Introduction:
The human brain produces a staggering 75 litres of CO2/day. We recently published data demonstrating CO2 mediates neurovascular coupling, affecting cerebral blood flow (CBF) (Hosford et al., 2022). A meta-analysis published by our group suggested that nitric oxide and cyclooxygenase (COX) are key modulators of this response (Hosford et al., 2022). However, the mechanisms behind this phenomenon and the implications for cerebrovascular disease remain incompletely understood.
Aims
In this study, we aim to characterise the molecular mechanism behind hypercapnia induced vasodilation of the cerebrovasculature. We also aim to characterise this effect on cerebrovascular reactivity in the hyperhomocysteinemia (HHcy) mouse model of vascular dementia.
Methods
Vessel diameter changes in cortical arterioles were recorded from ex vivo rat brain slices, pre-constricted with L-phenylephrine (200µM) in artificial cerebral-spinal-fluid (aCSF). Isohydric hypercapnia (pH 7.35) was induced by bubbling aCSF with 13.5% CO2 containing 80mM sodium bicarbonate. Ex vivo and in vivo dye loading experiments were carried out on brain slices and open cranial windows on 10-week-old male C57BL6 mice using aCSF containing 200µM carboxyfluorescein. Hypercapnia in in vivo studies was delivered by free-breathing of 10% CO2. The diet-induced HHcy model of vascular dementia, known to develop cognitive impairment and cerebral microbleeds (Sudduth et al., 2013), was carried out on C57BL6 mice of both sexes (n=20). CBF and cerebrovascular reactivity to hypercapnia were quantified using arterial-spin-labelling MRI (9.4T Bruker) under medetomidine/isoflurane anaesthesia.
Results
Recording of cortical arteriolar diameter changes in these slices confirmed CO2 induced vasodilation independent of CO2 induced acidification (p<0.05, n=9, t-test). Vasodilation was prevented by COX-1 inhibition (p<0.001, n=6, one-way-ANOVA) and was affected by neuronal nitric oxide synthase (NOS) inhibition (p<0.001, n=6, one-way-ANOVA), but not endothelial NOS inhibition (p>0.05, n=6, one-way-ANOVA) or neuronal activity inhibition (TTX) (p>0.05, n=6, one-way-ANOVA). Additionally, we found that hypercapnia-induced dilation depends on purinergic signalling (n=6) and is modulated by connexin/pannexin channel availability (p<0.01, n=6, one-way-ANOVA). We confirmed the role of these channels through loading of carboxyfluorescein, a connexin/pannexin channel permeable dye. Concentration dependent CO2 ex vivo dye loading in brain slices was blocked by connexin channel inhibitors. This was confirmed in vivo in the mouse somatosensory cortex, using hypercapnia (p<0.01, n=5, one-way-ANOVA), and CO2 release induced by forepaw stimulation (p<0.01, n=5, one-way-ANOVA).
In the HHcy model of vascular dementia, no differences in basal CBF or cognitive impairment were detected after 10 weeks on the HHcy inducing diet. However, cerebrovascular reactivity to hypercapnia was significantly reduced in mice on the HHcy diet compared to those on the control diet (p<0.05, n=20, t-test). This is suggestive of early disrupted cerebrovascular reactivity, a translational biomarker which could precede cognitive deficits and changes to basal perfusion.
Conclusion
In conclusion, our studies indicate that hypercapnia-induced cerebrovascular dilation is dependent upon neuronal nitric oxide release, purinergic signalling, and CO2 induced connexin channel availability. Further studies are underway to understand the mechanisms driving the impairment of cerebrovascular reactivity to hypercapnia in the HHcy mouse model, highlighting its potential to be used as a much-needed early biomarker of vascular dementia.