Hemodynamic responses evoked by neuronal stimulation via channelrhodopsin-2 can be independent of intracortical glutamatergic synaptic transmission

Abstract

Maintenance of neuronal function depends on the delivery of oxygen and glucose through changes in blood flow that are linked to the level of ongoing neuronal and glial activity, yet the underlying mechanisms remain unclear. Using transgenic mice expressing the light-activated cation channel channelrhodopsin-2 in deep layer pyramidal neurons, we report that changes in intrinsic optical signals and blood flow can be evoked by activation of a subset of channelrhodopsin-2-expressing neurons in the sensorimotor cortex. We have combined imaging and pharmacology to examine the importance of glutamatergic synaptic transmission in this form of neurovascular coupling. Blockade of ionotropic glutamate receptors with the antagonists CNQX and MK801 significantly reduced forepaw-evoked hemodynamic responses, yet resulted in no significant reduction of channelrhodopsin-evoked hemodynamic responses, suggesting that stimulus-dependent coupling of neuronal activity to blood flow can be independent of local excitatory synaptic transmission. Together, these results indicate that channelrhodopsin-2 activation of sensorimotor excitatory neurons produces changes in intrinsic optical signals and blood flow that can occur under conditions where synaptic activation of neurons or other cells through ionotropic glutamate receptors would be blocked.

Figure 1. IOS and laser speckle imaging reveals ChR2 and sensory (forepaw) evoked hemodynamic responses.

A, Schematic of experimental set-up. Responses to ChR2 stimulation (473 laser activation) and electrical sensory stimulation were recorded via IOS imaging (635 nm) and laser speckle imaging (785 nm). B, Averaged EEG responses (20 trials) to 100 ms and 1000 ms of ChR2 stimulation in a representative animal. C, left: Green light image of the cortical surface. Peak IOS response (averaged over 1 s from peak) to 1 s of ChR2 stimulation (middle) and 1 s of sensory stimulation (right). D, Maximum normalized laser speckle contrast response (averaged over 1 s at peak) to no stimulation (left), 1 s of ChR2 stimulation (middle) and 4 s of sensory stimulation (right) in the same animal. C,D Data shown are averaged 20 trials from a representative animal. E, Temporal profiles of IOS responses evoked by 1 s of ChR2 stimulation (n = 23) compared to 1 s of sensory stimulation (n = 12). F, Temporal profiles of laser speckle responses evoked by ChR2 stimulus train durations of 1 s (n = 13) compared to 4 s of sensory stimulation (n = 12). B,E–F Grey and black bars represent onset and duration of stimulation. Blanked data in temporal profiles reflect laser stimulus artifact. The peak amplitudes of ChR2-evoked IOS and speckle contrast responses were compared as a function of distance from the centre of laser activation (G). The spatial extents differed within animals at distances greater than 1000 µm (n = 6, both p<0.001, paired t-test). H, A similar effect was observed in sensory-evoked IOS and speckle contrast responses at distances of 500 and 1000 µm (n = 4, both p<0.05, paired t-test). Error bars represent SEM.

Figure 2. Effect of varying ChR2 stimulus train duration on evoked hemodynamic responses.

A, Temporal profiles of laser speckle responses evoked by 100 ms (grey, n = 7) and 1000 ms of ChR2 stimulation (black, n = 7). B, Peak amplitudes of temporal profiles in (A), compared within the same animals (p = 0.0294, n = 7). C, Temporal profiles of IOS responses evoked by 100 ms (grey, n = 7) and 1000 ms of ChR2 stimulation (black, n = 7). D, Peak amplitudes of temporal profiles in (C), compared within the same animals (p = 0.0012, n = 7).

Figure 3. Blockade of synaptic transmission by CNQX and MK801 inhibits forepaw stimulation responses but not ChR2.

A, left: Pre-antagonist IOS responses to sensory stimulation (yellow, indicated by yellow arrow) and ChR2 stimulation (green, indicated by green arrow) thresholded to 50% of maximum and superimposed on an image of the cortical surface. Right: Temporal evolution of IOS responses to 1 s of sensory stimulation (upper) and ChR2 stimulation (lower). B, left: Post-antagonist IOS response to sensory stimulation is inhibited (yellow arrow indicates pre-antagonist map location), while the IOS response to ChR2 stimulation is preserved (green, indicated by green arrow). Right: Temporal evolution of IOS responses to 1 s of sensory stimulation (upper) and ChR2 stimulation (lower). A,B Shown are the averaged results of 20 trials from a representative animal; blanked data at 0 s represents stimulus artifact. C, Temporal profiles of IOS responses to 1 s of sensory stimulation before and after CNQX/ MK801 incubation (n = 7); responses are displayed until pre-CNQX/MK801 responses return to baseline. D, Temporal profiles of IOS responses to 1 s of ChR2 stimulation before and after CNQX/MK801 incubation (n = 7); blanked data indicate laser stimulus artifact. C,D Stimulus onset denoted by vertical dotted line. E, Peak amplitudes in sensory-evoked IOS responses were significantly reduced following antagonist incubation (paired t-test, p = 0.0009) while peak ChR2-evoked IOS responses were not (paired t-test, p = 0.9358). Error bars represent SEM.

Figure 4. ChR2-evoked blood flow is not inhibited by blocking intracortical ionotropic glutamatergic synaptic transmission by CNQX/MK801.

A, Temporal profiles of laser speckle responses evoked by 1 s of ChR2 stimulation before and after antagonist incubation. B, Comparison of peak amplitudes (averaged over 1 s) in laser speckle contrast responses in (A) before and after antagonist incubation. No significant difference was observed (n = 5, paired t-test, p = 0.0789). C, Temporal profiles of laser speckle responses evoked by forepaw sensory stimulation before and after antagonist incubation. D, Comparison of peak amplitudes (averaged over 1 s) in laser speckle contrast responses in (C) before and after antagonist incubation. A significant difference was observed (n = 8, paired t-test, p = 0.0144). Error bars represent SEM.

Figure 5. ChR2-evoked and sensory-evoked changes in blood flow are partially reduced by the mGluR5 antagonist MPEP.

A, Temporal profiles of laser speckle responses evoked by 1 s of ChR2 stimulation before and after antagonist incubation. B, Comparison of peak amplitudes (averaged over 1 s) in laser speckle contrast responses in (A) before and after antagonist incubation. A significant difference was observed (n = 8, paired t-test, p = 0.0251). C, Temporal profiles of laser speckle responses evoked by sensory stimulation before and after antagonist incubation. D, Comparison of peak amplitudes (averaged over 1 s) in laser speckle contrast responses in (C) before and after antagonist incubation. A significant difference was observed (n = 4, paired t-test, p = 0.0155). E, Temporal profiles of laser speckle responses evoked by ChR2 stimulation before and after injection of saline vehicle. F, Comparison of peak amplitudes (averaged over 1 s) in laser speckle contrast responses in (E) before and after saline injection. No significant difference was observed (n = 3, paired t-test, p = 0.8211). Error bars represent SEM.