Scanless two-photon excitation of channelrhodopsin-2
Articles
Scanless two-photon excitation of channelrhodopsin-2
© 2010 Nature America, Inc. All rights reserved.
Eirini Papagiakoumou1,5, Francesca Anselmi1,5, Aurélien Bègue1,5, Vincent de Sars1, Jesper Glückstad2, Ehud Y Isacoff 3,4 & Valentina Emiliani1 Light-gated ion channels and pumps have made it possible to probe intact neural circuits by manipulating the activity of groups of genetically similar neurons. What is needed now is a method for precisely aiming the stimulating light at single neuronal processes, neurons or groups of neurons. We developed a method that combines generalized phase contrast with temporal focusing (TF-GPC) to shape two-photon excitation for this purpose. The illumination patterns are generated automatically from fluorescence images of neurons and shaped to cover the cell body or dendrites, or distributed groups of cells. The TF-GPC two-photon excitation patterns generated large photocurrents in Channelrhodopsin-2– expressing cultured cells and neurons and in mouse acute cortical slices. The amplitudes of the photocurrents can be precisely modulated by controlling the size and shape of the excitation volume and, thereby, be used to trigger single action potentials or trains of action potentials.
Since early studies1, an important component of what we know about the brain has come from electrical stimulation or local drug application, but these approaches have poor spatial resolution, require physical contact or slow exchange and cannot target specific cell types. Advanced microscopy and optogenetics have recently greatly advanced the precision of these manipulations. Optical stimulation can be less invasive, allows superior spatial and temporal resolution as well as specificity for cell type and quick reversibility. The most widely used optogenetic tool is Channelrhodopsin-2 (ChR2)2, a cation-selective channel that, like other genetically encoded pumps and channels, can be functionally expressed in mammalian neurons under the control of cell-specific promoters. Upon illumination with blue light, the cation flux generated through ChR2 produces rapid membrane depolarization, which can evoke reliable trains of action potentials at frequencies up to 200 Hz (ref. 3). Several methods have been used to activate ChR2, including widefield lamp illumination4, laser-scanning illumination5 or illumination with a micro-LED array6. For in vivo applications, ChR2 is usually stimulated by light sources coupled to optical
fibers7–9 or by miniaturized LEDs10. The limitations imposed by the low penetration depth of blue light and the lack of optical sectioning inherent for single-photon excitation are partly mitigated by the low excitation levels necessary for photoactivation (~1 mW mm−2)7 and by genetic targeting of ChR2 expression, permitting applications in freely moving mice, including the optical control of whisker movement7, locomotion8, the probing of Parkinsonian neuronal circuits9 and the restoration of visual function in retinal degeneration11. The precision of these manipulations would be considerably enhanced if two-photon excitation could be used to confine the stimulation in three dimensions to select subsets of neurons that cannot be distinguished based on cell-specific promoters or distinct subcellular compartments in a neuron. Several factors make it challenging to use two-photon excitation to stimulate ChR2. First, ChR2 has a low conductance (~80 femtosiemens)12, making it difficult to drive action potentials by photoactivation with the standard small two-photon excitation volume (~2–5 μm3). Increasing excitation density would not help, as saturation of excited ChR2 channels is quickly reached owing to the high two-photon absorption cross-section of ChR2 (~260 Goeppert-Mayer units at 920 nm) and the long lifetime of the conducting excited states (~10 ms) as demonstrated in the first paper, to our knowledge, reporting action-potential generation by two-photon ChR2 photoactivation13. Increasing the fraction of the cell membrane stimulated by two-photon excitation by underfilling the objective back aperture13 or fast scanning of the laser beam through multiple positions13,14 suffers from substantial loss in axial (z axis) or temporal resolution. The above considerations suggest that the optimal illumination for ChR2 stimulation in two-photon excitation experiments is with low excitation density, large excitation area, and millisecond and microscale resolution. We recently proposed a solution that could satisfy these requirements: a method for generating two-photon light patterns by combining digital holography15–17 with a dispersive optical setup for temporal focusing18 (TF-DH), in which large two-dimensional areas are excited rapidly with a depth resolution of
Scanless two-photon excitation of channelrhodopsin-2
© 2010 Nature America, Inc. All rights reserved.
Eirini Papagiakoumou1,5, Francesca Anselmi1,5, Aurélien Bègue1,5, Vincent de Sars1, Jesper Glückstad2, Ehud Y Isacoff 3,4 & Valentina Emiliani1 Light-gated ion channels and pumps have made it possible to probe intact neural circuits by manipulating the activity of groups of genetically similar neurons. What is needed now is a method for precisely aiming the stimulating light at single neuronal processes, neurons or groups of neurons. We developed a method that combines generalized phase contrast with temporal focusing (TF-GPC) to shape two-photon excitation for this purpose. The illumination patterns are generated automatically from fluorescence images of neurons and shaped to cover the cell body or dendrites, or distributed groups of cells. The TF-GPC two-photon excitation patterns generated large photocurrents in Channelrhodopsin-2– expressing cultured cells and neurons and in mouse acute cortical slices. The amplitudes of the photocurrents can be precisely modulated by controlling the size and shape of the excitation volume and, thereby, be used to trigger single action potentials or trains of action potentials.
Since early studies1, an important component of what we know about the brain has come from electrical stimulation or local drug application, but these approaches have poor spatial resolution, require physical contact or slow exchange and cannot target specific cell types. Advanced microscopy and optogenetics have recently greatly advanced the precision of these manipulations. Optical stimulation can be less invasive, allows superior spatial and temporal resolution as well as specificity for cell type and quick reversibility. The most widely used optogenetic tool is Channelrhodopsin-2 (ChR2)2, a cation-selective channel that, like other genetically encoded pumps and channels, can be functionally expressed in mammalian neurons under the control of cell-specific promoters. Upon illumination with blue light, the cation flux generated through ChR2 produces rapid membrane depolarization, which can evoke reliable trains of action potentials at frequencies up to 200 Hz (ref. 3). Several methods have been used to activate ChR2, including widefield lamp illumination4, laser-scanning illumination5 or illumination with a micro-LED array6. For in vivo applications, ChR2 is usually stimulated by light sources coupled to optical
fibers7–9 or by miniaturized LEDs10. The limitations imposed by the low penetration depth of blue light and the lack of optical sectioning inherent for single-photon excitation are partly mitigated by the low excitation levels necessary for photoactivation (~1 mW mm−2)7 and by genetic targeting of ChR2 expression, permitting applications in freely moving mice, including the optical control of whisker movement7, locomotion8, the probing of Parkinsonian neuronal circuits9 and the restoration of visual function in retinal degeneration11. The precision of these manipulations would be considerably enhanced if two-photon excitation could be used to confine the stimulation in three dimensions to select subsets of neurons that cannot be distinguished based on cell-specific promoters or distinct subcellular compartments in a neuron. Several factors make it challenging to use two-photon excitation to stimulate ChR2. First, ChR2 has a low conductance (~80 femtosiemens)12, making it difficult to drive action potentials by photoactivation with the standard small two-photon excitation volume (~2–5 μm3). Increasing excitation density would not help, as saturation of excited ChR2 channels is quickly reached owing to the high two-photon absorption cross-section of ChR2 (~260 Goeppert-Mayer units at 920 nm) and the long lifetime of the conducting excited states (~10 ms) as demonstrated in the first paper, to our knowledge, reporting action-potential generation by two-photon ChR2 photoactivation13. Increasing the fraction of the cell membrane stimulated by two-photon excitation by underfilling the objective back aperture13 or fast scanning of the laser beam through multiple positions13,14 suffers from substantial loss in axial (z axis) or temporal resolution. The above considerations suggest that the optimal illumination for ChR2 stimulation in two-photon excitation experiments is with low excitation density, large excitation area, and millisecond and microscale resolution. We recently proposed a solution that could satisfy these requirements: a method for generating two-photon light patterns by combining digital holography15–17 with a dispersive optical setup for temporal focusing18 (TF-DH), in which large two-dimensional areas are excited rapidly with a depth resolution of
