The Possible Role of CO(2) in Producing A Post-Stimulus CBF and BOLD Undershoot

Abstract

Comprehending the underlying mechanisms of neurovascular coupling is important for understanding the pathogenesis of neurodegenerative diseases related to uncoupling. Moreover, it elucidates the casual relation between the neural signaling and the hemodynamic responses measured with various imaging modalities such as functional magnetic resonance imaging (fMRI). There are mainly two hypotheses concerning this mechanism: a metabolic hypothesis and a neurogenic hypothesis. We have modified recent models of neurovascular coupling adding the effects of both NO (nitric oxide) kinetics, which is a well-known neurogenic vasodilator, and CO(2) kinetics as a metabolic vasodilator. We have also added the Hodgkin-Huxley equations relating the membrane potentials to sodium influx through the membrane. Our results show that the dominant factor in the hemodynamic response is NO, however CO(2) is important in producing a brief post-stimulus undershoot in the blood flow response that in turn modifies the fMRI blood oxygenation level-dependent post-stimulus undershoot. Our results suggest that increased cerebral blood flow during stimulation causes CO(2) washout which then results in a post-stimulus hypocapnia induced vasoconstrictive effect.

Keywords: CMRO2; carbon dioxide; cerebral blood flow; fMRI; nitric oxide; vasodilation.

Figure 1

A Schematic representation of the proposed model. ADP, adenosine diphosphate; AP, action potential; ATP, adenosine triphosphate; BOLD, blood oxygenation level-dependent; Ca²⁺, calcium; CBF, cerebral blood flow; CBV, cerebral blood volume; CO₂, carbon dioxide; dHb, deoxyhemoglobin; GAP, glyceraldehyde-3-phosphate; GLC, glucose; K, potassium; LAC, lactate; Na, sodium; NAD⁺, nicotinamide adenine dinucleotide; NADH, reduced nicotinamide adenine dinucleotide; NO, nitric oxide; PEP, phosphoenolpyruvate; PYR, pyruvate.

Figure 2

The input function of the model: a series of action potentials at 150 Hz (A), only a portion is shown as an example. The Ca²⁺ and Na⁺ concentration as a response to the stimulus. Results are given (B) and (D) for a 20-s stimulus and (C) and (E) for a 100-s stimulus.

Figure 3

Our model results for an evoked change in CBF considering the vasoactive role of only NO (dashed line), only CO₂ (dotted line), both NO and CO₂ (solid line). The relative change in CMRO₂ is indicated by the blue line. Results are given (A) and (B) for a 20-s stimulus and (C) and (D) for a 100-s stimulus. In each case we considered the stimulus as a train of action potentials at a repetition frequency of 150 Hz as input to our model. Note that in (A) we increased the scale of the only CO₂ effect (dotted line) by ×20. CBF, cerebral blood flow; CMRO₂, cerebral metabolic rate of oxygen.

Figure 4

Our model results for an evoked change in BOLD considering the vasoactive role of only NO (dashed line), only CO₂ (dotted line), and both NO and CO₂ (solid line). Results are given (A) and (B) for a 20-s stimulus and (C) and (D) for a 100-s stimulus. In each case we considered the stimulus as a train of action potentials at a repetition frequency of 150 Hz as input to our model. BOLD, blood oxygenation level-dependent.

Figure 5

Change in NO levels in smooth muscle (A,B) and the partial pressure of CO₂ in the precapillary arteriole (C,D) are plotted considering the flow response from NO alone (dashed line), CO₂ alone (dotted line), and both NO and CO₂ (solid line) for the 20-s stimulus (A,C) and the 100-s stimulus (B,D). pCO₂, partial pressure of CO₂.

Figure 6

Modeled (A) CBF and (B) BOLD versus the partial pressure of CO₂ compared with experimental data from (Kety and Schmidt, 1948) circles, (Grubb et al. , 1974) squares, and (Hoge et al., 1999b) diamonds. Our model results are indicated by the solid line. CBF, cerebral blood flow; BOLD, blood oxygenation level-dependent.

Figure 7

Peak CBF undershoot versus stimulus duration (A). BOLD undershoot versus stimulus duration (B). Experimental data (circles), model results (solid line), model results when the model is forced to have a faster CMRO₂ increase following stimulus onset and recovery post-stimulus (dotted-line). CBF, cerebral blood flow.