One neuron can make a difference, especially if it sheds light on how its billions of neighbors conspire to form functional neural circuits in the brain. In the December 23 Nature Methods online, researchers led by Michael Häusser, University College London, England, report a new technique for measuring electrical activity in single neurons in the living brain. Unlike currently used methods, which rely on expression of transgenic fluorescent proteins to identify single cells, this one is applicable to wild-type cells. Called “shadowpatching,” the technique can also be used for electroporation and is poised to be a valuable tool in the study of neuronal circuitry and plasticity. “The technique allows you to look at the effect of a single gene in a single cell, and do it in small, defined populations,” said Häusser in an interview with ARF. The work was carried out in collaboration with Winfried Denk’s lab at the Max Planck Institute for Medical Research, Heidelberg, Germany, and Masanobu Kano’s lab at Osaka University, Japan.

Shadowpatching is a variation of two-photon targeted patching (TPTP), which uses the three-dimensional, pinpoint accuracy of the two-photon microscope to guide patch clamps to neurons expressing a fluorescent marker—often green fluorescent protein. TPTP has helped researchers to overcome a major hurdle in recording neurons in vivo, namely the uncertainty in knowing exactly what type of neuron has been clamped. However, because TPTP requires the expression of transgenes, it is restricted to specific animal lines or those transfected by viruses or other vectors. Shadowpatching needs no such manipulation.

The key to shadowpatching is turning the fluorescent marker idea inside out. Instead of labeling the neurons, first authors Kazuo Kitamura, Benjamin Judkewitz, and colleagues labeled the extracellular space. They flooded the neocortex or cerebellum of rodents with a fluorescent dye, Alexa 594, which is not taken up by cells, and found that neurons could be seen as dark cells, or shadows, against the bright background. This “shadowing” could even identify specific neuronal types, such as pyramidal cells and interneurons, by their gross morphology. Guided by the two-photon microscope, the researchers were then able to patch-clamp electrodes to individual shadowed cells. In fact, they used the same micropipette that delivers the fluorescent dye as the patch clamp. This turns out to be more than just convenient. Because there is a constant flow of dye from the pipette, the researchers were able to visualize when the pipette came in contact with a neuron because of a fluorescent dimple made in the cell membrane. Applying suction at just that moment creates a gigaseal or patch clamp in the cell membrane. Shadowpatching is a more controlled version of the “blow and seal” technique originally developed for patch-clamping cells in brain slices.

Kitamura and colleagues shadowpatched and recorded activity from a variety of cells in living mice. They made reliable electrical recordings from pyramidal cells in layers 2/3 of the barrel cortex, Purkinje neurons in the cerebellum, and interneurons in both regions. The success rate of around 70 percent is very high and is similar to that seen in isolated preps, said Häusser. Typical success rates for patch-clamping in vivo are about 20 percent. The researchers were also able to clamp dendrites, though at a reduced efficiency.

What is perhaps most useful about this technical development is the combination of patch-clamp recordings and electroporation. “This gives us a much more precise and targeted way to change properties of specific cell types in the brain,” said Häusser. Other techniques target entire groups or populations of neurons, with more system-wide consequences. “If you knock out a critical gene, for example, an NMDA receptor, then the whole system may compensate for the loss, but if you simply do the knockout in a single cell, then the network will not notice,” he said. Häusser predicts that the technique will give researchers new means to study mouse models of disease.—Tom Fagan


  1. This article provides some practical alternatives for visualizing and patching neurons in an in vivo preparation using two-photon imaging techniques. Although currently there are relatively few researchers conducting these types of experiments, perhaps these “tricks” can make these techniques more accessible to others, and increase the overall productivity rate in this low-throughput, but high-gain, approach to studying neuronal function in vivo. For somatic patching, shadowpatching will certainly improve productivity compared to blind patching, and for dendritic patching, it is likely a necessity.

    Usually, the excess fluorescent dye that is ejected into the extracellular space is considered an undesirable effect in imaging experiments, yet here, it is used to generate a shadow image of neurons—a dark soma against the fluorescent background noise. The contrast generated appears high, but the inherent optical sectioning ability of two-photon imaging requires a comprehensive z-stack of the neuron in order to determine morphology. Ironically, this optical sectioning feature is one of the significant advantages of two-photon imaging, but works against the shadowpatching objectives here. Yet, creating images from stacked z-planes need not be difficult with most imaging software applications and will provide a nice 3D image of the neuron recorded from. Additionally, one could also accelerate the process by focusing up and down to quickly ascertain the neuronal morphology, and verify cell type upon patching and filling the neuron.

    The authors also coupled their patching approach with single-cell electroporation of plasmid DNA in vivo. By placing an electrode filled with fluorescent dye and plasmid DNA encoding eGFP close to the soma and administering negative voltage pulses, individual neurons took up the dye and plasmids. Expression of the eGFP was determined at 24 and 48 hours afterwards. Identified single neurons in the cortex were electroporated; however, in more dense regions where neurons are closely packed, electroporation of individual cells may be more problematic.

    These helpful tips can certainly facilitate the patching and imaging of individual neurons in vivo, an already challenging and daunting task. I would imagine some practice and fine-tuning would be needed to optimize the amount of positive pressure required to eject the dye into the extracellular space. Likewise, there is likely a fine line between the voltage needed for a successful electroporation and a stimulus that is damaging or lethal to the cell. Once these and other empirical issues are worked out, not seeing your neurons may be the way to go for targeted in vivo imaging.

    There may be some useful variants of this approach that could be applied to neurodegenerative disease studies, such as AD. As any fluorescently labeled object can be imaged, in vivo two-photon imaging is not limited to cells. One thought is to fill the patch pipette with thioflavin S or a similar fluorescent plaque indicator and a conventional fluorescent dye, then shadowpatch neurons to measure effects of plaque deposits on neuronal morphology and physiology. Likewise, plaque markers can be injected systemically, and neurons in varying proximity to these labeled plaques can be shadowpatched using a fluorescent dye of a different emission wavelength. These plaque and filled neuron images can be overlaid, or, simultaneously acquired if multiple emission detections systems are in place. Analogous applications may involve apoptotic markers, Lewy body markers, or other pathogenic indicators. As long as it fluoresces, it can likely be detected with two-photon imaging.

    View all comments by Grace Stutzmann

Make a Comment

To make a comment you must login or register.


No Available References

Further Reading

No Available Further Reading

Primary Papers

  1. . Targeted patch-clamp recordings and single-cell electroporation of unlabeled neurons in vivo. Nat Methods. 2008 Jan;5(1):61-7. PubMed.