The role of nitric oxide in flow-induced and myogenic responses in 1A, 2A, and 3A branches of the porcine middle cerebral artery

Abstract

Myogenic and flow-induced reactivity contribute to cerebral autoregulation, with potentially divergent roles for smaller versus larger arteries. The present study tested the hypotheses that compared with first-order (1A) branches of the middle cerebral artery, second- and third-order branches (2A and 3A, respectively) exhibit greater myogenic reactivity but reduced flow-induced constriction. Furthermore, nitric oxide synthase (NOS) inhibition may amplify myogenic reactivity and abolish instances of flow-induced dilation. Isolated porcine cerebral arteries mounted in a pressure myograph were exposed to incremental increases in intraluminal pressure (40–120 mmHg; n = 41) or flow (1–1,170 µL/min; n = 31). Intraluminal flows were adjusted to achieve 5, 10, 20, and 40 dyn/cm² of wall shear stress at 60 mmHg. Myogenic tone was greater in 3A versus 1A arteries (P < 0.05). There was an inverse relationship between myogenic reactivity and passive arterial diameter (P < 0.01). NOS inhibition increased basal tone to a lesser extent in 3A versus 1A arteries (P < 0.01) but did not influence myogenic reactivity (P = 0.49). Increasing flow decreased luminal diameter (P ≤ 0.01), with increased vasoconstriction at 10–40 dyn/cm² of shear stress (P < 0.01). However, relative responses were similar between 1A, 2A, and 3A arteries (P = 0.40) with and without NOS inhibition conditions (P ≥ 0.29). Whereas NOS inhibition increases basal myogenic tone, and myogenic reactivity was less in smaller versus larger arteries (range = ∼100–550 µM), neither NOS inhibition nor luminal diameter influences flow-induced constriction in porcine cerebral arteries.

NEW & NOTEWORTHY This study demonstrated size-dependent heterogeneity in myogenic reactivity in porcine cerebral arteries. Smaller branches of the middle cerebral artery exhibited increased myogenic reactivity, but attenuated NOS-dependent increases in myogenic tone compared with larger branches. Flow-dependent regulation does not exhibit the same variation; diameter-independent flow-induced vasoconstrictions occur across all branch orders and are not affected by NOS inhibition. Conceptually, flow-induced vasoconstriction contributes to cerebral autoregulation, particularly in larger arteries with low myogenic tone.

INTRODUCTION

In the cerebral circulation, although segmental differences exist, large surface arteries serve as a site of vascular resistance and contribute to cerebral autoregulation (1–4). Among the interrelated mechanisms that contribute to cerebral autoregulation are the intrinsic mechanical properties of cerebral arteries that respond to hemodynamic inputs: myogenic and flow-induced reactivity (3). Myogenic reactivity refers to dilation or constriction of arteries following decreases or increases in transmural pressure, respectively. In a cross-sectional study using cerebral arteries from cortical biopsies from patients with brain tumors (ages 15–75), Bevan et al. (5) demonstrated that the capacity to develop tension in a wire myograph was inversely related to passive arterial diameter. Similarly, in a sample of 12 human pial arteries ranging in diameter from ∼200 to 1,000 µM from a similar patient base, Thorin-Trescases et al. (6) demonstrated that myogenic tone in a pressure myograph was inversely related to passive arterial diameter. Work from Cipolla et al. (7, 8), comparing parenchymal arteries with the middle cerebral arteries in rats, reveals the endothelial contribution to myogenic tone, including the role of nitric oxide (NO) (9), may be related directly to arterial diameter across different vascular segments of the cerebral circulation. Together, these findings provide evidence that diameter dependence of myogenic tone may be related to divergent functional roles for NO in smaller versus larger intracranial arteries.

Flow-induced reactivity refers to dilation or constriction responses to increases in intraluminal flow (3). Conceptually, in the cerebrovasculature, flow-induced dilation may facilitate functional hyperemia, but also impede myogenic reactivity and therefore decrease the efficiency of cerebral autoregulation. In contrast, flow-induced constriction, by increasing vascular tone, may complement myogenic reactivity and therefore increase the efficiency of cerebral autoregulation (3, 10). This latter prospect is based on observations that in larger cerebral arteries myogenic responses alone may be insufficient to compensate for pressure-induced increases in flow (6, 10–12). However, it must be noted that, under in vivo conditions, neither flow-induced constriction nor dilation has been observed in large pial arteries, and only evidence of flow-induced dilation has been observed in basilar arteries (13, 14).

The conditions under which flow mediates dilation or constriction remain unclear, with conflicting results possibly related to differences between the species and vascular bed studied as well as experimental techniques (e.g., in vivo versus ex vivo, wire versus pressure myography, flow rates, etc.). Increasing flow through an artery augments wall shear stress, which has been reported to elicit constriction or dilation, with the latter occurring in an endothelial nitric oxide synthase (NOS)-dependent as well as biphasic manner (3, 10, 14–23). For example, Ngai et al. (20) reported an endothelial NOS-dependent dilation at low flow rates (10 µL/min) followed by a vasoconstriction (restoration toward baseline) at higher flow rates (20 µL/min) in downstream branches of rat middle cerebral arteries (38–55 µM). However, Shimoda et al. (22) reported constriction at low flow rates (70–200 µL/min) and an endothelial NOS-dependent vasodilation (restoration toward baseline) at higher flow rates (∼400–1,500 µL/min) in porcine anterior and middle cerebral arteries (∼493 µM external diameter). Furthermore, in contrast to the work from Ngai et al. (20), Bryan et al. reported flow-induced constriction in the rat middle cerebral artery (15) and similar responses have been observed in rat parenchymal arteries (15, 19). Beyond technical and methodological differences between studies, perhaps similar to the diameter dependence of myogenic tone, arterial diameter or branch order influences flow-induced reactivity and NOS-dependent dilation. Therefore, in the present study, the effect of arterial branch order on myogenic tone and flow-induced responses in porcine pial arteries were examined. A porcine model was selected, as cerebral arterial diameters are similar to those observed in humans (5, 6, 10–12). It was hypothesized that second and third order branches of the middle cerebral artery (2A and 3A, respectively) would display greater myogenic responsiveness, but decreased flow-induced constriction compared with first-order (1A) branches of the middle cerebral artery. Furthermore, it was hypothesized that NOS inhibition would amplify myogenic reactivity in larger arteries, and consistent with previous observations in swine (22), abolish instances of flow-induced dilation at higher levels of shear stress.

METHODS

Animals

All experiments were planned and conducted in accordance with the “Principles for the Utilization and Care of Vertebrate Animals Used in Testing Research and Training” and approved by the University of Missouri as well as the University of Saskatchewan Animal Care and Use Committees and conformed to the NIH guidelines (24). Commercial Landrace pigs (sex = female, n = 20, age = 3 ± 0 mo, mass = 29 ± 1 kg) were housed in the animal care unit under temperature-controlled conditions, with a 12-h/12-h light/dark cycle and consumed a standard commercially available chow diet (5L80, Lab Diet; 2.98 kcal/g, carbohydrate = 70%, protein = 21%, and fat = 9%) with ad libitum access to water. A single sex was studied with the goal of collecting homogeneous data. Intact females were selected owing to superior availability. Pigs at 3 mo of age were selected, reflecting the postweaning, prepubertal period (25), during which pigs have a large brain-to-body mass ratio (26), and are more easily manageable and relatively low-cost compared with market weight pigs. Following an overnight fast, pigs were anesthetized with an intramuscular injection of telozol (5 mg/kg) and xylazine (2.25 mg/kg) and anesthesia was maintained with inhaled isoflurane (5% for 20 min). Pigs were then euthanized by removal of the heart and subsequent exsanguination.

Isolated Artery Studies

The brain was removed and a portion of the brain containing the middle cerebral artery was harvested for ex vivo vasomotor control experiments as described previously (27–30). Briefly, the brain was placed in ice-cold physiological saline solution (PSS: NaCl 145 mM, KCl 4.7 mM, CaCl₂ 2.0 mM, MgSO₄ 1.17 mM and with a pH of 7.4) and 1A, 2A, and 3A branches of the middle cerebral artery were dissected. Where possible, arteries were dissected longitudinally along the same vascular segment. Subsequently, they were transferred to a Plexiglass chamber filled with PSS and cannulated with two resistance and flow-matched glass micropipettes filled with PSS. The chambers were transferred to the stage of an inverted microscope (Nikon Diaphot 200) attached to a video camera (Javelin Electronics, Los Angeles, CA, USA), video micrometer (Microcirculation Research Institute, Texas A&M University), and a Powerlab data acquisition system (ADInstruments, Colorado Springs, CO), as previously described (27–33). Fluid-filled reservoirs were set at an intraluminal pressure at 60 mmHg, chamber temperature was maintained at 37°C, and luminal diameter was monitored throughout the experiment.

Arteries were allotted 45 min to stabilize, at which point maximal arterial vasoconstriction in response to 80 mM KCl was determined. Arteries that did not constrict ≥20% to KCl were discarded. After a 30-min recovery from KCl, arteries either remained untreated (vehicle) or were incubated with the NOS inhibitor N^G-nitro-l-arginine methyl ester (l-NAME; 300 μM) for 20 min before experimentation. Experimentation consisted of three protocols to examine 1) pressure-induced responses, 2) passive mechanical properties, and 3) flow-induced responses.

In protocol 1, intraluminal pressure was decreased to 40 mmHg and then increased in 20 mmHg increments every 5 min (or until a stable diameter was achieved for 2 min) up to an intraluminal pressure of 140 mmHg (30). At the end of the protocol, all vessels were washed twice with Ca²⁺ free PSS to determine maximal passive diameter and pressure curves were repeated in Ca²⁺ free conditions. Myogenic tone at each pressure increment was calculated using the following equation (30):

$$\text{\%Myogenic tone }=\left[\left(I{D}_p-I{D}_a\right)/I{D}_p\right]\times 100,$$

where ID_p is the passive internal lumen diameter in Ca²⁺-free PSS and ID_a is the active internal lumen diameter in PSS. In protocol 2, to examine passive mechanical properties of pial arteries, in a subset of arteries from protocol 1, intraluminal pressure was reduced to 20 mmHg and increased in increments of 20 mmHg until 120 mmHg and both intra- and extra-luminal diameters were recorded. To calculate circumferential stress, mmHg was converted to N/m² (1 mmHg = 1.344 × 10² N/m²). Subsequently, it was calculated at each pressure increment using the following equation (34):

$$\text{Circumferential stress }=(P_{IL}\times ID)/(2WT),$$

where P_IL is the intraluminal pressure, ID is the internal lumen diameter and WT is the wall thickness. Circumferential strain was calculated at each pressure increment using the following equation (34):

$$\text{Circumferential strain }=\left(ID-I{D}_{\text{20}}{}_{\text{mmHg}}\right)/I{D}_{\text{20}}{}_{\text{mmHg}},$$

where ID_20mmHg is the internal lumen diameter at 20 mmHg. Circumferential stress versus strain curves for 1A, 2A, and 3A arteries were curve fitted with an exponential regression equation: y = ke^βx, where y represents the wall stress at a given strain x, k represents the intercept term, and β represents the constant corresponding to the rate of rise of the stress-strain curve (for each branch order, r² > 0.97).

In protocol 3, intraluminal flow was adjusted by changing proximal and distal pressure reservoirs in equal but opposite directions to establish a pressure gradient across the artery while maintaining mean pressure at the midpoint of the artery at 60 mmHg (33, 35, 36). The pressure gradients required to generate the desired flow rates across resistance-matched pipettes were established before experimentation and confirmed with a ball flowmeter (Omega Engineering) as well as direct measurement of fluid across reservoirs. Pressure gradients were variable and selected to produce the estimated flow rate to achieve 5, 10, 20, and 40 dyn/cm² of shear stress. These levels of shear stress were selected to represent a physiological range (37, 38). The estimated wall shear stress stimulus was calculated using the following equation (3, 15, 39):

$$W\text{all shear stress }=32\mu Q/\pi D{i}^3,$$

where µ is the viscosity at 37°C (0.008 P), Q is the volumetric flow rate and Di is the internal arterial diameter. To predict whether flow was laminar, the Reynold’s number was calculated for each experiment using the following formula (40–42):

$$\text{Reynold}’\text{s number}=\rho Div/\text{η},$$

where ρ is the density of the fluid, Di is the internal arterial diameter, v is the velocity, and η is the viscosity of the fluid. Values >3,000 predict turbulent flow, values between 2,000 and 3,000 predict the transition to turbulent flow and values <2,000 predict laminar flow (40, 42). Calculations using the combination of different diameter and flow values from the current study reveal all Reynold’s numbers were <2,000 indicative of laminar flow (1A = 27 ± 27; 2A = 12 ± 13; 3A = 4 ± 4). Flow-induced responses were expressed using the following equation:

$$\text{\%Flow-induced response }=\left[\left(I{D}_f\hspace{0.17em}-I{D}_b\right)/I{D}_b\right]\times 100,$$

where ID_f is the internal lumen diameter in response to varying flow rates and ID_b in the internal lumen diameter at baseline under zero flow conditions.

Statistics

Passive arterial characteristics among 1A, 2A, and 3A arteries were compared using a one-way ANOVA. The %myogenic tone at 40, 60, 80, 100, and 120 mmHg between untreated and NOS inhibition conditions as well as between 1A, 2A, and 3A arteries were compared using a condition × pressure repeated-measures ANOVA. Planned comparisons for 60, 80, 100, and 120 versus 40 mmHg were completed using a Dunnett’s test. Flow-induced changes in absolute diameter for all arteries were compared using a one-way ANOVA. The %Δchange from baseline at 5, 10, 20, and 40 dyn/cm² of wall shear stress between untreated and NOS inhibition conditions as well as between 1A, 2A, and 3A arteries were compared using a condition × shear stress repeated-measures ANOVA. Planned comparisons for 10, 20, and 40 versus 5 dyn/cm² of wall shear stress were completed using a Dunnett’s test. The NOS inhibition data were normalized to the mean of untreated conditions in 1A, 2A, and 3A branches for both myogenic and flow-induced responses and compared using a one-way ANOVA. Planned comparisons for 2A and 3A versus 1A arteries were completed using a Dunnett’s test. Simple linear regression analyses were used to examine the diameter dependency of myogenic and flow-induced responses. Significance was considered when P ≤ 0.05. Where possible, individual data are plotted, and all grouped data are plotted as means ± SD.

RESULTS

Artery Characteristics

Eighty-two isolated cerebral arteries were studied, divided into 25–1A, 28–2 A, and 28–3 A arteries. Overall, 2A arteries had smaller passive diameters than 1A arteries (P < 0.01), and 3A arteries had smaller passive diameters than either 1A or 2A arteries (P < 0.001; Fig. 1A). Although there was overlap in passive diameter across branch orders, with one exception, within each vascular network 1A > 2A > 3A. The KCl-induced vasoconstriction was similar across branch order (P = 0.19; Fig. 1B). Wall thickness was decreased in 2A versus 1A arteries as well as in 3A versus both 1A or 2A arteries (P < 0.001); however, wall:lumen ratios were similar among branch orders (P ≥ 0.98; Fig. 1, C and D). Passive mechanical properties were examined in a subset of arteries (N = 14 per branch order). Among these, circumferential stress and strain were similar among branch orders (P ≥ 0.84) and stress-strain relationships were similar between 1A, 2A, and 3A arteries (k values, P = 0.95; β values, P = 0.38; Fig. 1E).

Figure 1.

Arterial characteristics. Passive arterial diameters (A), KCl-induced constriction (B), wall thickness (C), wall:lumen ratio (D), and circumferential wall stress plotted against circumferential wall strain (E) in 1A (filled circles), 2A (filled squares), and 3A (filled triangles) arteries branching off the middle cerebral artery. For panels A–D, the total number of arteries is as follows: 1A, n = 25; 2A, n = 28; 3A, n = 28. For panel E, the total number of arteries is as follows: 1A, n = 14; 2A, n = 14; 3A, n = 14. A one-way ANOVA was used to compare arterial characteristic across branch order. *P < 0.05 vs. 1A; §P < 0.05 vs. 2A.

Pressure-Induced Responses

Myogenic reactivity was examined in 47 arteries, divided into 16–1A, 15–2 A, and 16–3 A arteries. Arteries that did not display myogenic tone (1–1 A; 3–2 A; 2–3 A) were removed from the analysis. Thus, the final sample size was 41 arteries, divided into 15–1 A, 12–2 A, and 14–3 A arteries. In 1A arteries, NOS inhibition increased myogenic tone (P < 0.01; Fig. 2A), but the effect of pressure only approached significance (P = 0.07). In 2A arteries, both NOS inhibition and pressure increased myogenic tone (P < 0.01; Fig. 2B). Pairwise comparisons revealed within 2A arteries, myogenic tone was greater at 60, 80, 100, and 120 versus 40 mmHg (P < 0.02; Fig. 2B). Within 3A arteries, increased myogenic tone with NOS inhibition only approached significance (P = 0.06; Fig. 2C) and myogenic tone increased with pressure (P < 0.01; Fig. 2C). Pairwise comparisons indicated that myogenic tone was similar between 40 and 60 mmHg (P = 0.32; Fig. 2C), but greater at 80, 100, and 120 versus 40 mmHg (P < 0.02; Fig. 2C). When grouped, both NOS inhibition and pressure increased myogenic tone (P < 0.01), but NOS inhibition did not influence myogenic reactivity (P = 0.49). When divided according to branch order, there was a significant branch order by pressure interaction (P < 0.01). Pairwise comparisons indicated that differences in myogenic tone increased at 60 versus 40 mmHg in 1A, 2A, and 3A arteries (P < 0.01; Fig. 2D), and at 80, 100, and 120 versus 40 mmHg in 2A and 3A arteries (P ≤ 0.02; Fig. 2D). Furthermore, myogenic tone was greater in 3A versus 1A arteries at 80, 100, and 120 mmHg (P ≤ 0.04; Fig. 2D).

Figure 2.

Pressure-induced responses. Myogenic curves in 1A (A), 2A (B), and 3A (C) arteries branching off of the middle cerebral artery in control (open squares) and nitric oxide synthase inhibition (NOS inh.) conditions (Xs). Myogenic curves comparing 1A (filled circles) vs. 2A (open circles) vs. 3A (filled triangles) branch orders (D). A condition × pressure repeated-measures ANOVA was used to compare myogenic tone between conditions at each pressure increment. *P < 0.05, main effect of NOS inhibition; §P < 0.05 vs. 40 mmHg; #P < 0.05 vs. 1A.

Flow-Induced Responses

Flow-induced responses were examined in a subset of 34 arteries, divided into 9–1 A, 13–2 A, and 12–3 A arteries. Among these arteries, increasing intraluminal flow resulted in a vasoconstriction in 33/34 arteries. One artery displayed flow-induced dilation (−3%, 14%, 22%, and 23% vasodilation at 5, 10, 20, and 40 dyn/cm² of wall shear stress, respectively) that was blocked with NOS inhibition in a second artery from the same vascular segment (1%, 1%, 1%, and 10% vasoconstriction at 5, 10, 20, and 40 dyn/cm² of wall shear stress, respectively). Also, one artery exhibited vasomotion (e.g., vasoconstriction ranged 2%–11% at 40 dyn/cm² wall shear stress). These data (n = 3) were excluded from the final analysis and the final samples size was 31 arteries, divided into 9–1 A, 11–2 A, and 11–3 A arteries. Individual flow-induced constriction data in control and NOS inhibition conditions are presented in Fig. 3, A–C (all data grouped). To examine the relationship between flow and diameter, arterial diameter changes were subdivided into quartiles according to the flow rate. The change in absolute arterial diameter was similar between the first and second quartiles (P = 0.28), greater in the third than the first quartile (P = 0.01; Fig. 3D), and greater in the fourth versus first, second, and third quartiles (P ≤ 0.01; Fig. 3D).

Figure 3.

Flow-induced responses. Flow-diameter curves in 1A (A), 2A (B), and 3A (C) arteries branching off of the middle cerebral artery in control (full line) and nitric oxide synthase inhibition (NOS inh.) conditions (dashed line). ΔDiameter at each quartile of flow for all arteries in control (open squares) and NOS inh. (Xs) conditions (D). First quartile = 1–9 µL/min; second quartile = 11–30 µL/min; third quartile = 31–92 µL/min; fourth quartile = 100–1,170 µL/min. In A–C, the thin lines represent individual flow-diameter curves and the thick lines represent the mean. A one-way ANOVA was used to compare Δdiameter at each quartile of flow. *P < 0.05 vs. first quartile (1–9 µL/min); §P < 0.05 vs. second quartile (11–30 µL/min); #P < 0.05 vs. third quartile (31–92 µL/min).

To evaluate relative responses, changes in absolute arterial diameter were normalized to baseline diameter and flow rates were converted to the estimated wall shear stress. Further, to evaluate the role of NO, data for control and NOS inhibition conditions were compared. Vasoconstriction at varying magnitudes of wall shear stress were not different between NOS inhibition and control conditions in either 1A, 2A, or 3A arteries (P ≥ 0.29; Fig. 4, A–C). Within 1A arteries, constriction was greater at 40 versus 5 dyn/cm² of wall shear stress (P < 0.01; Fig. 4A). Within 2A arteries, constriction was greater at 20 and 40 versus 5 dyn/cm² of wall shear stress (P < 0.01; Fig. 4B). Within 3A arteries, constriction was greater at 10, 20, and 40 versus 5 dyn/cm² of wall shear stress (P < 0.01; Fig. 4C). There was no effect of arterial branch order on flow-induced constriction (P = 0.40; Fig. 4D), but there was a main effect of shear stress (P < 0.01; Fig. 4D). Pairwise comparisons revealed that vasoconstriction was greater at 10, 20, and 40 versus 5 dyn/cm² of wall shear stress among all cerebral arteries (P < 0.01; Fig. 4D).

Figure 4.

Shear stress-induced responses. %Vasoconstriction at 5, 10, 20, and 40 dyn/cm² of wall shear stress in in 1A (A), 2A (B), and 3A (C) arteries branching off the middle cerebral artery in control (open squares) vs. nitric oxide synthase inhibition (NOS inh.) conditions (Xs). %Vasoconstriction at 5, 10, 20, and 40 dyn/cm² of wall shear stress comparing 1A (filled circles) vs. 2A (open circles) vs. 3A (filled triangles) branch orders (D). A condition × shear stress repeated-measures ANOVA was used to compare vasoconstriction at different levels of shear stress. *P < 0.05 vs. 5 dyn/cm².

Diameter Dependency

To examine the diameter dependency of myogenic reactivity and NOS inhibition, the %Δdiameter between 40 and 120 mmHg for all arteries were plotted against passive diameters and the NOS-inhibition-mediated increases in myogenic tone (normalized to control conditions) in 1A, 2A, and 3A branches were compared. Linear regression analysis revealed larger arteries displayed less myogenic reactivity (R = 0.51; P < 0.01; Fig. 5A). Compared with 1A, 3A branches exhibited less NOS-inhibition-mediated increases in myogenic tone (P < 0.01; Fig. 5B). To examine the diameter dependency of flow-induced vasoconstriction and NOS inhibition, the %Δdiameter between 0 and 40 dyn/cm² was plotted against passive diameters and the NOS-inhibition-mediated increases in flow-induced constriction (normalized to control conditions) in 1A, 2A, and 3A branches were compared. Linear regression analysis revealed no relationship between passive diameter and flow-induced constriction (R = 0.11; P = 0.57; Fig. 5C). Similarly, the effect of NOS inhibition on flow-induced constriction was similar among arterial branch orders (P ≥ 0.41; Fig. 5C).

Figure 5.

Diameter and NOS dependency of flow and pressure-induced responses. Myogenic reactivity (%Δdiameter between 40 and 120 mmHg) plotted against passive arterial diameters at 40 mmHg (A). The effect of nitric oxide synthase inhibition (NOS inh.) to increase myogenic tone in 1A (filled circles), 2A (open circles), and 3A (filled triangles) arteries branching off the middle cerebral artery (B). Flow-induced constriction (FIC; (%Δdiameter between 0 and 40 dyn/cm²) plotted against passive arterial diameters (C). The effect of NOS inh. to increase FIC in 1A, 2A, and 3A arteries branching off of the middle cerebral artery (D). In A and C, the filled line represents the line of best fit and the dotted line represents the 95% confidence intervals. A linear regression was used to examine diameter dependency of vasomotor responses and a one-way ANOVA was used to examine the role of NOS inh. on vasomotor responses across arterial branch orders. *P < 0.05 vs. 1A.

DISCUSSION

This study examined the dual roles of diameter or branch order and NOS signaling on myogenic and flow-induced reactivity in large surface arteries from the same vascular network along the cerebrovascular tree. The findings reveal that myogenic tone was greater in 3A versus 1A branches of the middle cerebral artery and myogenic reactivity was inversely related to passive arterial diameter. However, flow-induced vasoconstriction was similar across arterial branch orders and unrelated to passive arterial diameter. NOS inhibition increased basal myogenic tone, with increases being less pronounced in 3A versus 1A arteries; however, it did not affect constrictor responses to pressure or flow. Concerning flow-induced responses, in contrast to the hypothesis, constriction was observed at lower and higher shear rates, irrespective of arterial diameter or branch order. These data suggest NOS attenuates basal myogenic tone, but that myogenic reactivity and flow-induced constriction operate independently from NOS. Further, myogenic and flow-induced responses as well as functional NOS signaling may contribute to segmental differences in vasomotor control along the cerebrovascular tree.

In the present study, myogenic tone was greater in 3A versus 1A branches of the middle cerebral artery from 80 to 120 mmHg, and myogenic reactivity was greater in smaller versus larger arteries ranging from ∼100 to 550 µM. Furthermore, the modulatory role of NOS was reduced in 3A versus 1A arteries, suggesting that reduced NO signaling may be implicated in the elevated myogenic responses. Evidence indicates, that in the range of pressures studied, elevated myogenic tone may be the result of increased calcium entry into the vascular smooth muscle, and augmented myogenic reactivity may be the result of increased calcium sensitivity (11). Of note, NO may decrease calcium entry and sensitivity in vascular smooth muscle (43–45). Previous work indicates that basal myogenic tone and voltage-dependent calcium channel activity is greater in downstream parenchymal versus middle cerebral arteries in rats. Concurrently, the contribution of NOS to basal myogenic tone also appears reduced in parenchymal versus middle cerebral arteries (7, 8). Taken together, reduced NOS coupled with enhanced calcium signaling in 3A versus 1A arteries may contribute to the segmental differences observed herein. Importantly, although such differences may contribute to enhanced autoregulation in downstream arteries (46), they may result in segmental heterogeneity in hypertension-induced vascular remodeling. That is, hypertension may disproportionately affect vascular remodeling in downstream arteries, as they exhibit heightened myogenic reactivity and may obtain less protection against remodeling from basal NO.

In contrast to myogenic properties, flow-induced vasoconstriction at different levels of wall shear stress was similar among 1A, 2A, and 3A branches of the middle cerebral artery. There was no relationship between flow-induced vasoconstriction and passive arterial diameter. Thus, although myogenic reactivity was inversely related to passive arterial diameter, in this experimental set-up, relative flow-induced vasoconstriction (between 5 and 40 dyn/cm² of shear stress) was not. These data highlight that in large arteries with low myogenic reactivity, where basal NOS signaling is elevated and myogenic responses alone may be unable to compensate for pressure-induced increases in flow (6, 10–12), flow-induced constriction may contribute significantly to cerebral autoregulation (3, 10).

The conditions under which flow elicits cerebral dilation or constriction and the mechanisms responsible are not clear. In the current study, only one artery exhibited flow-induced dilation and it was abolished with NOS inhibition in a corresponding vessel segment from the same artery. Thus, with only a single observation, the current study is under powered to address the stated hypothesis that NOS inhibition would abolish instances of flow-induced dilation. Concerning flow-induced constriction, it has been suggested that it may result from a wash-out effect of local vasodilator substances, including NO. Likewise, because NO can inhibit the production of 20-hydroxyeicosatetraenoic acid, a candidate mechanism of flow-induced constriction, inhibition of NOS could augment flow-induced constriction (3, 10). However, findings from the current study refute such prospects, as NOS inhibition did not augment flow-induced constriction. This observation is in agreement with previous work that indicates flow-induced vasoconstriction is NOS independent [and involves a combination of cytoskeletal integrin signaling (15, 47, 48), increased cyclooxygenase activity, reactive oxygen species production, as well as 20-hydroxyeicosatetraenoic acid acting on thromboxane A₂ receptors (10)]. Given flow-induced dilation appears to be NOS dependent (14, 16, 20, 22, 49), flow-induced constriction may represent a mechanistically independent phenomenon that contributes differentially to cerebral autoregulation. However, this has yet to be confirmed under in vivo conditions.

Work from Toth and Koller indicates that myogenic reactivity and flow-induced constriction must act in concert to maintain constant cerebral blood flow in response to increases in intravascular pressure (3, 10). Of note, evidence indicates the magnitude of flow-induced vasoconstriction is either the same across a range of pressures within the autoregulatory range (40–140 mmHg) (10) or may increase with increasing intravascular pressure (from 60 to 90 mmHg) (18). Given that myogenic reactivity appears to be greater in smaller versus larger arteries, it is possible the combination of flow-induced constriction and myogenic reactivity leads to more efficient cerebral autoregulation in smaller versus larger arteries, particularly at higher intravascular pressures. However, without corresponding in vivo data to support this notion, it remains speculative.

Limitations of the current experimental approach must be noted when interpreting the present data. Specifically, only a single sex was studied, and several experimental factors differed from in vivo conditions. For example, the study was conducted in isolated cerebral arteries from postweaned, prepubescent female swine under ex vivo conditions and myogenic reactivity was examined in the absence of flow. Further, flow-induced responses were examined using a physiological salt solution and without a peristaltic pump. As a result, the viscosity of the perfusate was less than blood and the flow profiles were not pulsatile. Further, the composition of the perfusate differed from blood and cerebral spinal fluid, which may have influenced vasomotor responses. Of note, these factors which differ from in vivo conditions may have decreased endothelial NOS signaling (14, 16, 21, 34, 50) and possibly enhanced myogenic tone. Thus, the current data do not undermine the prospect that endothelial NOS signaling mediates flow-induced dilation in cerebral arteries under normal physiological conditions.

This study demonstrated that myogenic reactivity was greater whereas NOS inhibition-mediated increases in myogenic tone were less in 3A versus 1A branches of the middle cerebral artery. Furthermore, in contrast to myogenic responses, flow-induced vasoconstriction was similar across branch orders and unrelated to passive luminal diameter or NOS signaling. Taken together, these findings provide evidence that flow-induced vasoconstriction may contribute to cerebral autoregulation, particularly in large arteries with low myogenic reactivity. These findings provide functional insight into segmental differences in autoregulation in large surface arteries along the cerebrovascular tree.

GRANTS

T.D.O. is supported by NSERC (Discovery Grant 05323) as well as a College of Veterinary Medicine Pilot Grant, University of Missouri. J.P. is supported by National Institutes of Health (NIH) Grant K01-HL-125503.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

C.J.M., E.M.B., M.W.M., J.P., and T.D.O. conceived and designed research; C.J.M. and T.D.O. performed experiments; C.J.M. and T.D.O. analyzed data; C.J.M., E.M.B., M.W.M., J.P., and T.D.O. interpreted results of experiments; T.D.O. prepared figures; C.J.M. and T.D.O. drafted manuscript; C.J.M., E.M.B., M.W.M., J.P., and T.D.O. edited and revised manuscript; C.J.M., E.M.B., M.W.M., J.P., and T.D.O. approved final version of manuscript.