Article Figures & Data
Figures
- Figure 1.
Spatially restricted optogenetic silencing of visual cortex neurons. a, Top, Schematic of recording configuration and example of a fixed coronal section of visual cortex in which PV neurons are transfected with AAV9-DiO-hChR2-eYFP (green). The area of illumination over LM is indicated in blue. Scale bar, 0.5 mm. b, Example traces depicting the selection of optimal LED intensity to stimulate PV neurons. We noted that high LED intensity could induce abnormal firing, defined by a reduction in the action potential amplitude >20% during LED stimulation. Example traces show a cell-attached recording of a L2/3 PV neuron at low (top, 7 μW), optimal (center, 25 μW), and high (bottom, 130 μW) LED intensities, LEDon trials indicated by blue boxes. The LED intensity range used in this manuscript was 22–295 μW (mean value 96 μW). Scale bar, 1 mV, 2 s. c, Reliability of LED activation of PV neurons. Top, Example trace of a cell-attached recording of a L2/3 PV neuron during 8 consecutive LED illuminations at optimal intensity lasting 3 s each, with a 9 s interval between illuminations. Scale bar, 2 mV, 6 s. Bottom, Average response of three individual PV neurons to consecutive LED stimulations. Data are mean ± SEM of three to four trials for each neuron. LED intensities used were, from top to bottom: 95, 128, and 22 μW. d, Average response of an example LM PV neuron located 250 μm below the pia surface in response to LED illumination of increasing intensity. Data are mean ± SEM of eight to 16 trials. Optimal LED intensity was determined to be 128 μW, dashed line and top trace). Scale bar, 2 mV, 3 s. e, Example traces of two LM excitatory neurons recorded in L2/3 (top left, 230 μm below the pial surface) and L5a (bottom left, 450 μm below the pial surface, from the same animal as d during visual stimulation (black arrows show the direction of the visual stimuli). The firing rate of each trial is plotted to the right. Scale bar, 2 mV, 5 s. f, Trial averages for all intensities tested for the L5A neuron shown in e, mean ± SEM of 8 trials. g, Spatial activation of the three PV neurons shown in c, mean ± SEM. The inset shows an individual neuron activation map in x, y coordinates. h, Visually evoked responses from an example excitatory neuron were silenced when the LED was positioned directly over the cell soma (0 μm) and were not affected by LED when the light was positioned 400 μm away from the soma.
- Figure 2.
Retinotopic mapping and alignment of recording and silencing sites. a, Schematic of the visual presentation during intrinsic signal imaging; the screen was positioned 25 cm from the contralateral eye at an angle of 70° to the midline. b, Retinotopic maps in response to a vertically (top left) and horizontally (bottom left) moving bar. LM borders (black) were defined by rentinotopic contours. Red and blue circles indicate V1 recording and LM silencing regions, respectively. Bottom right, Silencing and recording sites from 10 of 12 animals were projected onto an averaged coordinate space defined by the lamboid suture and V1–LM phase transition of 10 animals. The center of silencing and recording sites were located 821 ± 50 μm and 657 ± 66 μm from the lamboid suture, respectively, and separated by 896 ± 39 μm.
- Figure 3.
Modulation of visually evoked V1 neuron responses by LM feedback is cell-type specific. a, Schematic of recording configuration. b, Example traces a V1 L2/3 excitatory neuron in response to visual stimulation (black box, arrows indicate presentation angle) in the presence (LEDon) and absence (LEDoff) of LED illumination. Scale bar, 2 mV 0.1 s. Right, Orientation tuning curves of the same neuron as in the left panel in the presence (blue) and absence of LED illumination. c, Impact of LED illumination expressed as percent facilitation of the evoked response at the preferred orientation, averaged across neurons. d–f, Average orientation tuning curves of V1 L2/3 excitatory (n = 16 neurons, 12 mice), L4 excitatory (n = 8 neurons, 6 mice), and L2/3 PV (n = 10 neurons, 5 mice) neurons aligned to their peak response. Each inset shows a representative spike waveform of the cell type recorded. Data are mean ± SEM.
- Figure 4.
Visually evoked responses of L2/3 excitatory neurons are reduced at the preferred orientation during LM silencing. Spike rates of individual neurons in during LM silencing (LEDon) versus control trials at the preferred (0°, left) and orthogonal orientation (90°, right) are shown in relation to the unity line (black). Scale is adjusted to maximize plotting area. Each inset shows a representative spike waveform of the cell type recorded. Scale bar, 2 mV, 1 s.
- Figure 5.
Location and axonal projection pattern of the subcortical region targeted for electrical stimulation. a, Schematic of viral injection location and the rostra-caudal position of coronal slices shown in b–g, sagittal view. b, c, Traced axonal fibers projected from three imaging planes. Yellow line indicates the slice edges and the dashed line indicates the border of L4. Scale bar, 0.5 mm. Inset, Example of a single image plane of the GFP signal used for tracing. Scale bar, 200 μm. d, Quantification of axonal fibers within LM and V1 (n = 8 animals). There was not a significant difference between LM and V1 (all layers, p = 0.613; specifically L2/3, p = 0.831, paired t test). Data are mean ± SEM. e, Hoechst stain reveals the increased thickness of L4 throughout V1, same slice as c. The white arrow indicates the border between V1 and LM. f, g, Characterization of injection site location across animals. In an example slice (f), the blue line indicates the distance from the bifurcation in the white matter and the center of mass of GFP fluorescence (Fcom; yellow plus sign). White scale bar, 0.5 mm (GFP epi-fluorescence image); black scale bar, 1 mm (transmitted light image of the same slice). The medial-lateral distance of Fcom from the white matter bifurcation point was averaged across animals (g, dashed line), as well as the ventral-dorsal distance of the Fcom from the dashed line to create an x, y point (right, black plus sign) that represents the average location of the eight injections. To characterize the average target location in the eight animals examined, the x, y point was overlaid onto a scaled atlas image using the white matter bifurcation as the reference coordinate point (left). B, Nucleus basalis; BLA, basal lateral amygdala; CPu, caudate putamen; GP, globus pallidus; LV, lateral ventricle; IC, internal capsule.
- Figure 6.
BF stimulation desynchronizes cortical EEG and increases responsiveness of V1 neurons. a, Schematic of electrode placement, sagittal view. b, Representative EEG recording during BF stimulation. Top, Single stimulation trial. Bottom, Spectrogram averaged across 12 trials. Stimulation time is in orange and stimulation artifact on the spectrogram is blank. c, Time course of prefrontal EEG desynchronization induced by BF stimulation. Power ratio, EEG power at 10–100 Hz divided by EEG power at 1–10 Hz. Black line is the mean and the gray line is the SEM (n = 7 mice). d, e, V1 L2/3 excitatory neuron visually evoked responses and CV at the preferred orientation (n = 12 neurons, 7 mice) in the absence (baseline, filled circles) and presence (BF stim, open circles) of BF stimulation. Circles joined by lines indicate paired measures of individual neurons; plus signs indicate mean values. f, Average orientation tuning curves of V1 L2/3 excitatory neurons aligned to their peak response during baseline (filled circles) and BF stimulation (open circles). g, EEG power index, 1-EEG power Post-Stim1-10Hz/EEG power Pre-Stim1-10Hz, of the individual animals shown in c. *p < 0.05; **p < 0.001.
- Figure 7.
LM response properties are unaltered during BF stimulation. a, b, LM L2/3 excitatory neuron visually evoked responses and CV, recorded at the preferred orientation in the absence (baseline, filled circles) and presence of BF stimulation (BF stim, open circles). Circles joined by lines indicate paired measures of individual neurons (n = 10 neurons, 5 mice); plus signs indicate mean values. There was not a significant difference between conditions, Response, p = 0.598 and CV, p = 0.986. c, EEG power index from the experiments shown in a and b. Black line indicates the mean value. d, Average orientation tuning curves of LM L2/3 excitatory neurons aligned to their peak response during baseline (filled circles) and BF stimulation (open circles). Data are mean ± SEM.
- Figure 8.
LM and V1 neurons are differentially modulated by BF stimulation. Shown are average orientation tuning curves aligned to their peak response during baseline (filled circles) and BF stimulation (open circles) for PV neurons in V1 (n = 8 neurons, 5 mice) and LM (n = 6 neurons, 3 mice). Data are mean ± SEM.
- Figure 9.
LM-mediated facilitation is reduced by half during BF stimulation. a, b, V1 L2/3 excitatory neuron visually evoked responses and CV recorded at the preferred orientation in the absence (baseline, filled circles) and presence of BF stimulation in conjunction with LM silencing. Circles joined by lines indicate paired measures of individual neurons (n = 12 neurons, 7 mice); plus signs indicate mean values. c, Impact of LM silencing expressed as percentage facilitation of the evoked response at the preferred orientationaveraged across excitatory neurons in the absence (control, n = 16 neurons, 12 mice; left) and presence (BF stim, n = 12 neurons, 7 mice) of BF stimulation. The predicted reduction (10.3%) for a linear relationship is plotted as a dashed line. Control values are replotted from Figure 3c averaged across PV neurons in the absence (right; control, n = 10 neurons, 5 mice) and presence (BF stim, n = 8 neurons, 5 mice) of BF stimulation. Data are mean ± SEM. d, Silencing and recording sites for the 7 animals in Group 2 were projected onto an averaged coordinate space as in Figure 2. The center of silencing and recording sites were located 967 ± 44 μm and 700 ± 50 μm from the lamboid suture, respectively, and separated by 985 ± 36 μm. Scale bar, 0.5 mm. *p < 0.05; **p < 0.001.
