Hyperoxia evokes pericyte-mediated capillary constriction

Abstract

Oxygen supplementation is regularly prescribed to patients to treat or prevent hypoxia. However, excess oxygenation can lead to reduced cerebral blood flow (CBF) in healthy subjects and worsen the neurological outcome of critically ill patients. Most studies on the vascular effects of hyperoxia focus on arteries but there is no research on the effects on cerebral capillary pericytes, which are major regulators of CBF. Here, we used bright-field imaging of cerebral capillaries and modeling of CBF to show that hyperoxia (95% superfused O₂) led to an increase in intracellular calcium level in pericytes and a significant capillary constriction, sufficient to cause an estimated 25% decrease in CBF. Although hyperoxia is reported to cause vascular smooth muscle cell contraction via generation of reactive oxygen species (ROS), endothelin-1 and 20-HETE, we found that increased cytosolic and mitochondrial ROS levels and endothelin release were not involved in the pericyte-mediated capillary constriction. However, a 20-HETE synthesis blocker greatly reduced the hyperoxia-evoked capillary constriction. Our findings establish pericytes as regulators of CBF in hyperoxia and 20-HETE synthesis as an oxygen sensor in CBF regulation. The results also provide a mechanism by which clinically administered oxygen can lead to a worse neurological outcome.

Keywords: 20-HETE; Hyperoxia; cerebral blood flow; pericyte; reactive oxygen species.

Figure 1.

Hyperoxia-evoked capillary constriction in rat cerebral cortical tissue. (a) Representative bright-field images of a rat cortical capillary, with a pericyte indicated by the yellow arrow, in the normoxic condition (left) and 30 min after application of hyperoxic aCSF oxygenated with 95% O₂ (right), showing capillary constriction evoked by hyperoxia. The white lines in both images display the vessel diameter. (b) Time course of capillary diameter on applying aCSF oxygenated with 95% O₂ (n = 20) or aCSF oxygenated with 20% O₂ (also present in the baseline period, n = 5) for 30 min. Dotted line indicates the time when application of hyperoxic solution started. (c) Constriction in (b) at t = 28–30 min, showing a significant capillary constriction in the presence of hyperoxia in rat cortex. (d–e) As in (c) but in the presence of 2 µM TTX (d, n = 3 in 20% and n = 6 in 95% O₂) or 10 µM DNQX (e, n = 5 in 20% and n = 4 in 95% O₂). (f) Mean constriction as a function of distance along the capillary from 20 pericyte somata, on changing from 20% to 95% O₂ in the superfusate and (g) Mean constriction from (f) after 30 mins of 95% O₂ (dots show individual measurements).

Figure 2.

Hyperoxia-evoked capillary constriction in human cerebral cortical tissue. (a) Bright-field images of a human cortical capillary with a pericyte (yellow arrow) in the normoxic condition (left) and 60 min after application of hyperoxic aCSF oxygenated with 95% O₂ (right), showing capillary constriction. The white lines represent vessel diameter. (b) Capillary constriction at pericyte locations in human cortical tissue was found after application of hyperoxic aCSF (95% O₂, n = 5) but not with normoxic aCSF (20% O₂, also present in the baseline period, n = 5). Dotted line indicates the time when application of hyperoxia started. (c) Constriction in (b) at t = 58-60 min, showing that hyperoxia evoked a significant capillary constriction in human cerebral cortex.

Figure 3.

Hyperoxia-evoked [Ca²⁺]_i increase in mouse cerebral cortex pericytes. (a and b) Fluorescent images of a cortical pericyte (yellow arrows) from NG2-CreERT2 x PC::G5-tdT mice, which express TdTomato (red) and the calcium indicator GCaMP5G (green) in pericytes, in the normoxic condition (a) and 4 min after application of hyperoxic solution (b). Superimposed squares show regions of interest (ROIs) over the pericyte and over a background (BG) part of the image. There was a rise in GCaMP5G fluorescence in the pericyte (but not in the background) after superfusing hyperoxic aCSF, indicating that [Ca²⁺]_i is increased. (c) Time course of the increase of GCaMP5G fluorescence (ΔF/F) on application of aCSF oxygenated with 95% O₂ (n = 8) or aCSF oxygenated with 20% O₂ (also used for the baseline period, n = 16) for 20 min, showing a peak increase of [Ca²⁺]_i at 4–6 min followed by a decrease in the maintained presence of high [O₂]. Background ROIs showed no significant change. (d) ΔF/F in (C) averaged over t = 4-6 min, showing a significant increase in GCaMP5G fluorescence in pericytes after application of aCSF oxygenated with 95% O₂, with no significant change in background ROIs.

Figure 4.

Hyperoxia induces cytosolic and mitochondrial ROS production but ROS and endothelin do not mediate hyperoxic capillary constriction. (a) Dihydroethidium (DHE) staining of rat cortical slices incubated in aCSF with 20% O₂ (left) and with 95% O₂ (right), to visualise cytosolic superoxide production. Cytosolic reactive oxygen species (ROS) production was increased when the slices were made hyperoxic. (b) Average intensity of DHE fluorescence from rat cortical slices incubated in aCSF oxygenated with 20% O₂ or with 95% O₂, showing a significantly higher level of fluorescence in hyperoxic slices (n = 18 for both conditions). (c) MitoSOXstaining for mitochondrial ROS production in rat brain slices in normoxic and hyperoxic conditions, showing an increase in mitochondrial ROS production when the slices were in hyperoxic solution. (d) Average intensity of MitoSOX staining of rat cortical slices incubated in aCSF oxygenated with 20% O₂ (normoxic) or incubated in aCSF oxygenated with 95% O₂ (hyperoxic), showing that mitochondrial ROS production was enhanced in hyperoxia (n = 18 for both conditions). (e) Hyperoxic aCSF (n = 20, red trace) induced capillary constriction occurred with a similar time course in the presence of the NOX4 inhibitor GKT137831 (GKT, 0.45 µM, n = 5, blue trace), the nonspecific NOX blocker DPI (10 µM, n = 5, green trace) or the mitochondrially-targeted antioxidant MitoQ (500 nM, n = 11, black trace). Inhibitors were applied for 15 min before application of 95% O₂ (dotted line) and diameters were normalised to the pre-hyperoxia baseline. (f) Constriction in (e) averaged over t = 28–30 min, showing that hyperoxic vasoconstriction was not inhibited by GKT137831, DPI or MitoQ. (g) Time course of capillary diameter on applying aCSF oxygenated with 95% O₂ (n = 20, red trace) alone, or in the presence of the nitric oxide synthase inhibitor N_ω-nitro-L-arginine (L-NNA, 100 µM, n = 9, blue trace) or the endothelin-1 type A receptor (ET_A) inhibitor BQ-123 (1 µM, n = 9, green trace). Diameters were normalised to the pre-hyperoxic phase (before the dotted line). (h) Constriction in (g) averaged over t = 28-30 min, showing that the hyperoxia-evoked capillary constriction was not significantly blocked by L-NNA or BQ-123.

Figure 5.

Hyperoxia-induced vasoconstriction is partly mediated by 20-HETE. (a) Hyperoxic vasoconstriction in rat cortical slices (n = 20) was partly inhibited by application of HET0016 (100 nM, n = 8), an inhibitor of the synthesis of 20-hydroxyeicosatetraenoic acid (20-HETE) which did not affect the baseline capillary diameter. Diameters were normalised to the pre-hyperoxic phase (before the dotted line). (b) Constriction in (a) averaged over t = 28-30 min, showing that application of HET0016 partially blocked hyperoxia-evoked capillary constriction in rat cortical slices. (c) Time course of capillary diameter on applying aCSF oxygenated with 20% O₂ (n = 5) alone, or in the presence of the adenosine A1 receptor inhibitor 8-cyclopentyl-1,3-dipropylxanthine (DPCPX, 500 nM, n = 6, green trace) or the adenosine A2a receptor inhibitor ZM241385 (1 μM, n = 5, orange trace). Dotted line indicates the time when application of drugs started. (d) Constriction in (c) averaged over t = 13-15 min, showing that the application of DPCPX and ZM241385 did not have a significant effect on capillary diameter.