135. Cholinergic modulation of neurovascular coupling in the rat somatosensory cortex
C. Sandoe, P. Fernandes, X. Tong, C. Lecrux and E. Hamel
Department of Neurology and Neurosurgery, Montreal Neurological Institute, McGill University, Montreal, QC, Canada
Background and objectives: Functional hyperemia, or neurovascular coupling, is the highly coordinated increase in cerebral blood flow (CBF) seen in areas of increased neuronal activity. Evidence suggests that the perfusion response is driven by the afferents to the target areas and, more importantly, by their local processing. The exact mechanisms by which this occurs, however, as well as the cell types and messengers involved, are still poorly understood. The cholinergic system is involved in many brain functions from memory to attention and arousal, and increasing acetylcholine (ACh) levels through basal forebrain (BF) stimulation leads to cortical activation. Particularly, increased cortical ACh broadens somatosensory cortex receptive fields and uncovers hidden ones (Lamour et al, J Neurophysiol 1988;60:725–50), while reduced ACh levels after BF lesion decreases sensory stimulus-dependent plasticity and whisker stimulation-evoked functional activity (Juliano et al, PNAS 1991;88:780–4; Jacobs et al, Brain Res 1991; 560:342–5), suggesting that cholinergic tone modulates thalamocortical sensory information processing. Here, we further investigated this role using CBF as an index of neurovascular coupling and c-Fos immunohistochemistry to identify the cellular basis of the modulation.
Methods: Endogenous brain ACh levels were increased with the ACh-enhancing drug linopirdine (10 mg/kg, intraperitoneal), or decreased via intracerebroventricular injection of the selective cholinotoxin saporin (4 μg/2 μL). In all experiments, the whisker-to-barrel cortex pathway was used as a model of sensory input-induced CBF increase. Whiskers were mechanically stimulated and CBF was monitored by Laser Doppler flowmetry in urethane-anesthetized rats. Brains were then perfused and stained for neuronal activation using single (c-Fos alone) or double immunohistochemistry for c-Fos and markers of pyramidal cells or different GABA interneuron subtypes.
Results: Saporin treatment decreased cholinergic afferents to the cortex and decreased the CBF response to whisker stimulation (−27.8%±1.7%, P<0.001, n = 18); this was accompanied by a decrease in stimulus-induced neuronal activation in the barrel cortex as detected with c-Fos immunostaining. In contrast, ACh increase via linopirdine treatment increased the CBF response to whisker stimulation (+30.8%±1.9%, P<0.001, n = 6) without clearly altering the extent of c-Fos activation in the barrel cortex. Instead, a shift in the types of interneurons activated in specific layers of the somatosensory cortex seemed to occur.
Conclusions: Our results clearly support a role for ACh in the modulation of sensory input to the somatosensory cortex, showing that enhancing or decreasing cortical ACh neurotransmission had a direct impact on the neurovascular coupling response to thalamocortical afferent activation. Further, they suggest that this modulation may be due, in part, to a shift in the neuronal networks activated by the incoming stimuli, although a direct effect of ACh on neuronal electrical activity cannot be ruled out. These studies provide important insight into the modulation of sensory information processing by ACh. In addition, they may be of relevance to neurodegenerative conditions such as Alzheimer's disease, where a cholinergic deficit is observed, selected groups of interneurons are affected and neurovascular coupling is impaired.
Sponsored by CIHR, MOP-84209 (EH); Jeanne Timmins studentship (CS); and Heart and Stroke Foundation of Canada/Canadian Stroke Network postdoctoral fellowship (CL).
909. In vivo pial arteriolar regulation associated with ATP hydrolysis in the rat
F. Vetri, H.L. Xu, L. Mao and D.A. Pelligrino
Neuroanesthesia Research Laboratory, University of Illinois at Chicago, Chicago, Illinois, USA
Objectives: ATP and adenosine (ADO) have been shown to have important roles in intercellular communication both in physiological and pathophysiological conditions. Furthermore, ATP has been linked to neural activity-induced vasodilation. Whether that vascular response relates to ATP itself, via interactions with purinergic P2 receptors, or is more closely linked to products of ATP hydrolysis, like ADO (and interactions with purinergic P1 receptors), is not precisely known. In the present study, we tested the hypothesis that extracellular ATP elicits vasodilation primarily via its rapid conversion to ADO.
Methods: We assessed vascular effects of exogenous ATP or ADO, topically applied through a closed cranial window, by measuring pial arterial dilation in the absence and presence of the ecto-5′-nucleotidase (e-N) inhibitor α,β-methylene ADP (AOPCP) or the selective ADO A2A receptor blocker, ZM241385. ATP induced a dose-dependent dilation of pial arterioles (14% and 27% at 1 and 10 μmol/L, respectively). Those dilations could be reduced to 2% and 10% in the presence of ZM241385 (10 μmol/L). When co-applied with AOPCP (300 μmol/L), ATP (1 and 10 μmol/L) elicited only 3% and 13% dilations, respectively. On the other hand, ADO elicited 15% and 32% pial arteriolar dilation at 10 and 100 μmol/L, but, when suffused with ZM241385, the dilation was completely blocked. Conversely, as expected, AOPCP did not interfere with ADO-induced pial arteriolar dilation.
Discussion: In the present study, we showed that ATP acts to dilate pial arterioles mainly in the form of ADO. Upon the conversion of ATP to AMP, via ecto-nucleotide pyrophosphatase phosphodiesterase (e-NPP), e-N removes a Pi to produce ADO, which, in turn, dilates vessels through ADO A2A receptors. This cascade can be effectively blocked either by inhibiting e-N with AOPCP or the ADO A2A receptor with ZM241385. The specificity of ZM241385 was proven by its ability to completely block ADO dilatory responses. Similarly, the specificity of AOPCP was demonstrated by the lack of any effect on ADO-induced dilation. As a further indication of the primacy of the ecto-pyrophosphatase (i.e., ecto-diphosphohydrolase; apyrase) pathway, over ecto-monophosphohydrolase activity (i.e., ATP → ADP → AMP), we previously reported that ADP-induced pial arteriolar dilation was unaffected by ADO receptor blockade with 8-sulfophenyltheophylline.1 This indicates that comparatively little ADP is formed from ATP extracellularly.
Conclusions: These data suggest that ADO is the major mediator of ATP-induced pial arteriolar dilation in vivo in rats.
500. Excessive oxygen supply does not change cerebral blood flow responses to increased neuronal activity
U. Lindauer1,2, C. Leithner2, H. Kaasch2, B. Rohrer2, G. Royl2, N. Offenhauser2, M. Füchtemeier2, M. Kohl-Bareis3 and U. Dirnagl2
1Department of Neurosurgery, Technical University Munich, Munich; 2Experimental Neurology, Charité—University Medicine Berlin, Berlin; 3Department of Mathematics and Technology, University of Applied Sciences Koblenz, Remagen, Germany
Introduction: Neuronal activation leads to an increase in regional cerebral blood flow (rCBF) and blood oxygenation (rCBO). Several mechanisms of oxygen metabolism induced delivery of vasoactive substances have been discussed to couple neuronal activity and metabolism to blood flow in the brain. It was suggested that during cellular activity oxygen consumption locally increases which causes deoxygenation of hemoglobin and release of nitric oxide (NO) (a) from deoxygenised hemoglobin (Stamler et al, 1997), (b) due to nitrite reduction by deoxyhemoglobin (Gladwin et al, 2000), or (c) due to activation of purinergic P2y-receptors at vascular endothelial cells during deoxygenation-induced ATP release from erythrocytes (Dietrich et al, 2000). According to these elegant models of combined oxygen and vasodilator release, blood flow would be directly matched to tissue oxygen demands.
Methods: We measured rCBF and rCBO responses to increased neuronal activity induced by functional stimulation or cortical spreading depression (CSD) under hyperbaric hyperoxygenation (3 or 4 ATA, FiO2 1.0) in the anesthetized rat. During hyperbaric hyperoxygenation oxygen supply to tissue is entirely provided through physically dissolved oxygen and no deoxygenation of hemoglobin occurs. Laser Doppler flowmetry combined with optical microfiber-spectroscopy was performed using a cranial window preparation (dura mater intact) in rats to measure relative changes in rCBF (functional activation, CSD) and rCBO (functional activation). Averaged rCBF and rCBO responses to electrical forepaw stimulation (3 Hz, 16 s stimulation period; 3ATA: n = 8; 4ATA: n = 5) or CSD (only LDF, elicited by 160 to 180 mmol/L potassium at a small remote burr hole; 3ATA:n = 6; 4ATA: n = 5) were recorded under normobaric normoxia and compared with responses during hyperbaric hyperoxygenation at 3 or 4 ATA. Neuronal activity was measured by recording somatosensory evoked potentials (functional activation study) or direct current potentials (CSD study).
Results: Hyperbaric hyperoxygenation increased arterial pO2 to >2000 mm Hg and cerebral hemoglobin saturation within the microcirculation to 100%. Resting blood flow as well as functional activation- or CSD-induced neuronal activity did not change compared to baseline conditions at normobaric normoxia. During forepaw stimulation, the deoxy-Hb decrease normally occurring during functional activation disappeared. The rCBF responses to functional activation as well as to CSD did not change under hyperbaric hyperoxygenation compared with control responses under baseline conditions.
Discussion: Our results suggest that rCBF regulation during physiological neuronal activation as well as during the much stronger metabolic stimulus of CSD involves other mechanisms than oxygen metabolism-dependent delivery and release of vasoactive mediators during deoxygenation of hemoglobin.
798. Cortical and sub-cortical activations in rat brain by fMRI
B.G. Sanganahalli1,2, P. Herman1,2,3, D.L. Rothman1,4 and F. Hyder1,2,4
1Diagnostic Radiology, Yale University School of Medicine, New Haven; 2QNMR, Yale University School of Medicine, New Haven, Connecticut, USA; 3Semmelweis University, Budapest, Hungary; 4Biomedical Engineering, Yale University School of Medicine, New Haven, Connecticut, USA
Objectives: Current understanding about BOLD signal and the underlying neurophysiology is based predominantly on functions of the cerebral cortex. BOLD activations of subcortical regions, in contrast, are hard to detect because of low sensitivity and/or difficult access. The goal of the present work is to study subcortical mechanisms underlying dispersed cortical activations during sensory stimulation in rat brain by fMRI. Our results demonstrate reproducible thalamus and superior colliculus activity during forepaw, whisker, and visual stimuli in anesthetized rats. These experiments should provide insights into understudied interactions between cortical/subcortical areas and provide a mechanistic basis to understand multisensory integration.
Methods: Sprague-Dawley rats were tracheotomized and artificially ventilated (70% N2O, 30% O2). During the animal preparation isoflorane (2% to 3%) used for induction. Intraperitoneal lines were inserted for administration of α-chloralose (46±4 mg/kg per h) and D-tubocurarine chloride (1 mg/kg per h). Forepaw stimulation (2 mA, 0.3 ms, 3 Hz): Stimulation was achieved by insertion of thin needle copper electrodes under the skin of the forepaw. Whiskers stimulation (8 Hz): Contralateral whiskers were trimmed to a length of ∼14 mm. Air puffs were used to stimulate the whiskers. Visual Stimulus delivery (8 Hz, blue light): Fibre optic cables (Φ: 1 mm) were used to guide the light of two strong LEDs, placed outside the scanner room, into the eyes of the animal as it lay positioned in the imaging bore. The details of the forepaw, whisker and visual stimulation procedure are available in our previous papers. All fMRI data were obtained on a modified 11.7T Bruker horizontal-bore spectrometer (Billerica, MA) using a 1H resonator/surface coil RF probe.
Results: Independent stimulation of forepaw or whisker or visual activated the contralateral S1FL or S1BF or V1 regions. Our results demonstrate reproducible thalamus and superior colliculus activity during forepaw, whisker, and visual stimuli in rats. Forepaw stimulation activates the medial portions of the laterodorsal (LD) thalamic nucleus. Whisker stimulation activates broader regions within the thalamus: a small caudal part of the lateral thalamic nucleus, the dorsal and medial parts of the lateral geniculate nucleus, and small portions of the dentate gyrus. Visual stimulation activates superior colliculus and lateral geniculate nucleus quite robustly and even parts of the periaqueductal gray. The stimulation frequencies used for forepaw, whisker and visual stimuli were 3, 8 and 8 Hz respectively. Cortical BOLD responses were significantly larger as compared to the thalamic responses during forepaw and whisker stimulation. However, we found no differences in the BOLD response in cortical and superior colliculus during visual stimulation.
Conclusion: We can apply high field fMRI (high S/N ratio) to study thalamo-cortical, colliculo-cortical, thalamo-colliculor as well as their reciprocal interactions with crossmodal sensory mixing. The three regions in the cortex represent the primary areas activated during tactile (somatosensory; S1FL, S1BF) and non-tactile (visual; V1) stimuli, which are connected to the subcortex (thalamus and superior colliculus, respectively). These results have significance in understanding the role of both cortical and subcortical areas during multisensory integration.
Acknowledgements: Supported by grants from NIH (R01 MH-067528, R01 DC-003710, P30 NS-52519).
688. Influence of intracranial pressure on neurovascular coupling: clarifying mechanisms of the bold post-stimulus undershoot
M. Füchtemeier1, N. Offenhauser1, C. Leithner1, J. Steinbrink1, M. Kohl-Bareis2, U. Dirnagl1, U. Lindauer1 and G. Royl1
1Department of Experimental Neurology, Charité Universitätsmedizin Berlin, Berlin; 2University of Applied Sciences Koblenz, RheinAhrCampus Remagen, Remagen, Germany
Objectives: Functional MRI (fMRI) with BOLD (blood oxygen level dependent) localizes activated brain areas non-invasively and is based on local vascular hyper-oxygenation during neuronal activation. However, the hemodynamic mechanisms of the underlying neurovascular coupling are still poorly understood. Therefore, the interpretation of the BOLD-signal is difficult. Which aspects can be attributed to passively occuring vascular effects and which aspects show the actual neuronal activation? Placed in the center of this discussion is the BOLD post-stimulus undershoot, a regularly seen transient hypooxygenation after the end of a stimulation period. Three main mechanisms have been introduced as explanation for the BOLD post-stimulus undershoot:
A neuronal inhibition follows immediately after cessation of the stimulus. Tight neurovascular coupling induces a cerebral blood flow (CBF) decrease leading to a transient relative increase of deoxy-Hemoglobin (deoxy-Hb) (Hoge et al, 1999).
After recovery of the CBF response, transient metabolic processes restore a steady-state equilibrium. These processes sustain an elevated cerebral metabolic rate of oxygen (CMRO2) that increases deoxy-Hb (Lu et al, 2004).
The BOLD post-stimulus undershoot is due to a temporal CBF/cerebral blood volume (CBV) mismatch.
After stimulus cessation, CBV returns to baseline values slower than CBF. The post-arteriolar compartment dilates passively during CBF elevation due to its compliance, similar to a balloon (Buxton et al, 1998) or windkessel (Mandeville et al, 1999). This dilated compartment stores CBV locally. After constriction of the feeding arterioles, this local CBV storage deoxygenates. We addressed these hypotheses by studying the effect of elevated intracranial pressure (ICP) on neurovascular coupling.
Methods: In anesthetized Wistar rats, a plastic catheter was positioned inside the cisterna magna. Over the somatosensory cortex, a closed cranial window was implanted. Using laser Doppler flowmetry and optical spectroscopy, CBF, deoxy-Hb, CBV and CMRO2 were measured, interleaved by blocks of 30 s of electrical forepaw stimulation (monitored by somatosensory evoked potentials). In between stimulation blocks, intracranial pressure was adjusted to four different levels (3.5, 7, 14, and 28 mm Hg) by infusion of artificial cerebro-spinal fluid.
Results: During ICP elevation, the amplitude of the deoxy-Hb post stimulus overshoot (correlate of the BOLD post-stimulus undershoot) was reduced and close to zero. Surprisingly, elevated ICP also reduced deoxy-Hb decrease during activation. At an ICP of 28 mm Hg the deoxy-Hb signal was even reversed. Time courses of CBF and CBV showed a trend towards less mismatch in the post-stimulus period (CBF returning to baseline values faster than CBV). CMRO2 in the post stimulus period was not elevated.
Conclusions: Our data shows that the deoxy-Hb post stimulus overshoot is caused by a passively occuring flow-volume mismatch rather than by metabolic afterload or neuronal inhibition. In addition, elevated ICP impaired, and even reversed the activity-induced deoxy-Hb decrease. Therefore, BOLD-fMRI might not be a reliable brain mapping method in states of ICP elevation.
693. A mathematical model of the neurovascular coupling suggesting both vasodilation and vasoconstriction
Y. Zheng1, Y. Pan1, S. Harris1, S. Billings2, D. Coca2, J. Berwick1, A. Kennerley1, D. Johnston1 and J. Mayhew1
1Psychology; 2Automatic Control and Systems Engineering, University of Sheffield, Sheffield, UK
Objectives: To investigate the coupling between neural activity and cerebral blood flow (CBF), a linear dynamic mathematical model was developed from data obtained using concurrent electrophysiology and laser Doppler flowmetry in rodent barrel cortex.
Methods: Urethane anesthetised Hooded Lister rats were used (200 to 300 g). Electrophysiological recordings were made using a 16-channel electrode probe (NeuroNexusTechnologies) coupled to a data acquisition device (TDT, Florida) with a custom-written Matlab interface. The electrode was inserted normal to cortical surface under to a depth of 1500 μm. The evoked field potentials were used to obtain spatio-temporal estimates of the major current sources and sinks within the cortical layers. Time series of the layer IV current source density (CSD) sink were used as the incoming neural activity. The LDF probe (PeriFlux 5010, Perimed, Stockholm, 780 nm illumination, 0.25 mm separation) was placed as close to the electrode as possible.
Electrical stimulation of the whisker pad was delivered at 5 Hz with intensity 1.2 mA and an individual pulse width of 0.3 ms. Each trial consists of a stimulus conditioning block of variable duration (2, 8, and 16 s) followed by a probing block of 1 s. The interval between the two blocks of stimulation also varies (0.6, 1, 2, 3, 4, 6, and 8 s). Each trial lasted 60 s, and was repeated 10 times per animal. Data were averaged over 10 animals.
Results: We show that the time series of CBF can be represented mathematically as two components: one increases with increased neuronal activity, while the other decreases. The dynamics of the two time series differ, and the summation of the two components provides excellent predictions of the measured CBF time series, shown in Figure 1. The CSD (blue) was normalised with respect to its first pulse, and the changes in CBF (green) was normalised with respect to its mean value during the control condition 4 s prior to stimulus onset. The length of the conditioning block is 16 s, with a resting gap of 8 s between the conditioning and the probing blocks. Significantly, the mathematical model can predict the ‘overshoot and plateau’ characteristics often observed in the haemodynamic responses during the onset period of the stimulation, as well as the delayed return-to-baseline characteristics present in the haemodynamic response.
Conclusions: We hypothesise that the two model components of CBF changes are mediated by arterial volume changes produced by vasodilation and vasoconstriction following changes in stimulation induced neural activity.
The model structure corroborates the finding that neuronal activation evokes both vasodilation and vasoconstriction, and demonstrates the possible involvement of both vasomotor responses in the regulation of CBF.1,2
References
1.
Am J Physiol Heart Circ Physiol2001;281:H2105–H2112.
