Effects of voluntary exercise on structure and function of cortical microvasculature

Abstract

Aerobic activity has been shown highly beneficial to brain health, yet much uncertainty still surrounds the effects of exercise on the functioning of cerebral microvasculature. This study used two-photon fluorescence microscopy to examine cerebral hemodynamic alterations as well as accompanying geometric changes in the cortical microvascular network following five weeks of voluntary exercise in transgenic mice endogenously expressing tdTomato in vascular endothelial cells to allow visualization of microvessels irrespective of their perfusion levels. We found a diminished microvascular response to a hypercapnic challenge (10% FiCO₂) in running mice when compared to that in nonrunning controls despite commensurate increases in transcutaneous CO₂ tension. The flow increase to hypercapnia in runners was 70% lower than that in nonrunners (p = 0.0070) and the runners' arteriolar red blood cell speed changed by only half the amount seen in nonrunners (p = 0.0085). No changes were seen in resting hemodynamics or in the systemic physiological parameters measured. Although a few unperfused new vessels were observed on visual inspection, running did not produce significant morphological differences in the microvascular morphometric parameters, quantified following semiautomated tracking of the microvascular networks. We propose that voluntary running led to increased cortical microvascular efficiency and desensitization to CO₂ elevation.

Keywords: Exercise; brain imaging; cerebral hemodynamics; cranial windows; two-photon microscopy.

Figure 1.

Maximum intensity projection normal to the cortical surface and centered over the primary motor cortex of a Cre x tdTomato mouse with vascular endothelial cells expressing tdTomato (red) and Oregon green dextran distributed within plasma (green) (a, scale 50 µm). The vessel-tracked vascular tree for this data set is seen in (d), scale 50 µm. Insets are of a penetrating vessel from the top (b, scale 20 µm) and the side showing the composite (c, scale 30 µm), green only (e, scale 30 µm), and red only (f, scale 30 µm) channel data.

Figure 2.

Maximum intensity projection of a two-photon fluorescence microscopy z-stack from the surface of the motor cortex (a, scale 50 µm) and side angle (b, scale 100 µm) with red showing vascular endothelial cells expressing tdTomato and green showing intravascular Oregon green. (c) and (d) (scale 10 µm) shows tdTomato labeling in the absence of Oregon green labeling (blue arrows), suggesting a new vessel.

Figure 3.

Samples of a bolus scan time series at a cortical depth of ∼150 µm, showing the bolus arrival first in an artery (a, middle panel; b, red pseudocolor), then in capillaries, and lastly in a vein (a, last panel; b, blue pseudocolor). The mean fluorescence signal change over time in the artery (red) and in the vein (blue) during air (c) and hypercapnia (d) in a nonrunner mouse. The bolus transit is faster in the sample artery than in the sample vein in both conditions; the bolus transit is shortened in each vessel under hypercapnic mixture versus air breathing (mean ± stdev based on signal variation within the vessel). Artery: TTP = 2.28 ± 0.04 s during air, TTP = 1.84 ± 0.05 s during hypercapnia. Vein: TTP = 2.93 ± 0.30 s during air, TTP = 2.35 ± 0.04 s during hypercapnia. The mean signal change over time for an exercise mouse during air (e) and hypercapnia (f); bolus transit time during hypercapnia is shortened for the vein, but minimally affected for the artery. Artery: TTP = 2.00 ± 0.02 s during air, TTP = 1.78 ± 0.06 s during hypercapnia. Vein: TTP = 3.30 ± 0.05 s during air, TTP = 2.42 ± 0.07 s during hypercapnia.

Figure 4.

Scatterplot of the estimates of the time-to-peak intensity of the fluorescent signal over the course of the bolus passage in each scanned vessel during hypercapnic challenge (y-axis) versus that during air breathing (x-axis) (N = 11 runners, N = 11 nonrunners). The slope of the Deming regression to these data was 0.78 ± 0.15 in the runners (green line) and 0.48 ± 0.18 in the nonrunners (orange line). Flow change to hypercapnia was estimated by the inverse of the slope of the regression: flow thus increased in response to 10% CO₂ inhalation by a factor of 1.3 ± 0.2 in runners and by a factor of 2.1 ± 0.8 in nonrunners.

Figure 5.

RBC speeds are measured in a penetrating vessel branch of an artery (a, scale 80 µm, and b, scale 10 µm, with pink line showing location of line scan). Line scan data samples from this vessel during air (c) and hypercapnia (d). RBC speeds thus estimated in this vessel were 1.32 ± 0.02 mm/s during air breathing and 1.86 ± 0.03 mm/s during breathing of 10% CO₂. Steeper slopes indicate smaller RBC speeds. Bar graph (e) showing change in arteriolar RBC speed induced by hypercapnia for nonrunners (N = 5) and runners (N = 4).