Category Archives: Imaging

A New Toy for the Optical Crowd

Increasingly, neuroscientists are using optical techniques to study neurons in the laboratory. The latest installment in their love affair with light is the discovery of light-sensitive ion channels called channelrhodopsins. Scientists have genetically altered neurons to express a channelrhodopsin called ChR2, which was originally isolated from algae. Shining light on these neurons causes positive ions to enter the cell, which depolarizes the neurons and triggers action potentials. Targeting ChR2 to specific types of neurons has allowed researchers to control the behavior of animals with fiber-optics. The New York Times picked up on these developments and recently published a nice review of the field.

Stimulating a specific class of neurons with channelrhodopsins can reveal the role of those neurons in neural circuits. But what if researchers want to test how different types of neurons interact? This would require different types of channelrhodopsins that are sensitive to different wavelengths of light. A new finding from the laboratory of Dr. Karl Deisseroth suggests that researchers may eventually have a whole color palate of channelrhodopsins at their disposal.

The paper, published in the latest issue of Nature Neuroscience, reports the discovery of a novel channelrhodopsin (VChR1) that responds to longer wavelengths of light than ChR2. The researchers scanned a genomic database to find microbial genes that resembled those coding for known channelrhodopsins. They tested the channel’s properties in Xenopus oocytes and HEK293 cells and confirmed that it was indeed a light-gated ion channel with an excitation spectrum distinct from ChR2. Then, by driving the gene with a CAMKII promoter, the researchers were able to express the protein in neurons and show that they could trigger action potentials with light.

As I mentioned, the real goal here is to use two wavelengths of light to selectively excite two types of neurons in the same preparation. Unfortunately, there is enough overlap in the excitation spectrums of ChR2 and VChR1 to make selective stimulation difficult. However, molecular refinement may eventually yield versions of these proteins with sufficiently distinct excitation profiles. Furthermore, the paper serves as a proof-of-concept for using bioinformatic tools to discover new channelrhodopsins.


Glue Sniffing and Time Stamping with Roger Tsien

Yesterday I went to a UPENN neuroscience retreat at The College of Physicians of Philadelphia, which happens to be the same building that houses the Mütter Museum. The highlight was an introductory lecture by Dr. Roger Tsien, inventor of calcium-sensitive dyes and Nobel favorite.

Recently his lab developed two novel techniques for imaging brain activity. First, Tsien discussed the Glutamate Sensitive Fluorescent Reporter (GluSnFR aka “Glue Sniffer”). The probe works by Fluroresence Resonance Energy Transfer (FRET). This technique involves two fluoresecent proteins that are matched so that the emission wavelength of the first is the excitation wavelength of the second. If the molecules are close together and oriented properly, then exciting the first protein (say, one that normally fluoresces blue) will in turn excite the second protein (say, one that normally fluoresces yellow). When the fluorescent proteins are separated, only blue light is emitted because FRET cannot take place. When they are oriented in the right way, only yellow light is emitted because all the energy is transferred via FRET.

GluSnFR works by linking blue and yellow fluorescent proteins with a third protein that changes conformation when it binds glutamate. Normally the two fluorescent proteins are close enough to engage in FRET, so excitation results in only yellow light. But glutamate released from neurons can bind the linker domain in GluSnFR and disrupt FRET, causing only blue light to be emitted. The ratio of blue to yellow light emission can be measured with high spatial resolution, facilitating time-lapse movies of glutamate spillover from synapses. This is a cool new tool for imaging neuronal activity. The graduate student who performed the work has a better explanation than I can provide. Paper here.

Tsien began the second half of his lecture by discussing competing theories for memory storage in the brain. The dominant theory has been that learning involves modulation of synaptic strength. However, Tsien stressed new evidence showing that learning depends on the formation of new synapses. TimeSTAMP is a new way of monitoring synapse formation.

Tsien’s lab engineered animals to express modified versions of the proteins normally expressed at the synapse, such as PSD-95. This version of the protein is linked to a hemagglutinin (HA) tag via another protein. The linker protein is actually a cis-acting protease, meaning that it spontaneously cleaves itself. Normally, the linker cleaves immediately after PSD-95 is translated, thus separating it from the HA tag. But when a protease inhibitor is added, all the newly translated PSD-95 will have the HA tag. After the experiment is over, the brain can be stained with anti-HA antibodies to see only the synapses that formed after the addition of the protease inhibitor. The ultimate goal is to administer the inhibitor before a learning experience and then observe the when and where of synapse formation compared to an animal that did not have that learning experience.

TimeSTAMP isn’t quite as cool as GluSnFR becasue the results are obtained retroactively with immunohistology instead of real time optical imaging. However, this approach is advantageous because it allows you to view the whole brain instead of just the superficial areas that light can penetrate. Another problem is that there may be high turnover of proteins like PSD-95 even at old synapses. The perfect marker will use a synaptic protein that is only translated during synapse formation, although it is unclear whether such a master molecule exists.

Needless to say, Tsien’s research is pretty awesome.