An optogenetics- and imaging-assisted simultaneous multiple patch-clamp recording system for decoding complex neural circuits

Abstract

Deciphering neuronal circuitry is central to understanding brain function and dysfunction, yet it remains a daunting task. To facilitate the dissection of neuronal circuits, a process requiring functional analysis of synaptic connections and morphological identification of interconnected neurons, we present here a method for stable simultaneous octuple patch-clamp recordings. This method allows physiological analysis of synaptic interconnections among 4-8 simultaneously recorded neurons and/or 10-30 sequentially recorded neurons, and it allows anatomical identification of >85% of recorded interneurons and >99% of recorded principal neurons. We describe how to apply the method to rodent tissue slices; however, it can be used on other model organisms. We also describe the latest refinements and optimizations of mechanics, electronics, optics and software programs that are central to the realization of a combined single- and two-photon microscopy-based, optogenetics- and imaging-assisted, stable, simultaneous quadruple-viguple patch-clamp recording system. Setting up the system, from the beginning of instrument assembly and software installation to full operation, can be completed in 3-4 d.

Figure 1

Manipulators for simultaneous multiple patch-clamp recordings. (a) Photograph of an array of eight MINI L&N motorized manipulators at a standard microscope. (b) Photographs of the side (top) and top (bottom) views of a 151-mm W × 151-mm D × 197-mm H MINI L&N motorized manipulator first invented in 1992 (left), a 104-mm W × 114-mm D × 130-mm H JUNIOR L&N motorized manipulator redesigned in 2010 (middle) and a 49-mm W × 114-mm D × 157-mm H JUNIOR Compact L&N motorized manipulator developed in 2013 (right) on a 25-mm-grid breadboard. Note ~50 mm width of the JUNIOR Compact manipulator in the y axis. Note that the red adapters can be custom-removed to further reduce the width of the manipulators.

Figure 2

Hardware wiring for simultaneous multiple patch-clamp recordings. Schematic sketch shows the hardware wiring of the computer, interfaces and manipulators for the simultaneous multiple (≥octuple) patch-clamp recordings setup. Inset image shows the arrangement of the eight patch pipettes and recording chamber. ADC, analog-to-digital converter; DAC, digital-to-analog converter; Ext. Comm., external command; I, current; PCI, peripheral component interconnect; TRIG, trigger; TTL, transistor-transistor logic pulse; V, voltage.

Figure 3

IGOR-based program for simultaneous multiple patch-clamp recordings. (a) Screenshot of IGOR-based program testing synaptic connections formed among eight recorded neurons. (b) Screenshot of the IGOR-based program measuring calcium transients at synapses of two neurons. Note that IGOR-based program displayed on multiple monitors can simultaneously run electrophysiology, two-photon laser-scanning imaging and/or optogenetics routines. (c) Flowchart showing generic sequence of operations during experiments. The laser light for optogenetic stimulation is delivered through the objective of the microscope (when no image data acquisition is required) and/or an optic fiber.

Figure 4

Laserspritzer-based synaptic connection ‘search’ technique. (a) Schematic graph shows the fabrication of laserspritzer fiber probe. (b) The tip of a laserspritzer (top) and light spot produced by the laserspritzer with laser illumination (bottom) under a microscope. (c) 3D transparent rendering of a sensorimotor cortical brain slice prepared from a 2-month-old VGAT-YFP-ChR2 (green)-positive mouse. Bottom left, L2 neuron (red) filled with Alexa Fluor 594 (0.2 mg/ml; Invitrogen, cat. no. A10438). ChR2-expressing cells (bottom middle) and positions of the soma of ChR2-expressing cell 3 in the red circle and stimulating laserspritzer (bottom right) under differential interference contrast and/or fluorescence microscopy. Arrowheads indicate the somata of interneurons expressing YFP-ChR2, and numbers 1–8 indicate cells stimulated by laserspritzer. Scale bar, 20 µm (top, bottom left, bottom middle) and 10 µm (bottom right). (d) Laserspritzer stimulation (8 ms; 0.1 mW/mm²) at two of eight ChR2-expressing cells elicited inhibitory postsynaptic potentials (IPSCs) with short (<6 ms) and fixed latencies in postsynaptic L2 neuron. Note ten consecutive recording traces shown in black, average subthreshold and suprathreshold responses shown in red and laser stimulation indicated by blue bars. These animal experiments were approved by the University of Wyoming Institutional Animal Care and Use Committee.

Figure 5

Single- and two-photon laser–based synaptic connection search technique. (a) Action potential (AP) thresholds for cortical L5 pyramidal neurons expressing either ChR2 (≤20 µW, n = 1; 20–200 µW, n = 7; 200–2,000 µW, n = 11; >2 mW, n = 4 out of 23 neurons tested) or CheRiff (<20 µW, n = 33; 20–200 µW, n = 4 out of 37 neurons tested) in response to a 5-ms 473-nm single-photon laser pulse. (b) Top, schematic graph shows the photostimulation started with a 20-ms 920-nm two-photon spiral laser scanning followed by a 5-ms 473-nm single-photon laser pulse. Bottom, responses in a CheRiff-GFP-expressing cortical L5 pyramidal neuron to a 5-ms 473-nm suprathreshold single-photon laser pulse stimulation (stim) alone (cyan trace), a 20-ms 920-nm subthreshold two-photon laser spiral scanning stimulation alone (red trace), a 5-ms 473-nm subthreshold single-photon laser pulse stimulation alone (blue trace) and the combination of the subthreshold two-photon laser spiral scanning and single-photon laser pulse stimulation (pink trace) under a 40×/0.8 numerical aperture (NA) Olympus objective lens. (c) Top, responses of a CheRiff-GFP-expressing cortical L5 pyramidal neuron to the suprathreshold single-photon laser pulse stimulation (cyan traces), and the combination of the subthreshold two-photon laser spiral scanning and single-photon laser pulse stimulation (pink traces) with the objective lens focusing point moving away from the soma. Bottom, plot of the average suprathreshold action potential responses of CheRiff-GFP-expressing L5 pyramidal neurons against the distances between the laser-focusing spot and soma of L5 pyramidal neurons. Note the error bars showing the standard errors of action potential incidences in L5 pyramidal neurons (n = 21), and note that the half-height spatial resolution of the combined single- and two-photon stimulation (~30 µm) is smaller than that of the single-photon stimulation alone (~60 µm). (d) Reconstruction of four L5 pyramidal neurons recorded simultaneously from an acute cortical slice superimposed on the transmitted light image captured during the recordings. The double-colored dots indicate the putative synaptic contacts identified by light microscopy. (e) The schematic drawing shows symbolically the synaptic connections. (f) Top, the combined single- and two-photon optogenetic stimulation of CheRiff-GFP–expressing cortical L5 pyramidal neurons (green) evoked unitary excitatory postsynaptic currents (uEPSCs) in one of the L5 pyramidal neurons (orange), but not in two others (blue and red). The monosynaptic connection was confirmed after the CheRiff-GFP–expressing neuron was patched and electrically stimulated in the whole-cell configuration. Note the slightly smaller amplitude and longer kinetics of the average uEPSC evoked by the optogenetic stimulation compared with that evoked by the current injection in postsynaptic neuron owing to the slight jittering of optogenetically evoked action potentials in the presynaptic neuron. The majority of the unconnected axonal branches of the pyramidal neurons are not reconstructed for simplicity, and these animal experiments were approved by the University of Virginia Institutional Animal Care and Use Committee.

Figure 6

Example of simultaneous multiple patch-clamp recordings. (a) An experiment of the rat sensorimotor cortical brain slice performed by a trainee with 3 weeks of experience on the simultaneous octuple patch-clamp recordings setup shown in Figure 1a. Single action potentials elicited in the presynaptic pyramidal neuron (red) evoked unitary excitatory postsynaptic potentials (uEPSPs) in the postsynaptic pyramidal neuron (orange), and single action potentials elicited in presynaptic interneuron (cyan) evoked unitary inhibitory postsynaptic potentials (uIPSPs) in postsynaptic pyramidal neurons (blue and pink) and interneuron (green). Scale bars apply to all recording traces with 80 and 2 mV bars applied to traces with and without action potentials, respectively. (b) Reconstruction of the eight recorded neurons reveals six pyramidal neurons and two interneurons, including a basket cell and an aspiny neuron that could not be morphologically identified owing to a large truncation of its axonal arborization. The double-colored dots indicate the putative synaptic contacts based on anatomical reconstruction. (c) The schematic drawing shows symbolically the synaptic connections. The majority of the unconnected axonal branches of the pyramidal neurons are not reconstructed for simplicity, and these animal experiments were approved by the University of Virginia Institutional Animal Care and Use Committee.