Controlling a neuron with light in a brain slice

Optogenetics makes chosen neurons sensitive to light by giving them the gene for a light-gated microbial protein, an opsin. Channelrhodopsin-2 (ChR2) opens a cation channel in blue light and depolarises the cell, so it excites; the engineered chloride pump eNpHR3.0 moves chloride ions into the cell in yellow light and hyperpolarises it, so it silences. Boyden and colleagues first showed that channelrhodopsin-2 can drive spikes in mammalian neurons with millisecond precision (Boyden et al., 2005).

The experiment is ex vivo: it records from neurons in an acute slice of mouse brain kept alive in oxygenated artificial cerebrospinal fluid (ACSF). The mice come from a published cross of a cell-type-specific Cre line with a Cre-dependent reporter line, so that only one type of interneuron expresses the opsin together with a fluorescent tag (Madisen et al., 2012). No viral injection or surgery is part of this slice experiment; in vivo optogenetics additionally requires approved surgery under anaesthesia to express the opsin and implant a light fibre, which is not described here.

The neuron is a teaching model. Its membrane follows the Hodgkin–Huxley equations with the kinetics accelerated to a reference temperature of about 18.5 °C, so that its action potentials are about half a millisecond wide at half amplitude, like those of a fast-spiking interneuron. Each opsin is a kinetic model with an opening time constant, a closing time constant and a light sensitivity, with approximate values based on published comparisons of opsins (Gradinaru et al., 2010; Mattis et al., 2011). The cell is a small sphere of about 1000 µm² (about 10 pF), so a current density of 1 µA/cm² is a whole-cell current of about 10 picoamperes.

The rig, the fluorescence check, the calibration of the light, the seal and break-in, and the light-only heating check are simplified illustrations, not part of the model, and are labelled as such. The photocurrents, the light-driven spikes and the silencing come from the model. The model membrane fires faster and more regularly than a real interneuron, because the Hodgkin–Huxley equations have no spike-frequency adaptation, so the numbers illustrate the logic rather than the exact physiology of any cell.

Learning objectives

• Match the opsin to the question: an excitatory channel (ChR2, blue) to drive firing, an inhibitory pump (eNpHR3.0, yellow) to silence it
• Confirm opsin expression from the fluorescent reporter, and use a non-expressing neighbour and an opsin-negative littermate as controls
• Calibrate the light as an irradiance at the sample (mW/mm²) rather than an instrument setting, and say why that is the comparable number
• Record channelrhodopsin-2 photocurrents in voltage clamp, read the peak, the steady state and the desensitisation, and see them saturate with irradiance
• Drive spikes with light trains and find where spike fidelity falls at high frequency, and where too much light causes extra spikes or depolarisation block
• Silence current-driven firing with eNpHR3.0, recognise post-inhibitory rebound, and analyse the result across animals with the correct test and the mouse as the unit

Ethics and safety

Slices for ex vivo optogenetics come from animals humanely killed by trained and authorised staff under approval by an Institutional Animal Ethics Committee; in India such work is governed by the CCSEA under the Prevention of Cruelty to Animals Act 1960, with equivalent frameworks in the European Union (Directive 2010/63/EU), the United States (Animal Welfare Act and PHS Policy) and the United Kingdom (Animals (Scientific Procedures) Act 1986). The 3Rs apply: many cells and slices are recorded from each animal (Reduction), and the opsin is expressed through a published genetic cross rather than surgery for this preparation (Refinement). Littermates of each genotype are the unit of analysis, not cells or slices, and the number of animals should come from a power calculation made in advance. In vivo optogenetics additionally requires approved surgery under anaesthesia to deliver the opsin and the light; no surgical steps, coordinates, titres or doses are given here. Light sources are bright and can damage the eyes, so appropriate eye protection and interlocks are used.

Key terms

• Optogenetics: Controlling neurons with light by expressing a light-gated protein in them. See optogenetics.
• Opsin: A light-sensitive protein. Microbial opsins such as channelrhodopsin and halorhodopsin are used to excite or silence neurons. See opsin.
• Channelrhodopsin-2 (ChR2): A light-gated cation channel from algae. Blue light opens it; cations flow in and depolarise the cell, so it excites.
• eNpHR3.0: An engineered chloride pump (a halorhodopsin). Yellow light drives chloride ions into the cell and hyperpolarises it, so it silences.
• Cre–loxP expression: A genetic system in which the enzyme Cre, made only in one cell type, switches on a gene flanked by loxP sites, so the opsin is expressed only where Cre is. See Cre recombinase and floxed allele.
• Reporter fluorescence: A fluorescent protein (here EYFP) joined to the opsin, so that cells expressing the opsin glow under the right light and expression can be confirmed before recording.
• Acute slice (ex vivo): A thin slice of living brain kept in oxygenated ACSF for a few hours, in which single cells can be recorded. See ex vivo.
• Irradiance: The light power per unit area reaching the cell, in milliwatts per square millimetre (mW/mm²); it, not the instrument setting, determines how strongly the opsin is driven.
• Photocurrent: The electrical current that flows through an opsin when it is lit, measured in voltage clamp.
• Desensitisation: The fall of the channelrhodopsin current from an initial peak to a smaller steady level while the light is still on, as some molecules enter an inactive state.
• Spike fidelity: The fraction of light pulses in a train that each evoke exactly one action potential; it falls at high frequency because the channel cannot close and reopen fast enough.
• Depolarisation block: A state in which a cell is held so strongly depolarised that its sodium channels stay inactivated and it stops firing, which too much light can cause.
• Post-inhibitory rebound: A burst of firing just after a hyperpolarisation ends, as inactivated sodium channels recover; a known caveat of optogenetic silencing.
• Access resistance: The resistance between the pipette wire and the inside of the cell; a high value slows the voltage clamp and underestimates fast photocurrents. See patch clamp.