Microcirculatory response to functional brain stimulation

Background: On the mesoscopic scale accessible by hemodynamically weighted modalities such as PET, fMRI, and OIS, the vascular response to brain stimulation is spatially restricted and temporally locked to neuronal activity under a wide range of conditions. However, little is known about the hemodynamic changes on the microscopic scale. Prior studies (Kleinfeld,1998; Schulte,2003; Chaigneau,2003) suggest a highly heterogeneous system, but suffer from very sparse sampling and use of a wide range of functional paradigms. This work aims to extend earlier investigations by employing a well-described functional paradigm and punctilious analysis to allow direct quantification of vessel volumes. Methods: Ten male SD rats were intubated, ventilated, and anesthetized with ?-chloralose. Electrical pulses (0.3-ms, 2-mA, at 3Hz) were delivered to the forepaw. Two-photon microscopy was performed in a closed cranial window over the S1FL using a Ti:S fs pulsed laser, with 805nm excitation. The passage of 150?300μL iv. bolus of rhodamine-labeled dextran (5mg/ml in PBS) was tracked using a single 1024x1024-μm2 slice, with a 8.0x8.0μm resolution, at 225ms/frame. These data were segmented into extra- and intra-vascular spaces, and a gamma variate fitted to each intravascular voxel's time course via NLS optimization to estimate transit times. Sixty high resolution slices (1-2μm in-plane, 1.5-3μm through-plane) were acquired during rest and activation to estimate steady-state intravascular volume changes. Exponential correction for depth intensity attenuation was performed on these stacks and followed by semi-automated segmentation based on region growing from manually defined seeds. Capillary line scans (1.7ms/line) were done to estimate RBC speeds. Results: The transit time across the imaged vasculature decreased by 20±8%: from 1.6±0.3s at rest to 1.2±0.2s during stimulation. Figure 1 displays representative transit time maps. The mean CBV increased by 10.2±0.6%. A result of semi-automatic segmentation of vasculature is shown in Figure 2. Smaller vessels dilated by 11.4±0.7% on average, a figure that was significantly (p<10-6) larger than the average 7.8±1.1% dilation of medium size vessels. Figure 3 shows histograms of ΔCBV in small and medium caliber vessels. No mean change in the capillary RBC speed was observed following stimulation (ΔvRBC∼-8.1±8.7%). Conclusion: The findings support very heterogeneous microcirculatory changes following stimulation. Average decrease of 0.5±0.1s (20±8%) in the transit time provides the all-important validation of the ROI. For the first time, a direct measurement of individual vessel volumes, rather than point estimation of the vessel diameter, was afforded by the present study. Larger dilatation of the smaller vessels underlay the mean stimulation-induced increase in blood volume of 10.2±0.6%, suggesting increased vascular homogeneity at activation relative to rest.