In Vivo Voltage-Sensitive Dye Study of Lateral Spreading of Cortical Activity in Mouse Primary Visual Cortex Induced by a Current Impulse

Abstract

In the mammalian primary visual cortex (V1), lateral spreading of excitatory potentials is believed to be involved in spatial integrative functions, but the underlying cortical mechanism is not well understood. Visually-evoked population-level responses have been shown to propagate beyond the V1 initial activation site in mouse, similar to higher mammals. Visually-evoked responses are, however, affected by neuronal circuits prior to V1 (retina, LGN), making the separate analysis of V1 difficult. Intracortical stimulation eliminates these initial processing steps. We used in vivo RH1691 voltage-sensitive dye (VSD) imaging and intracortical microstimulation in adult C57BL/6 mice to elucidate the spatiotemporal properties of population-level signal spreading in V1 cortical circuits. The evoked response was qualitatively similar to that measured in single-cell electrophysiological experiments in rodents: a fast transient fluorescence peak followed by a fast and a slow decrease or hyperpolarization, similar to EPSP and fast and slow IPSPs in single cells. The early cortical response expanded at speeds commensurate with long horizontal projections (at 5% of the peak maximum, 0.08-0.15 m/s) however, the bulk of the VSD signal propagated slowly (at half-peak maximum, 0.05-0.08 m/s) suggesting an important role of regenerative multisynaptic transmission through short horizontal connections in V1 spatial integrative functions. We also found a tendency for a widespread and fast cortical response suppression in V1, which was eliminated by GABAA-antagonists gabazine and bicuculline methiodide. Our results help understand the neuronal circuitry involved in lateral spreading in V1.

Fig 1. Verification of stimulation site locations.

(A) Time-lapse images evoked by a single-pulse 50 μA stimulus in V1 using a tungsten microelectrode (average of 12 trials). ΔF/F values above the level of statistical significance (SD×2.7) are represented in false color as indicated on the bar, superimposed over a greyscale image of the cortical surface. Delay after stimulation indicated on each frame. Stimulation in V1 evoked cortical activity around the stimulation site (cross), and at secondary independent locations (hollow arrows). In the same mouse, 3 stimulation sites in V1 and 2 sites around the first appearing secondary response (circle) were marked by the application of 50 μA for 4 s. (B) Tangential slice of mouse cortex stained for cytochrome oxidase C showing the 5 marked sites as round ruptures in the brain tissue (arrows). The approximate edge of the cranial window, as seen on the VSD image (dashed border) and the approximate border of the heavily stained area (V1) as seen on the slice (dotted curve) are shown in both A and B. The slice and the VSD images were roughly aligned by matching marked sites to the centers of cortical response sites using rotation, scaling and translation. Au: auditory cortex; S1: Somatosensory cortex; V2M: medial V2; A: anterior; L: lateral; P: posterior; M: medial.

Fig 2. Time courses of evoked VSD response at the stimulation site.

Evoked VSD response at 10, 25 and 50 μA stimulation intensities (each an average of 12 trials). Single-pulse electrical stimuli were delivered at 250 μm below the dura with glass micropipette electrodes with 5–6 μm tip diameter and ~1 MΩ resistance. The 4 examples displayed are from different animals. Typically, a fast initial increase of fluorescence was followed by a fast decrease, then a slow decrease or slow undershoot below the baseline. Stimulation at 0 ms; figure legends: stimulation intensity in μA; arrows: break in the downward slope of the signal.

Fig 3. Spatial propagation following high and low intensity stimulation.

High-intensity stimulation evoked regenerative propagation in V1. (A, B) Time-lapse images of the VSD signal (averages of 12 trials each) in the right visual cortex evoked by a 25 μA and 50 μA stimulus, respectively, delivered at 250 μm below the brain surface (layer II/III) in V1. Delay after stimulation indicated on each frame. ΔF/F values reproduced as shown on the right bars. A: anterior; P: posterior; L: lateral; M: medial; asterisk: stimulation site. In panel B, 25 ms frame, dashed line: approximate V1/V2 border; arrows: direction of cross-sections in C–F. (C–F) cross-sections of the fluorescence signal along the anteroposterior (C, D) and V1-V2L axes (E, F) (directions indicated by arrows on panel B, 25 ms frame) at increasing delays after stimulation, at 50 μA and 25 μA stimulus, respectively. Stimulus strength and direction indicated on each panel. Delay after stimulation (in ms) shown above each plot. Asterisk: stimulation site; dashed line: approximate V1/V2 border; V2A: anterior V2; V2L: lateral V2.

Fig 4. Stronger stimulation is more likely to evoke widespread activation in V1.

Activation was defined as fluorescence above the level of statistical significance (defined as SD×2.7) longer than 20 ms at every pixel in V1. Percentage of activated pixels in V1 is shown versus stimulation intensity in 8 animals. Data from each animal is represented with a different color. V1 area maps were drawn manually for each animal based on activation latency maps and activation patterns in the recordings.

Fig 5. VSD response profile around stimulation site.

(A) Time-lapse images of the VSD signal around the stimulation site (cross) in false colors as indicated on the right bar, superimposed over a greyscale image of the cortical surface. Only fluorescence levels above the level of statistical significance (SD×2.7) are displayed. Time after stimulation is indicated on each frame. The cortical response was elongated, the long and short axes of the elongated shape are shown on the 8 ms frame. (B–E) Cross-sections of VSD signal along the long (B, normalized in C) and short axis (panel D, normalized in E) of the cortical response (as shown in panel A, 8 ms), at increasing delays after stimulation. Orientation and direction of the distance axis indicated by inset on each plot. Delay (in ms) is indicated next to each line, and stimulation site is always at 0 distance. V2L: lateral V2. (A–E) are from the same animal, average of 12 trials. Different parts saturated at different ΔF/F levels, indicating that the cause was not VSD response saturation (compare the fluorescence profile on both sides of the stimulation site at 8–12 ms on panel B). (F) Average of cross-sections along the long axis of propagation in 13 mice following 50 μA stimulation. Delays after stimulation are indicated in the legend. Error bars at indicate SD, stimulation site was at 0 distance.

Fig 6. Lateral propagation velocities and anisotropy.

Lateral propagation velocities in V1 following 50 μA stimulation at different thresholds, along the long and short axes of the elongated cortical response. Thresholds are relative to peak maximum at each pixel. (A) Means ± SD (n = 17 mice for all panels) of propagation velocities at indicated thresholds. Data points shifted to avoid overlap. (B–C) histograms of spreading velocities at 50% of peak maximum (B) and at 5% (C). (D) Ratio of velocities along the two axes at indicated thresholds. Ratios were calculated for each mouse separately, then averaged. For (B–D), bin centers are indicated, bin widths were equal to the distance between the centers.

Fig 7. Closely timed falling phases indicate fast-spreading inhibition in V1.

(A) Time-lapse images of cortical VSD signal evoked by a single-pulse stimulus in V1. Time after stimulation indicated on each frame. ΔF/F values reproduced as shown in the right bar. Asterisk: stimulation site; V2L: lateral V2; V2M: medial V2; A: anterior; L: lateral; P: posterior; M: medial. (B–C) Latency maps at 0.3% ΔF/F threshold for the rising and falling phases of the initial fluorescence peak. Outlines of the initial responses in V1, V2L and V2M (thick contours) and the estimated functional border between V1 and V2 (dashed curves) were drawn on panels A–C. Latencies represented as shown on the bars. Contour lines were drawn for each millisecond. The very low falling phase latencies found in posterior V1 are calculation artifacts at this threshold due to low signal amplitudes. (D) Time courses of the fluorescence signal at locations 1–7 on panels B and C. Points 1–5 are located in V1; 6 and 7 are in anterior V2 (V2A, blue traces). Arrow: falling phases grouped together. (E) Cross-section of cortical activity along a line segment overlapping locations 1–7 on B and C at increasing delays after stimulation. Asterisk at 0 distance: stimulation site in V1; dotted line: estimated V1/V2 functional border; delay (in ms) indicated next to each profile line. All data in panels A–E are from the same animal, average of 12 trials. (F) Distribution of the ratio of rising and falling phase latency mean absolute errors (MAE) around the median for all 17 mice, between 50–75% of the maximum peak amplitude in V1 for each animal (n = 17, 50 μA stimulation). A >1 ratio indicates that falling phase latencies are more grouped together than the rising phase ones. For each bin, the center is indicated. Bin widths are equal to the distance between centers.

Fig 8. GABA_A-mediated inhibition strongly affects the deactivation phase in V1.

(A) VSD images of evoked cortical activity depicted in color as indicated on the right bar. Control (top row), after focal application of gabazine (GBZ, middle row), and their difference (GBZ-control, bottom row). Frames taken at indicated delays after 50 μA single-pulse stimulation in V1. Contour lines on the difference frames at every 0.1% ΔF/F level above 0. Asterisk: stimulation site; circle: GBZ ejection site; dashed curve: approximate V1/V2 border; A: anterior; L: lateral; P: posterior; M: medial. (B) Cross-section of fluorescence levels in V1 and anterior V2 (V2A) at increasing delays after V1 stimulation. Sampling location indicated by the dotted line segment on the 115 ms frames in panel A. Delays (in ms) are shown next to each line; the dashed vertical line indicates the approximate V1/V2A border, asterisk at 0: position closest to the stimulation site. (C) Time courses of cortical activity in V1 before (control) and after application of gabazine (GBZ), and after ~2 hours elimination. GBZ-control: control subtracted from GBZ. Sampling site indicated by a hollow rectangle between the stimulation and ejection sites on the 115 ms frames in panel A. Left: original data; right: normalized to peak maxima. All data are from the same animal, averages of 16 trials each.