Mesoscopic mapping of hemodynamic responses and neuronal activity during pharmacologically induced interictal spikes in awake and anesthetized mice

Abstract

Imaging hemodynamic responses to interictal spikes holds promise for presurgical epilepsy evaluations. Understanding the hemodynamic response function is crucial for accurate interpretation. Prior interictal neurovascular coupling data primarily come from anesthetized animals, impacting reliability. We simultaneously monitored calcium fluctuations in excitatory neurons, hemodynamics, and local field potentials (LFP) during bicuculline-induced interictal events in both isoflurane-anesthetized and awake mice. Isoflurane significantly affected LFP amplitude but had little impact on the amplitude and area of the calcium signal. Anesthesia also dramatically blunted the amplitude and latency of the hemodynamic response, although not its area of spread. Cerebral blood volume change provided the best spatial estimation of excitatory neuronal activity in both states. Targeted silencing of the thalamus in awake mice failed to recapitulate the impact of anesthesia on hemodynamic responses suggesting that isoflurane’s interruption of the thalamocortical loop did not contribute either to the dissociation between the LFP and the calcium signal nor to the alterations in interictal neurovascular coupling. The blood volume increase associated with interictal spikes represents a promising mapping signal in both the awake and anesthetized states.

Keywords

Neurovascular coupling epilepsy interictal event mesoscale optical imaging isoflurane

Introduction

Neurovascular coupling (NVC) refers to the relationship between local neural activity and subsequent changes in cerebral blood flow (CBF). NVC-based hemodynamic changes can be measured non-invasively using functional brain imaging techniques, such as single-photon emission computed tomography (SPECT), functional magnetic resonance imaging (fMRI), and positron emission tomography (PET), which makes these signals useful for human investigations.^1,2 In the management of patients with epilepsy, for example, such imaging techniques are recommended for presurgical evaluation by the neuroimaging subcommission of the International League Against Epilepsy (ILAE).³ Interictal events are attractive targets for hemodynamic-based seizure mapping since they occur frequently and do not cause abnormal movements or behaviors that might disrupt imaging. Therefore, imaging of the interictal period has great potential for clinical practice and interictal imaging may offer valuable information for identifying a seizure focus before surgical intervention.^4,5

Whether hemodynamic signals accurately reflect underlying neuronal activity in a variety of situations, both normal and pathological, has been the subject of numerous studies.^4
–7 One outstanding question concerns the effects of anesthesia on hemodynamic signals since the majority of studies in which both electrical activity and hemodynamic activity can be recorded simultaneously with sufficiently high resolution require laboratory animals to be placed under general anesthesia.^8
–10 Whether these findings apply to unanesthetized states has always been in question since different anesthetics have a variety of effects on NVC mechanisms.^11,12 More recently, awake unanesthetized preparations have been employed to dissociate the effects of anesthesia on the spatial specificity of hemodynamic signals for representing electrical activity.^13
–15 Our lab, for example, has shown that during ictal events, anesthesia blunts the hemodynamic response thereby amplifying focal ischemia within the hypermetabolic ictal focus, which leads to an increase in deoxygenated hemoglobin (Hbr), previously labeled the “epileptic dip”.¹³ The effects of anesthesia on interictal events and the specificity of hemodynamic mapping are less well understood given the paucity of studies with simultaneous high-resolution mapping of neuronal and hemodynamic signals.

In this study, we use wide-field optical imaging with a broad sampling of the neocortex combined with high spatial (tens of micrometers) and temporal (milliseconds) resolution, combined with local field potential (LFP) recording. We performed intrinsic optical spectroscopy (IOS) to examine hemodynamic changes and correlated them with the neuronal activity using genetically encoded calcium indicators in Thy-1 GCaMP6f mice, in whom pyramidal cells are predominantly labeled.¹⁶ Recordings were performed from both isoflurane-anesthetized and awake, behaving unanesthetized animals for comparison.

Methods and materials

Animal preparation and cranial window implantation

All experimental procedures were approved by the Weill Cornell Medical College Animal Care and Use Committee following the National Institutes of Health guidelines. Animal data reporting followed the ARRIVE 2.0 guidelines. Adult (8–12 weeks, 18–25 g) C57BL/6J-Tg(Thy1-GCaMP6f)GP5.5Dkim/J JacksonLab #024276, referred to as Thy1-GCaMP6f thereafter) mice of both sexes were employed in this study (8 mice for awake imaging, 5 for anesthetized imaging, 5 for ipsilateral thalamus inactivation, and 5 for contralateral thalamus inactivation). For cranial window implantation, all mice were anesthetized with isoflurane in 70% N₂:30% O₂, 5% induction, and 1–2% maintenance for the cranial window implantation. During the surgery, body temperature was maintained at 37 °C with a regulated heating blanket (Harvard Apparatus). The heart rate, SpO₂, and the end-tidal carbon dioxide (EtCO₂) were carefully monitored with small animal capnography (Surgivet) and were sustained throughout the experiment (heart rate: 300–450 beat/min, pO₂ > 95%, EtCO₂∼ 25–28 mmHg).

The head was fixed in a stereotaxic apparatus. A ∼ 4.5 × 6.5 mm cranial window was opened on the skull across bilateral hemispheres, between bregma and lambda. The skull within the cranial window, including the superior sagittal sinus area, was carefully removed. A ∼5 × 7 mm sterilized polydimethylsiloxane (PDMS, SYLGARD 184, Dow Corning, USA) film was prepared by mixing the base elastomer and curing agent in the proportion 10:1 (v/v) film (250–350 μm thick) and applied to fully cover the exposed brain tissue.¹⁷ The edges of the PDMS film were carefully fixed to the skull with surgical glue (3 M Vetbond Tissue Adhesive, Japan). A head plate was attached to the skull with dental cement (C&B-METABOND, Parkell Inc., Edgewood, NY). The mice were then individually housed in a 12:12 h light-dark cycle for recovery. Two weeks later, the mice were separated into two groups randomly, half for awake imaging and half for anesthetized imaging.

For the awake head-fixed imaging experiment, the mice were fixed on the imaging chamber for a 30-minute training every day for three continuous days. The head plate was fixed to a clamp to immobilize the head. The mouse was then placed in an air-floated chamber (Neurotar), allowing the animal to artificially ambulate within the chamber.

Acute interictal model and electrophysiology

Interictal-spike (IIS)-like events were triggered by injecting the GABA_A receptor antagonist bicuculline methiodide (BMI) (5 mM in 165 mM NaCl, pH 3.0; Sigma-Aldrich, St. Louis, MO) in both awake and anesthetized experiments. For the anesthetized experiment, the mouse was anesthetized with isoflurane in 70% N₂: 30% O₂, 5% induction, and 1.2–1.5% maintenance. The head was placed in a stereotaxic apparatus. For the awake experiment, the mouse was put in the head-fixed imaging system. 50–250 nl of BMI was injected 300–500 μm below the cortical surface through a glass electrode (50–100 μm tip opening) using a Nanoject II injector (Drummond Scientific, Broomall, PA). We screened the lowest dose that could reliably induce repeatable interictal events. The final dosage differed in the awake and the isoflurane-anesthetized state. Overall, more injections were required in the anesthetized animals to elicit similar stable interictal events. The injection site’s LFP was simultaneously recorded through the BMI injection electrode. The LFP was amplified (1000x), band-pass filtered (1–500 Hz) using an AM Systems 1800 amplifier, digitized using CED Power 1401, and recorded via a PC running Spike2 software (Cambridge Electronic Design, Cambridge, UK).

Widefield optical imaging

A “temporal separation” technique was employed to simultaneously image wide-field calcium and a multispectral intrinsic optical signal (IOS).^18,19 A CCD camera (J-MC023MGSY, Lighting Mind Inc., Changchun, China) using a tandem lens (85 × 50 mm) arrangement was focused 300–400 μm below the cortical surface. Three LEDs with coupled bandpass filters were employed as the illumination source, including a “blue” LED (470 ± 10 nm) for calcium imaging, a “green” LED (530 ± 10 nm), and a “red” LED (610 ± 10 nm) for IOS imaging. The illumination was directed to the cortex using fiber-optic light guides. A 510 nm long-pass filter was placed before the camera to prevent blue illumination contamination while permitting both calcium fluorescence signal and IOS. The ultra-fast multispectral switching among the three LEDs was time-locked to camera frames. An Arduino UNO board was used to couple the frame indicator between the camera and the LEDs. The calcium imaging was performed every other frame and green and red IOS was performed every 4th frame. The camera was running with a frame rate of 120 Hz, resulting in 60 Hz imaging for calcium and 30 Hz for green and red IOS, respectively.

Thalamic inhibition

To investigate the role of the thalamus in neurovascular coupling during IISs, we silenced the thalamus either ipsilateral or contralateral to the IIS generator (BMI injection site). To do this, tetrodotoxin (TTX) (5 µM, 1 µl) was injected either into the ipsilateral or the contralateral thalamus ∼60 minutes after the onset of the interictal events. A glass electrode filled with TTX was placed in the ventrobasal region: 1.88 mm posterior to Bregma, 1.5 mm lateral to the midline, and 3.3–3.8 mm ventral to the surface of the cortex on one or the other side to target both the ventral posteromedial nucleus (VPM) and the ventral posterolateral nucleus (VPL). Given the volume of fluid injected and the known diffusion of TTX in the thalamus, we estimated that the mean inactivated nervous tissue radius was 0.7–0.8 mm.²⁰ The optical imaging was performed before and after the TTX injection. The location of the TTX injection in the thalamus was confirmed in a pilot study with an injection of 1 μl blue dye RH-1692 (Optical Imaging Ltd, Supplementary Figure 1).

Data analysis

Custom-written software in MATLAB 2018 A (The Mathworks, Natick, MA, USA) was used for data processing and statistical analysis. The pulsation artifact from the heartbeat was eliminated with an offline algorithm.^18,21 Briefly, an average QRS interval was obtained for each pixel in each trial. The peaks of R waves were obtained from ECG and the maximal duration of the heartbeat cycle was detected. An averaged pulsation artifact was obtained by an R wave-triggered average using the maximal duration. This averaged pulsation artifact was repeatedly subtracted from each heartbeat cycle of the original data. To increase the signal-to-noise ratio, imaging data were convolved with a spatial Gaussian kernel (σ = 3 pixels).

A modified Beer-Lambert law with a path length correction factor was used to calculate the concentration of deoxygenated hemoglobin (Hbr), oxygenated hemoglobin (HbO), and total hemoglobin (HbT) change using 530 nm and 610 nm IOS data.²² A 2 Hz Butterworth low pass filter (1 order) was applied to the hemodynamic signal to reduce high-frequency noise.

For calcium signal processing, the functional hemodynamic artifact was separated using an approach previously described by Kramer and Pearlstein (1979)²³ and Ma et al., (2016).²⁴ In this approach, the calcium fluoresce intensity change was calculated using the following equitation:

F_{True} (t) / F_{True} (t_{0}) = (F (t) / F (t_{0}) / I (t) / I (t_{0}))

where F_True is the calcium fluoresce intensity absenting hemodynamic artifact, F is the recorded calcium fluoresce intensity, I is the recorded IOS signal at 530 nm, and t₀ is the time point for baseline.

The calcium and IOS were analyzed in a similar fashion using frame division triggered by the onset of the IIS. IIS-like events were identified from LFP. A 300-frame window (60 frames before and 240 frames after) was averaged over all IISs. This resulted in an average spike-triggered movie of reflection or fluorescence changes over time. The calcium imaging data was calculated as dF/F, where F was the baseline illumination level defined as an average of the three frames prior to IIS and dF was the signal change during IIS activity as identified by the LFP. Isoflurane induction elicited a transit change in hemodynamic baseline which become stable after ∼15 minutes (Supplementary Figure 2). Therefore, data acquisition started 20 minutes after isoflurane induction. The difference in baseline between awake and anesthetized states was small (∼1% in HbO, ∼0.5% in HbT, and <0.1% in Hbr, Supplementary Figure 2).

In order to calculate the spatial extent of calcium signal and IOS, a modified Chen-Bee method was employed.^25,26 Briefly, 30% of the maximal amplitude of the optical signal measured from the BMI injection site was selected as a threshold. All pixels with an amplitude above the threshold were considered active pixels. The area of spread was calculated by multiplying the number of active pixels by the area of each pixel (29.34 × 29.34 μm² per pix).

Since the hemodynamic amplitudes were affected by the inter-event interval²⁷ (Supplementary Figure 3), we chose only IISs with a pre-IIS interval ranging from 1.5 to 2.5 seconds to eliminate events with extremely high or low amplitudes.

Results

Anesthesia has a disproportionate effect on the hemodynamic response

IISs were recorded in the LFP a few minutes after BMI injection and reliably lasted for 2–3 hours in both the anesthetized and awake animals. As revealed in the LFP, the frequency of IISs was slightly higher in awake (0.742 ± 0.255 Hz; n = 2637 IISs, n = 8 animals) compared with anesthetized mice (0.564 ± 0.248 Hz; n = 889 IISs, n = 5 animals; p < 0.001, 2-tailed, unpaired t-test). Likewise, the amplitude of the IISs was larger in awake 1.42 ± 0.40 mV (n = 2637 IISs, n = 8 animals) compared with anesthetized mice (0.88 ± 0.23 mV; n = 889 IISs, n = 5 animals; p = 0.019, 2-tailed, unpaired t-test).

Each IIS elicited a clear and reproducible neuronal calcium concentration change along with an associated hemodynamic signal, both of which were spatially concordant with the LFP location and the presumed site the population of neurons participating in the IIS and the hemodynamic response to that activity (Figure 1). Signals arose predominantly from the ipsilateral hemisphere, with smaller signals recorded contralaterally.

Figure 1.

Impact of Anesthesia on calcium and hemodynamic changes during IISs. (a) Awake mouse cortex. The red box indicates a 3 × 3-pixel ROI covering the electrode tip, from which the LFP and calcium traces were derived. (b) Spatiotemporal dynamics of Ca and hemodynamic traces. The average (n = 347 events) Ca and hemodynamic traces from the ROI are shown on the left. Five vertical black lines corresponding to the baseline, onset, half maximum amplitude (rise), maximum amplitude, and half maximum amplitude (decay) indicate the time points from which the images are obtained on the right. (c) Anesthetized mouse cortex and (d) Spatiotemporal dynamics of Ca and hemodynamic traces. The average (n = 204 events) Ca and hemodynamic traces from the ROI are shown on the left. Five vertical black lines corresponding to the baseline, onset, half maximum amplitude (rise), maximum amplitude, and half maximum amplitude (decay) indicate the time points from which the images are obtained on the right. Note that the amplitude, ipsilateral, and contralateral areas of spread are much higher and broader in the awake state. Also, the initial increase in Hbr (initial ‘epileptic’ dip) is briefer in the awake state with a more profound delayed decrease in Hbr.

The ipsilateral calcium response was slightly larger and faster in the awake state compared with the anesthetized state, but these differences did not reach statistical significance. The awake calcium signal peaked at 0.096 ± 0.023 s (n = 2637 IISs; n = 8 animals) and decayed to 5% of the maximal amplitude in 0.768 ± 0.117 s (p > 0.05 in all comparisons, two-tailed, paired t-test). In anesthetized mice, the calcium signal peaked at 0.107 ± 0.013 s (n = 889 IISs; n = 5 animals) and decayed to 5% of the maximal amplitude in 0.980 ± 0.2452 s (p > 0.05 in all comparisons, two-tailed, paired t-test) (Figures 1 and 2). Thus, anesthesia caused a disproportionate dampening of the LFP compared with the calcium signal.

Figure 2.

Anesthesia decreases amplitude and increases the latency of ipsilateral hemodynamic response. (a) Averaged waveform of calcium and hemodynamic signals recorded from the BMI injection site in awake and anesthetized mice (mean ± sd). (b–c) Box plot of the amplitude (b) and time to peak (c) of calcium and hemodynamic signals in awake and anesthetized mice. Note that anesthesia decreases the amplitude and increases the latency of both signals but the effect on the hemodynamic signal is much more profound *: p < 0.05, **: p < 0.01, ***: p < 0.001.

The contralateral spread of IISs was recorded by imaging both hemispheres simultaneously. The contralateral calcium response was much smaller in the contralateral cortex than in the ipsilateral cortex in both awake and anesthetized mice. As with the ipsilateral responses, anesthesia had little impact on the peak amplitude or latency of the calcium signal (Figure 3).

Figure 3.

Anesthesia nearly eliminates contralateral hemodynamic response. (a) Averaged waveform of calcium and hemodynamic signals recorded from site contralateral to the BMI injection site in awake and anesthetized mice (mean ± sd). (b–c) Box plot of the amplitude (b) and time to peak (c) of calcium and hemodynamic signal in awake and anesthetized mice. Note that anesthesia dramatically decreases the amplitude and increases the latency of the hemodynamic signal with minimal effect on the calcium signal. Note: although it appears that the calcium signal is larger in the anesthetized state this was not statistically significant. *: p < 0.05, **: p < 0.01, ***: p < 0.001.

The hemodynamic responses, on the other hand, were disproportionally impacted by anesthesia compared with the calcium signals, both on the ipsilateral as well as the contralateral side. In the ipsilateral hemisphere, the HbT signal showed a monophasic increase that was both lower in amplitude and slower to evolve under anesthesia (Figure 2). The HbO and Hbr signals were biphasic, consistent with a short “initial ‘epileptic’ dip” in hemoglobin oxygenation followed by a delayed hyperoxygenation corresponding with the increase in HbT. Anesthesia dramatically dampened and delayed these responses ((p < 0.001 for all tests, two-tailed, paired t-test; Figures 2 and 3). The average percent decreases in the amplitude of the HbT, HbO, and Hbr (dip and overshoot) caused by anesthesia were 66.01%, 71.51%, 89.07%, and 73.34% respectively. The average delays in time to peak for these signals caused by anesthesia were 54.69%, 49.04%, 62.03%, and 45.59% respectively.

Although the calcium amplitude was very low in the contralateral cortex, the hemodynamic response, at least in the awake animal, remained measurable. In awake mice, clear HbT, HbO, and Hbr changes were recorded simultaneously with the IISs from the homotopic mirror cortex. While the morphology of these contralateral hemodynamic responses were similar to the ipsilateral responses, the peak amplitudes were significantly smaller (p < 0.001 for all tests, two-tailed, paired t-test; Figures 2 and 3).

Anesthesia caused dramatic decreases in amplitude and delays in time to peak for all contralateral hemodynamic signals. The average percent decrease in the amplitude of the HbT, HbO, and Hbr (dip and overshoot) caused by anesthesia was 83.68%, 93.53%, 72.13%, and 58.61% respectively. The average delays in time to peak for these signals caused by anesthesia were 19.20%, 44.13%, 66.50%, and 46.31% respectively (Figure 3). Overall, our data indicate that while anesthesia had a minimal impact on the calcium response, there was a dramatic impact on both the LFP and the hemodynamic responses, both ipsilaterally and contralaterally.

The spatial extent of the calcium and hemodynamic signals

Given the difference in hemodynamic amplitude in awake versus anesthetized animals, it remained unclear if the hemodynamic signals could be used to map the area of cellular activity under both conditions, as revealed by the calcium signal. We thus investigated the impact of isoflurane on the spatial spread of the calcium and hemodynamic signals (Figure 4(a) and (b)).²⁶ Once again, calcium responses were not significantly affected (p = 0.560, unpaired, two-tailed t-test). Surprisingly, unlike what occurred with the amplitude of the hemodynamic signal, the spatial spread of the hemodynamic signal in the ipsilateral hemisphere was minimally altered by isoflurane (p > 0.05 in all pairs, unpaired, two-tailed t-test, Figure 4). In the contralateral cortex, the calcium and hemodynamic signals were too small in the anesthetized animals to measure their areas. In the awake animals, the area could only be measured from the hemodynamic signals, so they could not be correlated with a calcium signal (Figure 4(c)).

Figure 4.

Spatial mapping and overlap between the calcium and hemodynamic signals. (a) Top: The field of view of the imaged cortex in a single mouse. The red point indicates the BMI injection site. The two red circles outline the imaged area in the ipsilateral and contralateral cortex. The surrounding bone and the central vein are excluded from the area calculation. Bottom: The maximal area of calcium signal detected with the modified Chen-bee method. (b) Top: the maximal area of HbT, HbO, Hbr increase, and Hbr decrease. Bottom: The overlap between the maximal calcium and maximal hemodynamic areas in the ipsilateral cortex with a two-dimensional correlation coefficient indicated. (c) Box plot of the maximal area of calcium and hemodynamic change in awake and anesthetized mice. Left: ipsilateral to the BMI injection side; Right: contralateral cortex and (d) Box plot of the correlation coefficients between calcium and hemodynamic signals in awake (left) and anesthetized (right) mice. *: p < 0.05, **: p < 0.01, ***: p < 0.001.

Restricting our analysis to the ipsilateral hemisphere, we then measured the spatial correlation between the calcium and hemodynamic signals during IISs to see how anesthesia impacted the ability of the hemodynamic signals to reflect the area of neuronal activation. Two-dimensional correlation coefficients between the spatial spread of calcium and different hemodynamic components were calculated in the ipsilateral cortex (Figure 4(d)). In both the awake and the anesthetized state, the area of the HbT signal had the highest spatial correlation with the area of calcium signal (Awake ANOVA: p = 0.014, f = 4.240, degree of freedom = 31, Anesthetized ANOVA: p < 0.001, f = 18.871, degree of freedom = 19).

Our spatial area analysis indicates that anesthesia dramatically blunted the hemodynamic response in the contralateral cortex and that HbT is the best surrogate of neuronal activity, in both the awake and the anesthetized state.

Thalamic inhibition in awake mice

After demonstrating that anesthesia had a disproportionate blunting effect on the hemodynamic response, and most dramatically in the contralateral cortex, we investigated the role of the thalamus in mediating this response as the potential target of anesthesia, since one of the primary mechanisms whereby general anesthesia causes unconsciousness is by disruption of thalamocortical connectivity.²⁸ To test this hypothesis, we injected TTX into the ipsilateral and then the contralateral VPM/VPL in awake behaving mice to silence thalamic activity ipsilateral or contralateral to the IIS focus to determine how this would impact the hemodynamic response. Since thalamic activity amplifies ipsilateral epileptic activity and the corpus callosum mediates contralateral propagation,²⁹ we hypothesized that ipsilateral thalamic inactivation would blunt both the ipsilateral calcium and the hemodynamic responses whereas contralateral inactivation would disproportionally blunt only the contralateral hemodynamic response.

When TTX was injected into the ipsilateral VPM/VPL in awake mice, unlike what was seen with anesthesia, we found a decrease in the ipsilateral calcium amplitude by 41.01% (p = 0.026, n = 5 mice, paired, two-tailed t-test), and a decrease in the time to peak by 16.67%. The LFP amplitude was also significantly reduced by 23% from 3.42 ± 0.78 mV to 2.63 ± 0.81 mV (p = 0.0036, n = 5 mice, paired, two-tailed t-test). A corresponding decrease in the amplitude of the HbT response occurred by 56.31% (p = 0.043), as well as in HbO by 61.56% (p = 0.040) (Figure 5). Surprisingly, Hbr was unaffected. The calcium and hemodynamic responses seemed unaffected in the contralateral hemisphere (p > 0.05).

Figure 5.

Changes in Calcium and hemodynamic signals induced by ipsilateral thalamic TTX injection. (a) Box plots of calcium and hemodynamic amplitude in both hemispheres before and after TTX injection in the ipsilateral thalamus and (b) Box plot of the peak time of calcium and hemodynamic signals in both hemispheres. Note: no significant change in the contralateral calcium and hemodynamic responses. *: p < 0.05, **: p < 0.01, ***: p < 0.001.

Silencing contralateral thalamic activity did not affect ipsilateral LFP (3.67 ± 1.64 before TTX v.s. 3.52 ± 1.62 mV after TTX, p = 0.869, n = 5 mice, paired, two-tailed t-test), calcium, or hemodynamic responses, but caused significant (p = 0.048 paired, two-tailed test, n = 5 mice) (45.75%) reductions in the amplitude of the contralateral calcium signal and a significant reduction in the contralateral hemodynamic responses (Hbt: p = 0.046 68.1% reduction; HbO: p = 0.032 and 69.61% reduction; Hbr decrease: p = 0.002 76.63% reduction)). (Figure 6(a)).

Figure 6.

Changes in calcium and hemodynamic signals induced by contralateral thalamic TTX injection. (a) Box plots of calcium and hemodynamic amplitude from both hemispheres before and after TTX injection in the contralateral thalamus and (b) Box plot of the peak time of calcium and hemodynamic signals in both hemispheres. Note: Significant change in the contralateral hemodynamic responses. *: p < 0.05, **: p < 0.01, ***: p < 0.001.

These experiments show that the effects of anesthesia on the hemodynamic responses are not a result of a disruption in the amplification effect of the thalamocortical loop since direct thalamic inactivation proportionally blunted the LFP, the calcium signal and the hemodynamic response whereas anesthesia had a disproportionate blunting effect on only the LFP and the hemodynamic signals.

Discussion

NVC-based hemodynamic responses (including CBF, CBV, and hemoglobin oxygenation^30
–33) form the basis of a variety of brain imaging modalities used both in the laboratory and the clinic as surrogates for neuronal activity. A detailed understanding of the hemodynamic response function is required for the correct interpretation of these signals, which can be altered by disease states such as epilepsy. In clinical practice, the hemodynamic response is often measured with EEG-correlated fMRI,^5,34,35 which has limited spatial and temporal resolution, making data analysis challenging.³⁶ In an effort to understand the origins of these signals, prior laboratory studies have coupled IOS with local field potentials, voltage-sensitive dye, or calcium imaging in anesthetized rodents.^8,37,38 The data from these studies is compromised by the fact that anesthesia can have a profound effect on neurovascular coupling, and therefore, to fully understand the effects of anesthesia, data acquired in awake animals is critical for comparison.

In this study, we used mesoscale optical imaging of excitatory neuronal (GCaMP6f) and hemodynamic changes (IOS) during BMI-induced interictal events in both isoflurane-anesthetized and awake mice to understand the impact of anesthesia on the interictal spike-related hemodynamic signals and provide data that are free of the effects of isoflurane and which are directly applicable to clinical epileptology imaging. Since isoflurane doesn’t just affect neurovascular coupling but also can impact neuronal activity, we first had to sort out how this anesthetic altered the neuronal signals. We first found that isoflurane differentially altered the LFP and the calcium signals. While the LFP signals were diminished by isoflurane, the calcium signals were not significantly altered. We also show that isoflurane blunted the amplitude and latency of the hemodynamic response, although not its area of spread. Whether this reflects merely a decrease in neuronal activation, or a dampening of the neurovascular coupling response is more difficult to discern and requires further discussion. We also confirm that in awake behaving animals, the CBV signal, recorded as the change in HbT, provided the best spatial estimation of neuronal activation, although the increase in Hbr and HbO were spatiotemporally correlated with neuronal activation as well.

Isoflurane effect on the neuronal activity during interictal events

Anesthetics affect neuronal activity in several ways. In general, anesthetics hyperpolarize neurons by increasing inhibition and/or decreasing excitation,^39,40 which usually leads to suppression in the average firing rate of neurons.^41,42 Isoflurane exerts its inhibitory effects on the central nervous system mainly via the potentiation of GABA-A receptors, which hyperpolarize the cell membrane.⁴³ Isoflurane also inhibits presynaptic sodium channels and action potential-evoked synaptic vesicle exocytosis by inhibiting presynaptic calcium influx into glutamatergic neurons to a greater degree than GABAergic neurons.⁴⁴ During sensory stimulation, for example, evoked neuronal responses are significantly delayed,^45,46 and their receptive fields are reduced under anesthesia.^47
–49

Isoflurane differentially altered the LFP and the calcium signal. There are several possible interpretations of this. The first is that these two signals represent different aspects of neuronal activation. Indeed, the LFP is derived predominantly from pre- and post-synaptic excitatory and inhibitory activity,^50,51 while the GCamp6-Thy1 calcium signal emanates mostly from excitatory cell cytosolic calcium concentration change primarily derived from action potentials.^16,52 While it is possible that anesthetics decreased synaptic activity but not cell spiking, this is unlikely for two reasons. First, the synaptic activation constituting the LFP signal is derived in large part from action potentials, so the two should be related. Second, the more likely explanation is that the calcium signal is not directly related to the frequency and number of spiking cells, since the calcium signals cannot reliably distinguish spiking frequencies above 3 Hz.⁵² Thus, a decreased reduction in cell spiking that remained above the 3 Hz threshold would be reflected in the LFP, as the calcium fluorescence might have been saturated. Since the temporal resolution of the LFP is very high, whereas the temporal resolution of the calcium signal is quite low, GCamp6 cannot discriminate high spike frequencies. Thus, if isoflurane either decreased pre- and post-synaptic activity, particularly in the interneuron pool, and/or decreased spike frequencies at ranges that remain above 3 Hz, the LFP would be altered but the calcium signal would remain relatively unchanged.

This dampening effect of anesthesia on interictal spikes, which required higher doses of BMI to achieve the same size interictal spike, is not unique to interictal spikes but also true of ictal activity, as we found in the acute 4-AP model.¹³ The anesthetic effect on the calcium signal, on the other hand, differed between the interictal and the ictal model. Whereas anesthesia only slightly diminished the calcium signal in the interictal model, it substantially blunted the rapid fluctuations and build-up of the calcium signal during ictal events.¹³ This difference is multifactorial. Firstly, ictal events last much longer than interictal ones, so the dampening effect of anesthetics on pre-synaptic calcium influx would multiply over time. Second, the slow temporal discrimination and decay of the calcium signal acts like a low pass filter for the epileptiform events, and the longer duration of ictal events results in a plateau in the accumulation of intracellular calcium, which does not occur during shorter duration interictal events.

Isoflurane effect on hemodynamic response during interictal events

Most anesthetics, including isoflurane, not only suppress neural activity but affect neurovascular coupling.⁵³ Isoflurane increases the vascular membrane permeability,⁵⁴ increases the membrane transport of endothelial cells,⁵⁵ and affects pericytes that control the vasoactive constriction/relaxation. Isoflurane is a potent vasodilator through its action on ATP-sensitive potassium channels and calcium currents in smooth muscle cells.^56
–58 Laser-Doppler studies show a reduction in both red blood cell velocity and concentration caused by isoflurane.⁵⁹ Isoflurane also binds to hemoglobin and changes the oxygen-carrying capacity of hemoglobin.⁶⁰

What is less clear from our data is the degree to which isoflurane’s impact on the hemodynamic response derives from the reduction in synaptic activity, as recorded by the LFP, versus a direct effect on vascular reactivity. Likely, both effects are meaningful since, during sensory processing, anesthetics suppress the hemodynamic response out of proportion with the dampening of neuronal responses themselves.^6,61 For example, the magnitude of the decrease in BOLD signal, CBV, or CBF is larger than the reduction of neuronal activity measured with multi-unit or LFP in rodent somatosensory and visual cortex.^45,61 Anesthesia has been shown to delay hemodynamic responses to normal sensory processing up to 2 s in several different species.^6,45,62

Interestingly, the effect of anesthesia on interictal neurovascular coupling was different than what we had found during ictal events. In the 4-AP model, the anesthetic blunting of the hemodynamic response caused a profound increase in ictal ischemia, or an ‘epileptic dip, since sustained metabolic demand could be met in the face of diminished blood flow.¹³ Interictal events, on the other hand, are so short-lived, that metabolic demand is met even under anesthesia and the dip is not augmented. In fact, the dip is diminished by anesthesia, likely from the decrease in the frequency of interictal events and an overall decrease in local metabolic demand.

Anesthetic effect on the neurovascular coupling is not related to thalamocortical dampening

One of the primary mechanisms whereby anesthesia causes unconsciousness is by disconnecting thalamocortical connectivity.^28,63 The thalamocortical loop is also known to amplify ipsilateral epileptic activity.²⁹ We investigated the extent to which the effects of anesthesia on neurovascular coupling were dependent on its effect on thalamocortical connectivity by inactivating one, and then the other, half of the thalamus. Ipsilateral silencing of the thalamus dampened ipsilateral LFP, calcium and hemodynamic responses whereas isoflurane only dampened LFP and the hemodynamic response, indicating that Isoflurane dampens both neuronal activity and vascular reactivity whereas thalamic inactivation only dampens neuronal activity, which, in turn, leads to a proportional decrease in the hemodynamic signal. Contralateral thalamic silencing dampened predominantly contralateral calcium and hemodynamic responses. Thus, thalamocortical connectivity does not appear to have an independent effect on cortical neurovascular coupling beyond its effect as an amplifier of epileptic activity. These findings also highlight the possibility of non-neuronal cells contributing to the spread of epilepsy.⁶⁴

In summary, we show that the anesthetic blunting of cortical neurovascular coupling mechanisms is derived both from a reduction in synaptic activity and a direct vasodilatory effect. Nevertheless, hemodynamic mapping of the interictal focus, whether under awake or anesthetized states, remains spatially accurate and HbT signals provide the best means for hemodynamic-based localization of the epileptic focus. Technical advances in chronic mapping of either calcium dynamics or HbT in human patients may be useful clinically to map ictal onset and propagation.

Supplemental Material

sj-pdf-1-jcb-10.1177_0271678X241226742 - Supplemental material for Mesoscopic mapping of hemodynamic responses and neuronal activity during pharmacologically induced interictal spikes in awake and anesthetized mice

Supplemental material, sj-pdf-1-jcb-10.1177_0271678X241226742 for Mesoscopic mapping of hemodynamic responses and neuronal activity during pharmacologically induced interictal spikes in awake and anesthetized mice by Jing Li, Fan Yang, Fengrui Zhan, Joshua Estin, Aditya Iyer, Mingrui Zhao, James E Niemeyer, Peijuan Luo, Dan Li, Weihong Lin, Jyun-you Liou, Hongtao Ma and Theodore H Schwartz in Journal of Cerebral Blood Flow & Metabolism

Footnotes

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by the National Natural Science Foundation of China (No. 81971205), the Jilin Province Science and Technology Development Plan Item (NO. YDZJ202101ZYTS084), and the China Scholarship Council Grant (201906170253).

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability

Any data relating to this study will be made available, within the bounds of our institution's formal guidelines and within what would otherwise be considered ethically appropriate.

Authors’ contributions

HM, THS, WL – designed the study, JLi, FY, FZ, JE, MZ, JEN, PL, DL, AI acquired and analyzed the data, JLiou, HM, THS– wrote the manuscript.

ORCID iDs

James E Niemeyer

Hongtao Ma

Supplementary material

Supplemental material for this article is available online.

References

Van Paesschen

Ictal SPECT. Epilepsia 2004; 45 Suppl 4: 35–40.

Bagshaw

Kobayashi

Dubeau

, et al. Correspondence between EEG-fMRI and EEG dipole localisation of interictal discharges in focal epilepsy. Neuroimage 2006; 30: 417–425.

(ILAE) NSotILAE. Commission on diagnostic strategies: recommendations for functional neuroimaging of persons with epilepsy. Epilepsia 2000; 41: 1350–1356.

von Oertzen

TJ.

PET and ictal SPECT can be helpful for localizing epileptic foci. Curr Opin Neurol 2018; 31: 184–191.

Khoo

Hao

von Ellenrieder

, et al. The hemodynamic response to interictal epileptic discharges localizes the seizure-onset zone. Epilepsia 2017; 58: 811–823.

Logothetis

Pauls

Augath

, et al. Neurophysiological investigation of the basis of the fMRI signal. Nature 2001; 412: 150–157.

Phillips

Chan

Zheng

, et al. Neurovascular coupling in humans: physiology, methodological advances and clinical implications. J Cereb Blood Flow Metab 2016; 36: 647–664.

Zhao

Suh

, et al. Hemodynamic surrogates for excitatory membrane potential change during interictal epileptiform events in rat neocortex. J Neurophysiol 2009; 101: 2550–2562.

Zhao

Schwartz

TH.

Dynamic neurovascular coupling and uncoupling during ictal onset, propagation, and termination revealed by simultaneous in vivo optical imaging of neural activity and local blood volume. Cereb Cortex 2013; 23: 885–899.

10.

Baird-Daniel

Daniel

AGS

Wenzel

, et al. Glial calcium waves are triggered by seizure activity and not essential for initiating ictal onset or neurovascular coupling. Cereb Cortex 2017; 27: 3318–3330.

11.

Franceschini

Radhakrishnan

Thakur

, et al. The effect of different anesthetics on neurovascular coupling. Neuroimage 2010; 51: 1367–1377.

12.

Gao

Zhang

, et al. Time to wake up: studying neurovascular coupling and brain-wide circuit function in the un-anesthetized animal. Neuroimage 2017; 153: 382–398.

13.

Yang

Song

, et al. Mesoscopic mapping of ictal neurovascular coupling in awake behaving mice using optical spectroscopy and genetically encoded calcium indicators. Front Neurosci 2021; 15: 704834.

14.

Sharp

Shaw

Boorman

, et al. Comparison of stimulus-evoked cerebral hemodynamics in the awake mouse and under a novel anesthetic regime. Sci Rep 2015; 5: 12621.

15.

Krogsgaard

Sperling

Dahlqvist

, et al. PV interneurons evoke astrocytic Ca(2+) responses in awake mice, which contributes to neurovascular coupling. Glia 2023; 71: 1830–1846.

16.

Jercog

Rogerson

Schnitzer

MJ.

Large-scale fluorescence calcium-imaging methods for studies of long-term memory in behaving mammals. Cold Spring Harb Perspect Biol 2016; 8: a021824.

17.

Heo

Park

Kim

, et al. A soft, transparent, freely accessible cranial window for chronic imaging and electrophysiology. Sci Rep 2016; 6: 27818.

18.

Harris

Rahmani

, et al. Wide-field in vivo neocortical calcium dye imaging using a convection-enhanced loading technique combined with simultaneous multiwavelength imaging of voltage-sensitive dyes and hemodynamic signals. Neurophotonics 2014; 1: 015003.

19.

Zhao

Harris

, et al. Simultaneous multi-wavelength optical imaging of neuronal and hemodynamic activity. In: Zhao

Schwartz

(eds) Neurovascular coupling methods. New York: Springer, 2014, pp.237–249.

20.

Sacchetti

Lorenzini

Baldi

, et al. Auditory thalamus, dorsal hippocampus, basolateral amygdala, and perirhinal cortex role in the consolidation of conditioned freezing to context and to acoustic conditioned stimulus in the rat. J Neurosci 1999; 19: 9570–9578.

21.

JY.

Initiation of spontaneous epileptiform events in the rat neocortex in vivo. J Neurophysiol 2004; 91: 934–945.

22.

Sheth

Nemoto

Guiou

, et al. Linear and nonlinear relationships between neuronal activity, oxygen metabolism, and hemodynamic responses. Neuron 2004; 42: 347–355.

23.

Kramer

Pearlstein

RD.

Cerebral cortical microfluorometry at isosbestic wavelengths for correction of vascular artifact. Science 1979; 205: 693–696.

24.

Pellegrino

Machado

von Ellenrieder

, et al. Hemodynamic response to interictal epileptiform discharges addressed by personalized EEG-fNIRS recordings. Front Neurosci 2016; 10: 102.

25.

Chen-Bee

Kwon

Masino

, et al. Areal extent quantification of functional representations using intrinsic signal optical imaging. J Neurosci Methods 1996; 68: 27–37.

26.

Berwick

Johnston

Jones

, et al. Fine detail of neurovascular coupling revealed by spatiotemporal analysis of the hemodynamic response to single whisker stimulation in rat barrel cortex. J Neurophysiol 2008; 99: 787–798.

27.

Geneslaw

Zhao

, et al. Tissue hypoxia correlates with intensity of interictal spikes. J Cereb Blood Flow Metab 2011; 31: 1394–1402.

28.

Hemmings

Jr Riegelhaupt

Kelz

, et al. Towards a comprehensive understanding of anesthetic mechanisms of action: a decade of discovery. Trends Pharmacol Sci 2019; 40: 464–481.

29.

Brodovskaya

Batabyal

Shiono

, et al. Distinct roles of rodent thalamus and corpus callosum in seizure generalization. Ann Neurol 2022; 91: 682–696.

30.

Girouard

Iadecola

Neurovascular coupling in the normal brain and in hypertension, stroke, and Alzheimer disease. J Appl Physiol (1985) 2006; 100: 328–335.

31.

Tian

Teng

May

, et al. Cortical depth-specific microvascular dilation underlies laminar differences in blood oxygenation level-dependent functional MRI signal. Proc Natl Acad Sci U S A 2010; 107: 15246–15251.

32.

Liu

Hirano

Nascimento

, et al. fMRI in the awake marmoset: somatosensory-evoked responses, functional connectivity, and comparison with propofol anesthesia. Neuroimage 2013; 78: 186–195.

33.

Hall

Reynell

Gesslein

, et al. Capillary pericytes regulate cerebral blood flow in health and disease. Nature 2014; 508: 55–60.

34.

Gotman

Epileptic networks studied with EEG-fMRI. Epilepsia 2008; 49 Suppl 3: 42–51.

35.

Thornton

Laufs

Rodionov

, et al. EEG correlated functional MRI and postoperative outcome in focal epilepsy. J Neurol Neurosurg Psychiatry 2010; 81: 922–927.

36.

Van Eyndhoven

Dupont

Tousseyn

, et al. Augmenting interictal mapping with neurovascular coupling biomarkers by structured factorization of epileptic EEG and fMRI data. Neuroimage 2021; 228: 117652.

37.

Suh

Bahar

Mehta

, et al. Temporal dependence in uncoupling of blood volume and oxygenation during interictal epileptiform events in rat neocortex. J Neurosci 2005; 25: 68–77.

38.

Arnal-Real

Mahmoudzadeh

Manoochehri

, et al. What triggers the interictal epileptic spike? A multimodal multiscale analysis of the dynamic of synaptic and non-synaptic neuronal and vascular compartments using electrical and optical measurements. Front Neurol 2021; 12: 596926.

39.

Alkire

Hudetz

Tononi

Consciousness and anesthesia. Science 2008; 322: 876–880.

40.

Ries

Puil

Mechanism of anesthesia revealed by shunting actions of isoflurane on thalamocortical neurons. J Neurophysiol 1999; 81: 1795–1801.

41.

Constantinople

Bruno

RM.

Effects and mechanisms of wakefulness on local cortical networks. Neuron 2011; 69: 1061–1068.

42.

de Kock

Sakmann

Spiking in primary somatosensory cortex during natural whisking in awake head-restrained rats is cell-type specific. Proc Natl Acad Sci U S A 2009; 106: 16446–16450.

43.

Ito

Izumi

Sato

, et al. Suppression of parasympathetic reflex vasodilatation in the lower lip of the cat by isoflurane, propofol, ketamine and pentobarbital: implications for mechanisms underlying the production of anaesthesia. Br J Anaesth 1998; 81: 563–568.

44.

Baumgart

Zhou

Hara

, et al. Isoflurane inhibits synaptic vesicle exocytosis through reduced Ca2+ influx, not Ca2+-exocytosis coupling. Proc Natl Acad Sci U S A 2015; 112: 11959–11964.

45.

Martin

Martindale

Berwick

, et al. Investigating neural-hemodynamic coupling and the hemodynamic response function in the awake rat. Neuroimage 2006; 32: 33–48.

46.

Sellers

Bennett

Hutt

, et al. Awake vs. anesthetized: layer-specific sensory processing in visual cortex and functional connectivity between cortical areas. J Neurophysiol 2015; 113: 3798–3815.

47.

Armstrong-James

George

MJ.

Influence of anesthesia on spontaneous activity and receptive field size of single units in rat Sm1 neocortex. Exp Neurol 1988; 99: 369–387.

48.

Chapin

Woodward

DJ.

Modulation of sensory responsiveness of single somatosensory cortical cells during movement and arousal behaviors. Exp Neurol 1981; 72: 164–178.

49.

Friedberg

Lee

Ebner

FF.

Modulation of receptive field properties of thalamic somatosensory neurons by the depth of anesthesia. J Neurophysiol 1999; 81: 2243–2252.

50.

Haider

Schulz

Hausser

, et al. Millisecond coupling of local field potentials to synaptic currents in the awake visual cortex. Neuron 2016; 90: 35–42.

51.

Teleńczuk

Dehghani

Le Van Quyen

, et al. Local field potentials primarily reflect inhibitory neuron activity in human and monkey cortex. Sci Rep 2017; 7: 40211.

52.

Hartung

Gold

MS.

GCaMP as an indirect measure of electrical activity in rat trigeminal ganglion neurons. Cell Calcium 2020; 89: 102225.

53.

Slupe

Kirsch

JR.

Effects of anesthesia on cerebral blood flow, metabolism, and neuroprotection. J Cereb Blood Flow Metab 2018; 38: 2192–2208.

54.

Noorani

Chowdhury

Alqahtani

, et al. Effects of volatile anesthetics versus ketamine on blood-brain barrier permeability via lipid-mediated alterations of endothelial cell membranes. J Pharmacol Exp Ther 2023; 385: 135–145.

55.

Spieth

Berghoff

Stumpf

, et al. Anesthesia triggers drug delivery to experimental glioma in mice by hijacking caveolar transport. Neurooncol Adv 2021; 3: vdab140.

56.

Flynn

Buljubasic

Bosnjak

, et al. Cerebral vascular responses to anesthetics. Adv Exp Med Biol 1991; 301: 237–246.

57.

Flynn

Buljubasic

Bosnjak

, et al. Isoflurane produces endothelium-independent relaxation in canine middle cerebral arteries. Anesthesiology 1992; 76: 461–467.

58.

Van Aken

Van Hemelrijck

Influence of anesthesia on cerebral blood flow and cerebral metabolism: an overview. Agressologie 1991; 32: 303–306.

59.

Takuwa

Matsuura

Obata

, et al. Hemodynamic changes during somatosensory stimulation in awake and isoflurane-anesthetized mice measured by laser-Doppler flowmetry. Brain Res 2012; 1472: 107–112.

60.

Ronzani

Woyke

Mair

, et al. The effect of desflurane, isoflurane and sevoflurane on the hemoglobin oxygen dissociation curve in human blood samples. Sci Rep 2022; 12: 13633.

61.

Aksenov

Miller

, et al. Effects of anesthesia on BOLD signal and neuronal activity in the somatosensory cortex. J Cereb Blood Flow Metab 2015; 35: 1819–1826.

62.

Pisauro

Dhruv

Carandini

, et al. Fast hemodynamic responses in the visual cortex of the awake mouse. J Neurosci 2013; 33: 18343–18351.

63.

Liu

Lauer

Ward

, et al. Differential effects of deep sedation with propofol on the specific and nonspecific thalamocortical systems: a functional magnetic resonance imaging study. Anesthesiology 2013; 118: 59–69.

64.

Gluck

Ferrari

Binini

, et al. Distinct signatures of calcium activity in brain mural cells. Elife 2021; 10: 10.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.42 MB

0.00 MB