so these systems must repeatedly scan the same outline of a biological component.
No moving parts
Furthermore, the Saggau system does not contain any moving parts such as mirrors, objectives or stages. Thus, no masses must be accelerated or decelerated. The theoretical positioning time is less than 1 μs/mm of the beam’s diameter.
Moreover, hopping between two points when using the system always takes the same amount of time, independent of the distance. This is in contrast to conventional scanning systems, whereby the fastest route for visiting multiple points must be determined by an algorithm that solves the “traveling salesman problem.”
The new Saggau system is an improved version of one the group designed previously, in which the beam passed through two single AODs offset by 90°. A stepwise change of the acoustic frequencies pivots the beam in the back focal plane of a microscope objective. Thus, the beam focus moves laterally. When rapidly and continuously changing the frequency of the acoustic waves (“chirping”), the AODs also will affect the collimation of the beam, which can move the beam focus axially.
However, in a chirped two-AOD system, the beam moves continuously in the lateral direction, which renders the device unable to dwell in one place to increase the signal-to-noise ratio — a hallmark of AOD-based scanning. The new Saggau system overcomes this limitation by using two pairs of AODs sending acoustic waves in opposite directions. The second AOD in a pair also compensates for the spatial dispersion effect of using ultrashort pulses.
The Saggau laboratory used custom AODs with a wide acceptance angle because the beam hits two deflectors at varying angles, whereas standard AODs can accept incoming beams at only a fixed angle. In all AODs, the beam’s deflection takes place in a crystal. The Saggau laboratory chose tellurium dioxide (TeO2) because it allows for wide apertures with high interaction bandwidth and diffraction efficiency. The AODs were custom-made by Isomet Corp. of Springfield, Va.
Because AODs can receive incoming laser beams of only one polarization and can rotate the orientation by 90°, the deflectors must be arranged in a specific order. The researchers also used a half-wave plate from Thorlabs Inc. of Newton, N.J., to adjust the polarization to the first deflector.
The investigators scanned the neurons with near-infrared pulses from a Coherent Ti:sapphire laser. The laser beam traveled to four AODs, four telescopes, a scan-angle magnifier that increased the lateral scan field and into the back aperture of a Nikon 60×, 1.0-NA water-immersion microscope objective. A Hamamatsu photomultiplier tube was used for fluorescence detection. The entire system can run automatically or semiautomatically on custom software.
In experiments detailed in the June 2008 issue of Nature Neuroscience, the researchers performed structural and functional imaging of neurons in rat brain slices, rapidly scanning many user-selected sites on dendrites.
The system visited 50,000 sites per second to create a 3-D functional image of the neurons, which Saggau says is the fastest 3-D scanning speed achieved by a microscopy system (Figure 2).
Figure 2. Researchers captured three-dimensional images of neurons firing in real time. Color represents axial depth, which is measured in microns relative to the objective lens focus. The system also can stimulate neurons optically.
The scientists found that there is one minor limitation of the system, which is that, with high-power objective lenses, it is restricted to a 50-μm axial scan range, but Saggau said that that limitation is easily overcome by decreasing the magnification.
Currently, the researchers’ setup is laid out on an optical table. Saggau said that if they licensed their patented technology to a company, it probably would be sold as a turnkey boxed instrument to be used with a commercial research microscope. He estimates that such a system, excluding the pulsed laser, would cost between $100,000 and $200,000, which is comparable to other accessory scanners for microscopes.
In collaborative work, the scientists will use their microscopy system to image neurons in the intact brains of mice. Because their system not only can image but also optically stimulate neurons rapidly in three dimensions, the investigators also plan to use their system to activate the acoustic nerve selectively in mice in research that they hope will one day restore hearing to humans who suffer from deafness.
Contact: Peter Saggau, Baylor College of Medicine, Houston; e-mail: firstname.lastname@example.org.
MORE FROM PHOTONICS MEDIA