Does hypercapnia-induced impairment of cerebral autoregulation affect neurovascular coupling? A functional TCD study

Transcription

1 J ppl Physiol 115: 91 97, 13. First published June 6, 13; doi:1.115/japplphysiol Does hypercapnia-induced impairment of cerebral autoregulation affect neurovascular coupling? functional TCD study Paola Maggio, 1 ngela S. M. Salinet, Ronney. Panerai,,3 and Thompson G. Robinson,3 1 Neurologia Clinica, Università Campus io-medico, Rome, Italy; Department of Cardiovascular Sciences, University of Leicester, Leicester, United Kingdom; and 3 National Institutes for Health Research (NIHR), iomedical Research Unit in Cardiovascular Sciences, Clinical Sciences Wing, Glenfield Hospital, Leicester, United Kingdom Submitted 18 March 13; accepted in final form 3 June 13 Maggio P, Salinet SM, Panerai R, Robinson TG. Does hypercapnia-induced impairment of cerebral autoregulation affect neurovascular coupling? functional TCD study. J ppl Physiol 115: 91 97, 13. First published June 6, 13; doi:1.115/japplphysiol Neurovascular coupling (NVC) and dynamic cerebral autoregulation (dc) are both impaired in the acute phase of ischemic stroke, but their reciprocal interactions are difficult to predict. To clarify these aspects, the present study explored NVC in a healthy volunteer population during a surrogate state of impaired dc induced by hypercapnia. This study aimed to test whether hypercapnia leads to a depression of NVC through an impairment of dc. Continuous recordings of middle cerebral arteries cerebral blood flow velocity (CFv), blood pressure (P), heart rate, and end-tidal CO were performed in 19 right-handed subjects (aged 5 yr) before, during, and after 6 s of a passive paradigm during normocapnia and hypercapnia. The CFv response was broken down into subcomponents describing the relative contributions of P (V P ), critical closing pressure (V CrCP ), and resistance area product (V RP ). V RP reflects myogenic activity in response to P changes, whereas V CrCP is more indicative of metabolic control. The results revealed that hypercapnia significantly affected NVC, with significant reductions in the relative contribution of V CrCP to the paradigminduced increase in CFv. The present study suggests that hypercapnia impairs both dc and NVC, probably acting through an impairment of the metabolic component of CF control. carbon dioxide; cerebral blood flow; dynamic cerebral autoregulation; functional transcranial Doppler; neurovascular coupling NEUROVSCULR COUPLING (NVC) and cerebral autoregulation (C) are two separate mechanisms involved in the control of cerebral blood flow (CF) (6, 7, 9). NVC involves a complex cascade of events following changes in neural activity and is usually assessed by the CF response to stimulation based on sensorimotor or cognitive paradigms. In particular, the introduction of transcranial Doppler (TCD) has facilitated its assessment by noninvasive means with high temporal resolution (35). NVC has been suggested as a prognostic factor for motor recovery after stroke (5) and as a therapeutic target in disorders associated with cerebrovascular dysfunction (1). C is defined as the inherent ability of brain blood vessels to keep CF relatively constant over a wide range of systemic blood pressure (P) changes (3). More recently, the transient response of CF to rapid changes in P has been the preferred method of assessment of what is now known as dynamic C (dc) (38). However, the effects of dc on NVC following neural stimulation are difficult to predict due to several factors influencing CF, such as changes in sympathetic activity, ddress for reprint requests and other correspondence: T. G. Robinson, Dept. of Cardiovascular Sciences, Robert Kilpatrick Clinical Sciences ldg., PO ox 65, Leicester LE 7LX, UK ( tgr@le.ac.uk). breathing, blood gases, and other cardiovascular variables such as heart rate and cardiac output (13, 6, ). In particular, dc is recognized to be impaired in the acute phase of ischaemic stroke (3), but it is unclear whether this is the sole cause of the depressed NVC reported in ischemic events (5, 3). efore attempting to address the interdependence of dc and NVC in cerebrovascular diseases, where the many confounding factors would require very large sample sizes to be investigated, it would be preferable to adopt a model based on healthy subjects where it should be possible to exert greater control. Hypercapnia can be used to emulate a state of impaired dc (1, ), thus enabling the simultaneous assessment of NVC and C in the presence of dc impairment. To our knowledge, the effect of hypercapnia on NVC has never been investigated by means of TCD during a motor activation paradigm. Rosengarten et al. (33) showed a reduction of the hemodynamic response to a visual paradigm during increased CO blood concentrations. However, definitive and general conclusions still cannot be drawn, because regional differences in NVC between motor and visual areas have been demonstrated (6), as well as regional heterogeneity of the cerebral vascular response to hypercapnia (15). Despite many functional magnetic resonance imaging (fmri) or positron emission tomography (PET) studies exploring the effect of hypercapnia on NVC, results are still conflicting (7, 19). limitation of these studies is the absence of concomitant measurements of peripheral hemodynamic changes during stimulation. The potential contribution of arterial P or arterial pco on NVC has been so far largely ignored (7), as well as the role of impaired dc during hypercapnia. In this scenario, TCD, enabling a beat-to-beat assessment of CF velocity (CFv), P, and end-tidal CO response to motor paradigm, as well as a reliable measure of dc, could be applied to provide a more comprehensive assessment of the impact of hypercapnia-induced impairment of dc on NVC. Therefore, this study aimed to test two interrelated hypotheses: 1) hypercapnia leads to a depression of NVC and ) hypercapnia affects NVC through an impairment of dc. To address the first question, the effect of increased CO blood concentrations on beat-to-beat cerebral and systemic hemodynamic responses to a passive motor paradigm was assessed by means of TCD. The second hypothesis was tested by analysis of the changes in instantaneous pressure-flow velocity relationships induced by hypercapnia, as a surrogate of impaired dc (6). MTERILS ND METHODS Subjects and measurements. Healthy volunteers were recruited from departmental staff and their relatives. Inclusion criteria were age of 5 yr and right-handedness according to the Edinburgh inventory /13 Copyright 13 the merican Physiological Society 91

2 9 The Effect of Hypercapnia on Neurovascular Coupling Maggio P et al. (3). Exclusion criteria comprised physical disease in the upper limb, poor insonation of both temporal bone windows, and any history of cardiovascular, neurological, or respiratory disease. The Nottingham Research Ethics Committee 1, United Kingdom (ref.: 11/EM/16), approved the study, and informed, written consent was obtained from all participants. Recordings were performed in a quiet research laboratory with constant ambient temperature of C. Subjects were asked to avoid caffeine, alcohol, and nicotine for 1 h before measurements. Each participant underwent a single recording. eat-to-beat P was recorded continuously using the Finometer device (FMS, Finapres Measurement Systems, rnhem, The Netherlands) attached to the middle finger of the left hand. The servo-correcting mechanism of the Finometer was switched on and then off before measurements. Heart rate interval was recorded using a three-lead electrocardiogram (ECG), and end-tidal CO was measured via nasal prongs (Salter Labs) by a capnograph (Capnocheck Plus). ilateral insonation of the middle cerebral arteries (MCs) was performed using TCD (Viasys Companion III; Viasys Healthcare) with a -MHz probe, which was secured in place using a head-frame. The MCs were identified according to their signal depth and velocity characteristics (). Hypercapnia was induced by the inhalation of a mixture of 5% CO /air through a mask. During CO breathing, the capnograph was connected to the mask via a sample line. During the entire procedure, subjects were in a supine position, and detailed instructions were given before taking measurements. fter a period of 15 min of stabilization, participants performed a 5-min baseline recording and two passive motor paradigms during air breathing. Passive motor stimulation was shown to have better reproducibility than active or imagery motor paradigms (36). The same sequence of measurements (5-min baseline and two passive motor paradigms) was repeated during CO inhalation. rterial P was measured with a sphygmomanometer before each measurement. The passive paradigm consisted of an examiner performing repetitive flexion and extension of the subject s elbow within a range of movement of 9 at a rate of 1 Hz; subjects were instructed to relax and not move the arm. The frequency was given by the sound of a metronome. ll paradigm recordings started with a 9-s baseline phase. Thereafter, the paradigm was performed over 6 s, with a 9-s recovery phase. During the rest and recovery periods, the examiner kept hold of the participant s arm. The paradigm was performed only with the dominant arm. Data analysis. Data were simultaneously recorded onto a data acquisition system (PHYSIDS, Department of Medical Physics, University Hospitals of Leicester) for subsequent offline analysis. ECG, end-tidal CO, P, and stimulus marker signals were sampled at 5 samples/s, and P was calibrated at the start of each recording. ll signals were visually inspected to identify artifacts and noise, and narrow spikes ( 1 ms) were removed by linear interpolation. The channels were subjected to a median filter, and all signals were low-pass filtered with a cut-off frequency of Hz. The R-R interval (i.e., the interval between two consecutive heartbeats recorded by the ECG) was then automatically marked from the ECG, and continuous heart rate was plotted against time. Occasional ectopic beats caused spikes in the heart rate signal; these were manually removed by remarking the R-R intervals for the time points at which they occurred. Mean P and CFv values were calculated for each cardiac cycle. The end of each expiratory phase was detected in the end-tidal CO signal, linearly interpolated, and resampled with each cardiac cycle. The instantaneous relationship between P and CFv was used to estimate critical closing pressure (CrCP) and resistance area product (RP) for each cardiac cycle using the first harmonic method (8). eat-to-beat data were spline interpolated and resampled at 5 samples/s to produce signals with a uniform time base. utoregulation index (RI) was computed on air and during 5% CO baseline measurement after a period of 1 s of stabilization, as previously described (16). In summary, a fast Fourier transform was applied to the data, and the cross- and auto-spectra were estimated using the Welch method. The transfer function of P-CFv was then calculated with P selected as the input and right then left CFv as the output variables. n inverse fast Fourier transform was then applied, converting data back into the time domain, to calculate the CFv step response (16). RI was assigned to each recording by using the best least-squares fit between the CFv step response and one of the 1 model RI curves proposed by Tiecks et al. (38). RI was computed for each subject separately for left and right hemisphere for normocapnia and hypercapnia. Mean values of each variable were extracted from the 3 s before the start of the paradigm for baseline. Using the electrical output from the metronome, coherent averages were calculated for each variable synchronized by the beginning of the paradigm. Temporal paradigmsynchronized population averages of the first and second motor paradigms were compared, and the maneuver that achieved the highest amplitude of contralateral CFv response was chosen to represent the participant s response (36). Decomposition of the CFv changes into its main subcomponents was performed as described previously (6). In summary, the percentage change in CFv was decomposed into standardized subcomponents describing the relative contributions of P (V P), resistance area product (V RP), and critical closing pressure (V CrCP). Therefore, the total change in CFv during activation was represented as the sum of the three subcomponents, reflecting the separate contribution of parallel changes in P, CrCP, and RP. It has been suggested that V RP might reflect myogenic activity in response to P changes, whereas V CrCP is more indicative of metabolic control (5, 6). Statistical analysis. Student s t-test for dependent variables was used to compare baseline values of CFv, P, heart rate, end-tidal CO, CrCP, RP, and RI between normocapnia and hypercapnia. To compare changes in CFv and its subcomponents between normocapnia and hypercapnia, the area-under-the-curve (UC) was calculated for their differences from the beginning of the maneuver up to s after the end of passive arm movement. Differences between ipsiand contralateral values of UC were tested by two-way, repeatedmeasures NOV, also including the effects of hypercapnia. In the absence of significant differences between sides, values for the right and left MC were averaged, and the effects of hypercapnia were tested by paired Student s t-tests. Student s t-test for dependent variables was also used to compare end-tidal CO UC values between the two groups. Multivariate NOV was performed to assess the overall effects of hypercania, taking into account the interactions between CFv, P, CrCP, and RP. ll statistical analyses were performed with Statistica software for Windows. value of P.5 ( ) was adopted to indicate statistical significance. onferroni correction for multiple testing was performed by recalculating the critical value of, taking into account the correlation coefficient between variables. RESULTS Nineteen participants (8 men) of mean age 58.1 yr (SD 7.1, range 8 77) were recruited. Participant s mean Edinburgh Inventory for right-handedness was 96.5% (SD 8.5). ll participants completed measurements under normocapnia. During hypercapnia, all subjects completed the baseline recording and the first motor paradigm. Only 16 participants completed the second paradigm during hypercapnia since three participants did not tolerate continuous use of the face mask. In these three cases, we analyzed the only available passive motor paradigm recording to represent the participant s response. aseline changes. The influence of hypercapnia on cerebral and peripheral hemodynamic variables during baseline recordings at rest is given in Table 1. Contralateral and ipsilateral baseline CFv, P, and end-tidal CO were significantly

3 The Effect of Hypercapnia on Neurovascular Coupling Maggio P et al. 93 Table 1. Mean (SD) values of cerebral and systemic hemodynamic variables for baseline air and CO measurements Variables ir CO *P Value CFv contr, cm/s 56.6 (11.8) 7. (1.8).1 CFv ipsi, cm/s 9. (9.1) 6. (11.).1 P, mmhg 99. (11.1) 19.1 (19.3).1 Heart rate, beats/min 67.1 (8.1) 7.8 (1.6).8 EtCO, Torr.7 (3.3) 7.9 (3.).1 CrCP contr, Torr 1.5 (1.5) 9. (18.7). CrCP ipsi, Torr 19. (13.3) 13. (17.6). RP contr, Torr s/cm 1.5 (.1) 1. (.39).7 RP ipsi, Torr s/cm 1.7 (.7) 1.6 (.8). RI contr 5. (1.). (1.1). RI ipsi 5. (1.).5 (1.3).1 Values are means (SD). CFv, cerebral blood flow velocity; P, mean arterial blood pressure; EtCO, end-tidal CO ; CrCP, critical closing pressure; RP, resistance-area product; RI, autoregulation index; contr, contralateral to the paradigm; ipsi, ipsilateral to the paradigm. *P value of Student s t-test dependent samples. higher during CO inhalation than during air breathing. Contralateral CrCP was significantly lower during hypercapnia than during normocapnia. Ipsilateral CrCP was also lower, although this did not reach statistical significance. Heart rate and RP were not different between air and 5% CO. oth contralateral and ipsilateral autoregulation indexes were significantly impaired by hypercapnia (Table 1). Neurovascular coupling. One subject did not show a CFv response to the paradigm during normocapnia and was excluded from further analysis. Population averaged changes of CFv, P, CrCP, and RP for normocapnia and hypercapnia are displayed in Figs. 1. Figure 5 depicts changes in UC due to hypercapnia for both ipsi- and contralateral sides. No significant differences were observed between values for the right and left MC, and these were averaged for further analyses. In addition to the shifts in baseline values described above, hypercapnia induced very significant effects on the response to passive motor stimulation of the arm (multivariate NOV F 3.; P.6). The increase of CFv induced by the paradigm during normocapnia (UC 5..1%) was significantly lower during hypercapnia (.7.68%; P.13). The paradigm-induced increase in P was reduced by hypercapnia (Figs. and 5), although the UC did not reach statistical significance ( vs %; P.1). P (mmhg) Positive changes in CrCP were observed during both normocapnia and hypercapnia (Fig. 3), but the contribution of V CrCP was significantly reduced during hypercapnia (3.3.97% to %; P.1, corrected.5). On the contrary, the contribution of V RP (Figs. and 5) was not significantly different between normocapnia and hypercapnia (.3 1.% vs %; P.). Figure 6 shows the absolute values (Fig. 6) and the normalized temporal pattern (Fig. 6) of the end-tidal CO population average. The passive exercise appeared to induce greater hyperventilation during normocapnia than during hypercapnia (Fig. 6), but the normalized UC was not significantly different (P.6). DISCUSSION Fig.. Changes in absolute values of blood pressure (P) population averages during passive motor paradigm. Normocapnia (continuous line) vs. hypercapnia (broken line). For clarity, only the largest 1 SE is represented at the point of occurrence. The horizontal black line shows the duration of the maneuver. Neurovascular coupling is a complex mechanism involved in the control of CF and is regulated by myogenic (17), metabolic (1), and neural components (13). Its impairment observed during the acute phase of stroke () could be explained by direct ischemic neuronal injury, reduced oxidative and glucose metabolism (9), and/or compromised C (3). To clarify these aspects, the present study explored NVC in a healthy volunteer population during a surrogate state of impaired dc, as induced by CO inhalation. The subcomponent analysis of the CF response to a passive motor stimulus was also applied to give further insights into different factors that might affect CFv changes during hypercapnia. Main findings. The significant changes in the NVC response to a passive motor paradigm, induced by hypercapnia, con- CFv (cm s -1 ) CFv (cm s -1 ) Fig. 1. Changes in absolute values of cerebral blood flow velocity (CFv) population averages during passive motor paradigm. : left hemisphere (contralateral to passive paradigm) during normocapnia (continuous line) vs. left hemisphere during hypercapnia (broken line). : right hemisphere (ipsilateral to passive paradigm) during normocapnia (continuous line) vs. right hemisphere during hypercapnia (broken line). For clarity, only the largest 1 SE is represented at the point of occurrence. The horizontal black line shows the duration of the maneuver.

4 9 The Effect of Hypercapnia on Neurovascular Coupling Maggio P et al. Fig. 3. Changes in absolute values of critical closing pressure (CrCP) population averages during passive motor paradigm. : left hemisphere (contralateral to passive paradigm) during normocapnia (continuous line) vs. left hemisphere during hypercapnia (broken line). : right hemisphere (ipsilateral to passive paradigm) during normocapnia (continuous line) vs. right hemisphere during hypercapnia (broken line). For clarity, only the largest 1 SE is represented at the point of occurrence. The horizontal black line shows the duration of the maneuver. CrCP (mmhg) CrCP (mmhg) firmed our first hypothesis as shown by the overall multivariate NOV and more specifically by the differences in the CFv response. Despite the absence of significant changes in the V RP response, our second hypothesis, that NVC was affected due to impairment in dynamic C, was also confirmed by the much reduced response in V CrCP. The demonstration that hypercapnia leads to the simultaneous impairment of dc and NVC thus suggests a possible association between what is usually regarded as two different mechanisms. C is a complex homeostatic mechanism influenced by myogenic (, 39), metabolic (3), and neural activity (13). Our results suggest that CO impaired the metabolic component of dc, as indicated by the changes in V CrCP (5, 6, 7, 8). Conversely, the lack of changes in V RP suggests that the myogenic pathway was not affected by hypercapnia. Neurovascular coupling during normocapnia. The CFv pattern response to the passive motor paradigm during normocapnia displayed an initial peak followed by a plateau profile and a delay in return to baseline after the cessation of stimulation, in very good agreement with data previously reported by others (18, 1, 35). Interestingly, the breakdown of the CFv response into its subcomponents revealed that, during normocapnia, the CFv increase is mainly determined by positive contributions of P and CrCP. Moody et al (1) described significant dynamic changes in heart rate induced by word or puzzle paradigm. However, these mental activation tasks are more demanding and possibly require a greater overall sympathetic activation than passive motor activity. Cerebrovascular effects of hypercapnia: baseline condition. Hypercapnia induced a bilateral increase of CFv, a rise in P and end-tidal CO, and a decrease of RI and contralateral CrCP, but did not affect RP. These effects are well known and are in agreement with those reported previously (1, ). The increase in inspiratory CO is followed by an extracellular rise in H ions, a ph reduction, and vessel smooth muscle cell relaxation, which are thought to be the major mechanisms of a CO -mediated increase of CFv. Other mechanisms such as nitric oxide (NO), prostaglandins, or neurogenic components might also be involved in hypercapnic vasodilation (1). Mean P increase is consistent with previous studies (1, 1) and could be explained by an effect of CO on the sympathetic nervous system () or as a CO -positive inotropic effect on the heart (1). Furthermore, changes in CO concentrations affect vasomotor tone, which may in turn affect CFv through a decrease of CrCP (1, 8). The lack of ipsilateral CrCP changes induced by hypercapnia could be attributed to its wide standard deviation and study power. RP did not change during 5% CO, as previously observed (). These results confirm a preliminary hypothesis raised by Panerai et al. (), which postulated that CO may induce a parallel shift of the flow-pressure curve with a reduction of CrCP rather than changes in the slope of the curve, thus with a lesser influence on RP. During motor paradigm. Previous studies demonstrated that hypercapnia reduces the early peak response and induces a slower increase of CFv in rat models during somato-sensorial stimulation (18) or in humans during visual stimulation (33). Our findings confirm these observations in humans using passive motor activation, a substantially reproducible paradigm (36), also feasible in stroke patients (), where motor deficits could make a voluntary activity not possible. Unfortunately, further comparisons between our results and those of Rosen- Fig.. Changes in absolute values of resistance area product (RP) population averages during passive motor paradigm. : left hemisphere (contralateral to passive paradigm) during normocapnia (continuous line) vs. left hemisphere during hypercapnia (broken line). : right hemisphere (ipsilateral to passive paradigm) during normocapnia (continuous line) vs. right hemisphere during hypercapnia (broken line). For clarity, only the largest 1 SE is represented at the point of occurrence. The horizontal black line shows the duration of the maneuver. RP (mmhg s cm -1 ) RP (mmhg s cm -1 )

5 The Effect of Hypercapnia on Neurovascular Coupling Maggio P et al CFv 6 V P 6 V CrCP 6 V RP UC (%) Fig. 5. Changes in mean area under the curve (UC) of the cerebral blood flow velocity (CFv) and the responses of P (V P), CrCP (V CrCP), and RP (V RP) to passive arm flexion at normocapnia (IR) and hypercapnia (CO ) for the ipsilateral (Œ; continuous line) and contralateral ( ; broken line) hemispheres. Error bars are 1 SE. - IR CO - IR CO - IR CO - IR CO garten et al. (33) are limited by methodological and physiological differences. In fact, regional differences in NVC between the motor and visual areas have been demonstrated (6), as well as regional heterogeneity of the cerebral vascular response to hypercapnia (15). Moreover, visual and motor paradigm-induced changes in peripheral variables, including respiratory rate, P, and heart rate, may be dissimilar, thus triggering different CFv changes. fmri or PET studies exploring the effect of hypercapnia on hemodynamic response to neural activation are so far conflicting. Two main hypotheses have emerged, the first stating that local CF changes induced by stimulation are independent of and additive to basal CF (19), and the second affirming that the magnitude of the hemodynamic response to stimulation is attenuated during increased basal CF, as induced by CO inhalation (7). This discrepancy may be explained by the type and the duration of the stimulus used and/or by methodological differences in end-tidal CO increase (steady state vs. transient modulation) and in respiratory pattern control. In agreement with our results, Stefanovic et al. (37) reported that hypercapnia reduces stimulus-evoked OLD signal changes in the human motor cortex, and at high end-tidal CO values an almost complete loss of the functional flow response to stimulation has been demonstrated (31). However, it is important to note that all these neuroimaging studies have largely ignored the role of task-induced P and arterial pco changes as well as the interplay between NVC and dc during hypercapnia; therefore, a comparison between our results and previous literature concerning peripheral changes in response to motor paradigms is not possible. Metabolic effects. The subcomponent analysis of the CFv response to neural activation revealed that hypercapnia significantly affects V CrCP, which mainly represents the metabolic pathway of CF control (5, 6, 8). The molecular mechanisms governing the interaction between neurons and cerebral blood vessels during hypercapnia and during functional hyperemia have been the subject of intense research. widely accepted hypothesis for NVC is that activation of glutamate receptors during synaptic transmission leads to postsynaptic increases in Ca, which in turn activates enzymes that produce vasoactive agents (1). Some of the mediators (including H,K, and NO) implicated in the CF response to neural activation are similarly demonstrated to be involved in CO -induced CF increase (3, 1). ccordingly, it is reasonable to postulate that, since the CFv increase induced by hypercapnia and by neural activation share common molecular pathways, a manipulation of CO blood concentration could interfere with neural or glial regulation of cerebral microvasculature, thus affecting NVC. Moreover, molecular (1) and electrophysiological evidence () support the hypothesis that CO acts by modulating neural cells, in other words that its action is neurogenic in origin. recent study by Kennerley et al. (18) analyzed the effect of hypercapnia on neural and hemodynamic responses during whisker stimuli in the rat. They observed that increased CO EtCO (mmhg) 8 EtCO (%) Fig. 6. Population averages of end-tidal CO (EtCO ) during passive motor paradigm. : absolute changes in EtCO during normocapnia (continuous line) and hypercapnia (broken line). : normalized changes in EtCO during normocapnia (continuous line) and hypercapnia (broken line). For clarity, only the largest 1 SE is represented at the point of occurrence. The horizontal black line shows the duration of the maneuver.

6 96 The Effect of Hypercapnia on Neurovascular Coupling Maggio P et al. concentrations inhibited both the hemodynamic and the neural response to a somato-sensorial stimulus, but subsequent modeling suggested that the difference in neural response might thoroughly explain hemodynamic changes. This result supports our findings, since it suggests that the impairment of CF response to activation may be attributed to a compromised neural (metabolic) ability to drive the hemodynamic shift. Myogenic effects. In our sample, neither the V RP nor the V P component of CFv response to motor paradigm were affected by hypercapnia. The exact mechanisms underlying the contribution of V RP on neurovascular coupling are not yet completely understood. V RP is supposed to reflect myogenic activity (5, 6), even if some evidence suggests an additional metabolic influence (5, 6, 7). One possible explanation for this finding is that the metabolic pathway of CF control is so compromised during hypercapnia that the myogenic influence is subordinate and not detectable by means of TCD measurements. Returning to the main hypotheses of this work, it is possible to accept that hypercapnia leads to a depression of NVC, but more caution needs to be applied to the second hypothesis, that depression of NVC results from dc impairment, due to the lack of significant differences in V RP, which have been associated with the myogenic component of dc. More work in this aspect of NVC impairment in acute stroke is needed, as discussed below. Clinical implications. Given the previous considerations and evidence from the literature (), how can we then explain that acute ischemic stroke patients have impairments of both C and NVC? Since both mechanisms are the result of myogenic, metabolic, and neural control, their impairment during acute ischemic stroke could be attributed to metabolic, myogenic, or neural involvement or, more likely, to their reciprocal interaction. Stroke subtype and volume, ischemic penumbra, small vessel disease, cerebrovascular risk profile, and extra- or intracranial atherosclerosis may have a different impact on metabolic or myogenic impairment of C, as well as on dc influence on NVC. For instance, an impairment of C myogenic control could be postulated in patients suffering from severe hypertension and small vessel disease and experiencing small lacunar stroke. On the other hand, patients with large MC infarction may present both metabolic and myogenic C impairment. The corresponding effect of dc on NVC is a composite, but intriguing, subject. It not only provides benefits for a detailed understanding of the damaging effects of stroke but is also crucial for the development of more specific and targeted rehabilitation therapies. Limitations In our study, we took into account only impairment of C due to blunting of the metabolic pathway, which could be regarded as a limitation to reproduce clinical conditions that also impairs myogenic control, as could be the case with some instances of acute ischemic stroke. Further work needs to be done in this direction and attempt to apply a more refined model of impaired C. The observation that Ca -channel blockade makes the cerebral microcirculation more vulnerable to systemic P fluctuations (39) and that purinergic receptors modulate the myogenic tone in cerebral parenchymal arterioles () may suggest novel strategies to induce a surrogate state of myogenic-impaired C. Therefore, it is worth mentioning that hypercapnia impairs C but also leads to vasodilation. t a high CO level, the pial arteries could be near to a limit of vasodilatation, which cannot be further influenced by neuronal activation induced by the paradigm. Using other surrogates of impaired dc, different from hypercapnia, would help to clarify this issue. Measurements of CFv reflect changes in CF as long as the diameter of the insonated vessel remains constant. Moreover, any variations in MC diameter taking place during activation tasks would also change RP, since it represents cerebrovascular resistance multiplied by MC cross-sectional area (8). However, several studies have demonstrated that the cross-sectional area of the MC changes minimally during large modifications in P or end-tidal CO (11), even if there is still limited evidence during motor paradigms. lthough the interpretation of changes in V CrCP and V RP as manifestations of metabolic and myogenic CFv regulation is currently supported (5 8), it is still speculative, and further experimental work is needed. In effect, it is still not possible to confirm that each parameter (CrCP or RP) is uniquely associated with the metabolic or myogenic control of CFv response to the paradigm, respectively (5, 7). Moreover, CrCP can also be influenced by cerebral venous pressure (8), a variable affected by intrathoracic pressure. We cannot rule out that part of the reduced CFv response to motor paradigm may be attributed to increased sympathetic activation triggered by CO. Pial arteries and large surface arterioles are innervated by nerve fibers that originate in autonomic and trigeminal sensory ganglia (13), and CO has been suggested to modulate sympathetic nervous system activity (). Future work taking into account the neural component of CFv response to motor paradigm would help to clarify this issue. Finally, the lack of biceps and triceps electromyography recordings cannot exclude the absence of voluntary muscular activity during the passive motor paradigm. However, similar to other brain activation studies (35, 36), subjects were closely monitored during the experiment for any active elbow movement. In conclusion, hypercapnia leads to impairment of dc and depression of NVC. nalysis of different subcomponents of the CFv response to neural activation demonstrated that depression of the NVC response was mainly caused by changes in CrCP, thus suggesting greater involvement of metabolic pathways rather than myogenic control of C, which would be expected to be mainly expressed through RP. These findings are pertinent to a better understanding of NVC alterations in pathological conditions such as stroke, where C is often impaired. Further work in this area should consider alternative models of C impairment, such as that induced by calcium channel blockers, to test the generalizability of our findings, as well as further refinements in analytical techniques to improve estimates of the different covariates influencing both C and NVC. CKNOWLEDGMENTS uthors acknowledge Dr. FabrizioVernieri for his contribution. We thank Dr. Nick Masca from the NIHR Cardiovascular iomedical Research Unit, University of Leicester, for expert advice on statistical analysis.

7 The Effect of Hypercapnia on Neurovascular Coupling Maggio P et al. 97 DISCLOSURES No conflicts of interest, financial or otherwise, are declared by the author(s). UTHOR CONTRIUTIONS uthor contributions: P.M. and.s.m.s. performed experiments; P.M.,.S.M.S., and R..P. analyzed data; P.M.,.S.M.S., R..P., and T.G.R. interpreted results of experiments; P.M. prepared figures; P.M. drafted manuscript; P.M.,.S.M.S., R..P., and T.G.R. approved final version of manuscript;.s.m.s., R..P., and T.G.R. conception and design of research;.s.m.s., R..P., and T.G.R. edited and revised manuscript. REFERENCES 1. aslid R, Lindegaard KF, Sorteberg W, Nornes H. Cerebral autoregulation dynamics in humans. Stroke : 5 5, aslid R, Markwalder T, Nornes H. Non-invasive transcranial Doppler ultrasound recording of flow velocity in basal cerebral arteries. J Neurosurg 57: , ries MJ, Elting JW, De Keyser J, Kremer P, Vroomen PC. Cerebral autoregulation in stroke: a review of transcranial Doppler studies. Stroke 1: 697 7, 1.. rayden JE, Li Y, Tavares MJ. Purinergic receptors regulate myogenic tone in cerebral parenchymal arterioles. J Cereb lood Flow Metab 33: 93 99, Calautti C, aron JC. Functional neuroimaging studies of motor recovery after stroke in adults: a review. Stroke 3: , Chiarelli P, ulte DP, Gallichan D, Piechnik SK, Wise R, Jezzard P. Flow-metabolism coupling in human visual, motor, and supplementary motor areas assessed by magnetic resonance imaging. Magn Reson Med 57: , Cohen ER, Ugurbil K, Kim SG. Effect of basal conditions on the magnitude and dynamics of the blood oxygenation level-dependent fmri response. J Cereb lood Flow Metab : 1 153,. 8. Dawson SL, Panerai R, Potter JF. Critical closing pressure explains cerebral hemodynamics during the Valsalva maneuver. J ppl Physiol 86: , Dietrich WD, Ginsberg MD, usto R. Effect of transient cerebral ischemia on metabolic activation of a somatosensory circuit. J Cereb lood Flow Metab 6: 5 13, Garnham J, Panerai R, Naylor R, Evans DH. Cerebrovascular response to dynamic changes in PCO. Cerebrovasc Dis 9: , Giller C, owman G, Dyer H, Mootz L, Krippner W. Cerebral arterial diameters during changes in blood pressure and carbon dioxide during craniotomy. Neurosurgery 7: , Girouard H, Iadecola C. Neurovascular coupling in the normal brain and in hypertension, stroke, and lzheimer disease. J ppl Physiol 1: , Hamel E. Perivascular nerves and the regulation of cerebrovascular tone. J ppl Physiol 1: , Iadecola C, Zhang F. Permissive and obligatory roles of NO in cerebrovascular responses to hypercapnia and acetylcholine. m J Physiol Regul Integr Comp Physiol 71: R99 R11, Ito H, Yokoyama I, Iida H, Kinoshita T, Hatazawa J, Shimosegawa E, Okudera T, Kanno I. Regional differences in cerebral vascular response to Pa CO changes in humans measured by positron emission tomography. J Cereb lood Flow Metab : 16 17,. 16. Katsogridakis E, ush G, Fan L, irch, Simpson DM, llen R, Potter JF, Panerai R. Detection of impaired cerebral autoregulation improves by increasing arterial blood pressure variability. J Cereb lood Flow Metab 33: , Kazama K, Wang G, Frys K, nrather J, Iadecola C. ngiotensin II attenuates functional hyperemia in the mouse somatosensory cortex. m J Physiol Heart Circ Physiol 85: H189 H1899, Kennerley J, Harris S, ruyns-haylett M, oorman L, Zheng Y, Jones M, erwick J. Early and late stimulus-evoked cortical hemodynamic responses provide insight into the neurogenic nature of neurovascular coupling. J Cereb lood Flow Metab 3: 68 8, Li TQ, Moseley ME, Glover G. FIR study of motor cortex activation under normo- and hypercapnia induced by breath challenge. Neuroimage 1: , Matteis M, Vernieri F, Troisi E, Pasqualetti P, Tibuzzi F, Caltagirone C, Silvestrini M. Early cerebral hemodynamic changes during passive movements and motor recovery after stroke. J Neurol 5: , Moody M, Panerai R, Eames PJ, Potter JF. Cerebral and systemic hemodynamic changes during cognitive and motor activation paradigms. m J Physiol Regul Integr Comp Physiol 88: R1581 R1588, 5.. Moreira TS, Takakura C, Colombari E, Guyenet PG. Central chemoreceptors and sympathetic vasomotor outflow. J Physiol 577: , Oldfield RC. The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia 9: , Panerai R, Deverson ST, Mahony P, Hayes P, Evans DH. Effects of CO on dynamic cerebral autoregulation measurement. Physiol Meas : 65 75, Panerai R, Eyre M, Potter JF. Multivariate modeling of cognitivemotor stimulation on neurovascular coupling: transcranial Doppler used to characterize myogenic and metabolic influences. m J Physiol Regul Integr Comp Physiol 33: R395 R7, Panerai R, Moody M, Eames PJ, Potter JF. Cerebral blood flow velocity during mental activation: interpretation with different models of the passive pressure-velocity relationship. J ppl Physiol 99: 35 36, Panerai R, Salinet S, Robinson TG. Contribution of arterial blood pressure and Pa CO to the cerebrovascular responses to motor stimulation. m J Physiol Heart Circ Physiol 3: H59 H66, Panerai R. The critical closing pressure of the cerebral circulation. Med Eng Phys 5: 61 63, Paulson O, Hasselbalch SG, Rostrup E, Knudsen GM, Pelligrino D. Cerebral blood flow response to functional activation. J Cereb lood Flow Metab 3: 1, Paulson O, Strandgaard S, Edvinsson L. Cerebral autoregulation. Cerebrovasc rain Metab Rev : , Posse S, Kemna LJ, Elghahwagi, Wiese S, Kiselev VG. Effect of graded hypo- and hypercapnia on fmri contrast in visual cortex: quantification of T(*)() changes by multiecho EPI. Magn Reson Med 6: 6 71, Rosenblum WI, Kontos H, Wei EP. Evidence for a K(TP) ion channel link in the inhibition of hypercapnic dilation of pial arterioles by 7-nitroindazole and tetrodotoxin. Eur J Pharmacol 17: 3 15, Rosengarten, Spiller, ldinger C, Kaps M. Control system analysis of visually evoked blood flow regulation in humans under normocapnia and hypercapnia. Eur J Ultrasound 16: , Salinet S, Haunton VJ, Panerai R, Robinson TG. systematic review of cerebral hemodynamic responses to neural activation following stroke. J Neurol. In press. 35. Salinet S, Panerai R, Robinson TG. Effects of active, passive and motor imagery paradigms on cerebral and peripheral hemodynamics in older volunteers: a functional TCD study. Ultrasound Med iol 38: , Salinet S, Robinson TG, Panerai R. Reproducibility of cerebral and peripheral haemodynamic responses to active, passive and motor imagery paradigms in older healthy volunteers: a ftcd study. J Neurosci Methods 6: 13 15, Stefanovic, Warnking JM, Rylander KM, Pike G. The effect of global cerebral vasodilation on focal activation hemodynamics. Neuroimage 3: 76 73, Tiecks FP, Lam M, aslid R, Newell DW. Comparison of static and dynamic cerebral autoregulation measurements. Stroke 6: , Tzeng YC, Chan GS, Willie CK, inslie PN. Determinants of human cerebral pressure-flow velocity relationships: new insights from vascular modelling and Ca channel blockade. J Physiol 589: , 11.. vanlieshout JJ, Secher NH. Point:Counterpoint: Sympathetic activity does/does not influence cerebral blood flow. Point: Sympathetic activity does influence cerebral blood flow. J ppl Physiol 15: , Wang Q, Pelligrino D, Koenig HM, lbrecht RF. The role of endothelium and nitric oxide in rat pial arteriolar dilatory responses to CO in vivo. J Cereb lood Flow Metab 1: 9 951, Xu F, Uh J, rier MR, Hart J Jr, Yezhuvath US, Gu H, Yang Y, Lu H. The influence of carbon dioxide on brain activity and metabolism in conscious humans. J Cereb lood Flow Metab 31: 58 67, 11.