Effect of Steady Versus Oscillating Flow on Porcine
Coronary Arterioles
Involvement of NO and Superoxide Anion
Oana Sorop, Jos A.E. Spaan, Terrence E. Sweeney, Ed VanBavel
Downloaded from http://circres.ahajournals.org/ by guest on June 15, 2017
Abstract—Coronary blood vessels are compressed by the contracting myocardium. This leads to oscillations in flow in
especially the subendocardium. We examined the effects of steady and oscillating flow on isolated, cannulated
subendocardial and subepicardial porcine arterioles. Steady flow–induced dilation in both vessel types, up to 12.9⫾0.8%
of the passive diameter in subendocardials and 9.6⫾1.4% in subepicardials at 40 dyne/cm2. Dilation was completely
abolished after treatment with 10 ␮mol/L L-NNA. Sinusoidal modulation of steady flow at 1.5 Hz and 50% to 200%
amplitude did not affect dilation. Oscillating flow without a net forward component with peak-peak shear values up to
100 dyne/cm2 caused no dilation at all in these vessels. However, in the presence of 100 U/mL superoxide dismutase
(SOD), oscillating flow induced dilation up to 19.5⫾2.3% in subendocardial vessels and 11.5⫾4.3% in subepicardials.
LNNA (10 ␮mol/L) blocked this dilation by approximately 50%. SOD did not affect the magnitude of steady
flow-induced dilation, but the response time after onset of steady flow shortened from 23.4⫾1.5 to 14.3⫾2.1 seconds.
Diphenyleneiodinium, an inhibitor of NAD(P)H oxidase, uncovered dilation to oscillating flow in subendocardial
vessels up to 9.5⫾1.6%. Flow causes production of both NO and O2⫺. During steady flow, the bioavailability of NO
is sufficient to cause vasodilation. During oscillating flow, NO is quenched by the O2⫺, suppressing vasodilation.
Considering the pulsatile nature of subendocardial flow and the vulnerability of this layer, pharmacological
manipulation of the balance between NO and O2⫺ may improve subendocardial perfusion. (Circ Res. 2003;92:13441351.)
Key Words: flow-induced dilation 䡲 coronary arterioles 䡲 oxidative stress
䡲 oscillating flow 䡲 superoxide dismutase
L
lack of cell alignment in the flow direction,17 increased expression of adhesion molecules,18 lack of eNOS upregulation,19,20
chronic increase in superoxide anion production,21 and different
intracellular calcium transients22 are found. However, these
differential effects of steady versus oscillating flow on endothelial biology cannot be extrapolated to vascular tone. Therefore,
we aimed to directly assess the effect of steady versus oscillating
flow on cannulated resistance arteries from the subendocardium
(ENDO) and subepicardium (EPI). We show that steady but not
oscillating flow caused dilation of ENDO and EPI vessels,
which fully depended on NO. Addition of superoxide dismutase
(SOD) or the NAD(P)H oxidase inhibitor diphenyleneiodinium
(DPI) did not affect the response to steady flow but revealed a
dilatory response to oscillating flow.
ocal coronary blood flow varies during the cardiac cycle.
This is caused by the rhythmic generation of intramyocardial pressure1 and elastance.2 The resulting compression
impedes perfusion during systole, when even retrograde flow
may appear. The compression in the subendocardium contributes to the vulnerability of this layer to ischemia. The
deleterious effects of this compression may be overcome by
dilatory effects of pulsatile pressure and flow. Using isolated
porcine coronary arterioles subjected to pulsatile transmural
pressure3 at zero flow, we previously found vasodilation and
increased sensitivity to vasodilators during pulsation compared with steady pressure. Variations in flow also could
affect vascular tone. Many small vessels,4 –15 including epicardial resistance vessels,8 show flow-dependent dilation to
steady flow. Yet, little is known about the effect of pulsatile
flow on these blood vessels. In addition, effects of flow on
subendocardial vessels have not been studied at all.
Some evidence for contrasting responses to steady and pulsatile flow comes from studies on cultured endothelial cells.16
When these cells are exposed to oscillatory versus steady flow,
Materials and Methods
Surgery and Vessel Cannulation
All animal protocols were approved by the institutional laboratory
animal care committee. Thirty-eight Yorkshire female pigs (H.W.
Vendrig farm, Amsterdam, The Netherlands; 17 to 26 kg, 12 to 18
Original received August 27, 2002; resubmission received February 27, 2003; revised resubmission received May 14, 2003; accepted May 14, 2003.
From the Department of Medical Physics, University of Amsterdam, Academic Medical Center, Amsterdam, The Netherlands.
Correspondence to Dr Ed VanBavel, Dept of Medical Physics, University of Amsterdam, Academic Medical Center, PO Box 22700, 1100 DE
Amsterdam, The Netherlands. E-mail [email protected]
© 2003 American Heart Association, Inc.
Circulation Research is available at http://www.circresaha.org
DOI: 10.1161/01.RES.0000078604.47063.2B
1344
Sorop et al
Oscillating Flow Effects in Coronary Arterioles
1345
Figure 1. Example of a coronary ENDO
arteriole subjected to steady flow with
initial shear of 20 dyne/cm2 (left), oscillating flow of 10 dyne/cm2 (middle), and
modulated flow with 10 dyne/cm2 and
100% amplitude (right). Panels depict
the applied left and right pressure levels,
the measured inner diameter, and the
calculated luminal pressure, flow, and
shear stress (see Materials and Methods). Gray boxes represent 2 second
expansions of the time scale.
Downloaded from http://circres.ahajournals.org/ by guest on June 15, 2017
weeks old) were anesthetized by 4% halothane, ketamine (20
mg/kg), midazolam (1 mg/kg IM), and atropine (0.05 mg/kg). The
animal was intubated and ventilated (O2/N2O, 1:2), the ear vein was
cannulated, and midazolam (0.2 mg/kg) was administrated intravenously. After midsternal thoracotomy and heparinization (0.1 mL/kg,
intravenously), the heart was fibrillated, excised, and immediately
placed in cold (4°C) MOPS-buffered Ringer’s solution (in mmol/L:
NaCl 145.0, KCl 4.7, MgSO4 1.17, CaCl2 2.0, NaH2PO4 1.2, glucose
5.0, and pyruvate 2.0; the solution was equilibrated with air; pH
7.35⫾0.02). Dissection was performed in MOPS-buffered Ringer’s
solution containing 1% albumin at 4°C. Subepicardial and subendocardial arterioles were dissected, cannulated at both ends with glass
micropipettes, tested for leaks, and set to their in situ lengths.
Each cannula was supplied by a reservoir of MOPS buffer
containing 1% bovine serum albumin. Each reservoir was separately
pressurized by a computer-driven Venturi valve (FAIRCHILD
T5200 –50). Internal diameter of the vessel was continuously measured using a video technique. The vessel was superfused with
MOPS buffer at a rate of 3 mL/min at 37°C. Drugs were added to the
superfusion medium.
The vessels had internal diameters of 100 to 200 ␮m at full
dilation and 60 mm Hg. Such vessels do not always develop
substantial basal tone. In order to have a consistent tone level in all
protocols, we preconstricted all vessels with 1 ␮mol/L U46619. All
vessels developed a similar level of constriction. Flow protocols
were started once the level of tone reached steady state.
Application of Flow Profiles
All flow profiles were applied at a constant intraluminal pressure of
60 mm Hg. To calculate the cannula pressures for the desired shear
stress profiles, ␶(t), we first determined the required flow profiles,
Q(t), from the following:
(1)
␲ 䡠 d3
䡠 ␶ 共t兲,
Q共t兲⫽
32 䡠 ␩
with ␶(t) the shear stress, d the internal diameter after development of
constant tone, and ␩ the viscosity of 0.001 N · s/m2. Then, the left and right
pressure profiles needed for these values of flow at a constant intraluminal
pressure of 60 mm Hg were determined from the following:
Pl 共t兲⫽60⫹Q 共t兲䡠Rl
(2)
Pr 共t兲⫽60⫺Q 共t兲䡠Rr,
with Rl and Rr the left and right cannula resistances, as determined
before start of the experiments. We used pipettes with matched
resistances. Based on geometry, the resistance of the cannulated
vessel segment was negligible compared with that of the cannulas.
Protocols
Experimental Group 1 (ENDO and EPI)
Steady flow (Figure 1 left traces), with shear between 1 and 90
dyne/cm2, was achieved by a step change in cannula pressure from
60 mm Hg to the values calculated according to the above formulas.
The different shear values were randomized. Flow was maintained
for 3 minutes; flow steps were separated by 3-minute periods of zero
flow. In some experiments, the pressure gradient, and consequently
the flow through the vessels were reversed during the flow period.
For purely sinusoidal oscillating flow without any net forward
component (oscillating flow; Figure 1, middle), computer-generated
sinusoidal pressure oscillations were applied to both cannulas. These
oscillations, superimposed on 60 mm Hg, were in anti-phase and had
a frequency of 1.5 Hz and amplitudes as calculated above. Peak-peak
shear values of 1 to 100 dyne/cm2 were applied in random order.
Each oscillating shear level was applied for 3 minutes, separated by
3 minutes no-flow. Sinusoidal modulation of flow (“modulated
flow”) was achieved by superimposing flow oscillations with a sine
waveform at 1.5 Hz, on previously initiated net steady forward flow
(Figure 1, right). The flow modulation amplitudes were 50%, 100%,
150%, and 200% of the steady flow value. For 100% modulation,
flow just reached zero. Larger amplitudes include back flow, such as
may occur in vivo during systole. These modulations were randomly
applied on top of steady shear values of 10, 30, 60, and 100 dyne/cm2
and maintained for 3 minutes, separated by 3 minutes steady flow.
Experimental Group 2 (ENDO and EPI)
To test the involvement of NO in flow-mediated dilation, steady flow
steps were applied before and 30 to 40 minutes after the start of
superfusion with 10 ␮mol/L L-NNA.
1346
Circulation Research
June 27, 2003
Group 3 (ENDO and EPI)
Steady and oscillating flow were applied before and 30 minutes after
the start of superfusion with 100 U/mL superoxide dismutase (SOD).
Groups 4 and 5 (ENDO Vessels)
The mechanisms of oscillating flow-induced dilation in the presence
of SOD were tested. After preconstriction, vessels were incubated in
100 U/mL SOD. Oscillating flow was applied. Subsequently, the
vessels were incubated in either 10 ␮mol/L L-NNA as eNOS
inhibitor or 80 U/mL catalase as H2O2 scavenger and the flow
oscillations were repeated.
Group 6 (ENDO Vessels)
Downloaded from http://circres.ahajournals.org/ by guest on June 15, 2017
The effect of diphenyleneiodinium (DPI), a blocker of flavincontaining enzymes, such as NAD(P)H oxidase (possible sources of
free radicals), was tested. Vessels were incubated with 30 ␮mol/L
DPI for 30 minutes. Steady and oscillating flow protocols were
applied before and during the incubation. Vessels in groups 3 to 6
were tested at 5 to 30 dyne/cm2. These vessels were smaller than
those in groups 1 to 2 and were found to be sensitive to lower shears.
To establish that a graded dilation to flow occurred in these small
vessels, such vessels were also subjected to steady shear between 1
to 5 dyne/cm2 (see online Figure 2 in the online data supplement
available at http://www.circresaha.org).
Group 7 (Rat Cremaster Vessels)
In order to compare flow effects on vessels with preconstriction
versus basal tone, we repeated the flow protocols of group 1 on
comparably sized first order resistance arteries isolated from the rat
cremaster muscle. These vessels develop stable, constant basal tone.
Accuracy of Flow Velocity Patterns
The accuracy of the experimental set-up in generating the desired
levels of flow and shear was evaluated in separate experiments using
an optical Doppler intravital velocimeter (Texas A&M University
System Health Science Center). Red blood cell velocity was measured in two cannulated arterioles perfused with MOPS buffer
containing rat erythrocytes. The measured centerline velocity
matched the value calculated on the basis of cannula resistances, left
and right cannula pressures, vessel diameter, and Poiseuille flow
within 10%. This was the case for steady flow as well as for
oscillating and modulated flow with frequencies between 0.5 and 2
Hz (online Figure 1). In order to estimate whether the velocity profile
in the vessel was fully parabolic during oscillating flow, we
calculated the Womersley number (WO). When WO exceeds unity,
the velocity profile is no longer parabolic, and the flow is phaseshifted in time relative to the oscillating pressure gradient.23 WO in
our experiments did not exceed 0.01, indicating quasi-steady behavior of the fluid with a parabolic flow profile during all protocols.
Drugs
Bradykinin, U46619, L-NNA, SOD S-2515, Catalase, and DPI were
purchased from Sigma Chemicals.
Data Analysis
All vascular diameters as well as the degree of dilation were
normalized to maximum dilated diameters of the vessels, as obtained
in response to 0.1 ␮mol/L bradykinin at 60 mm Hg intraluminal
pressure. For each 3-minute flow intervention, the time-averaged
diameter was obtained after development of a stable response. Data
are presented as mean⫾SEM. Data were analyzed by one-way and
two-way ANOVA, including the interaction term followed by
Bonferroni post hoc tests where appropriate. Differences were
considered significant for P⬍0.05.
Results
Effects of the various flow patterns were tested in 7 ENDO
(159⫾15 ␮m inner diameter) and 6 EPI (163⫾18 ␮m inner
diameter) arterioles (experimental group 1). In response to a
Figure 2. Steady and oscillating flow-induced responses. A, In 7
ENDO and 6 EPI arterioles, steady flow induced dilation. The
dilation was significantly larger in ENDO than in EPI vessels
(P⬍0.01, ANOVA). B, Oscillating flow up to 100 dyne/cm2 did
not affect the vascular diameter in 5 ENDO and 6 EPI arterioles
(P⫽NS, ANOVA).
sudden onset of steady flow, all vessels developed flowinduced dilation. Figure 1 (left) depicts a typical example.
Dilation developed after a lag time of 15 to 25 seconds from
the flow onset, and reached a plateau value in approximately
1 minute. This value of dilation remained constant for as long
as flow was maintained. After stopping flow, the vessel
regained its initial level of tone within 2 minutes. The middle
panels of Figure 1 depict the response of this vessel to
oscillating flow. In this example, flow-dependent dilation did
not occur in response to an oscillating shear stress of 10
dyne/cm2 at 1.5 Hz. The right panels indicate that 100%
modulation of 10 dyne/cm2 steady flow did not induce extra
vasodilation.
Figures 2 and 3 summarize the effects of the various flow
patterns on tone. For steady flow (Figure 2A), dilation in
ENDO and EPI vessels became significant at 10 dyne/cm2.
Dilation became larger with increasing shear stress up to
approximately 40 dyne/cm2 and tended to decrease again at
still higher shear values. Dilation was significantly larger in
ENDO as compared with EPI vessels (P⬍0.05, two-way
ANOVA).
A lack of dilation in response to oscillating flow was found
in all vessels; the diameter of some vessels remained unchanged, whereas others (both ENDO and EPI) showed a
small constriction. Figure 2B presents the average results of
experiments performed on 5 ENDO and 6 EPI arterioles.
Oscillating flow at peak-peak shear levels up to 100 dyne/cm2
Sorop et al
Downloaded from http://circres.ahajournals.org/ by guest on June 15, 2017
Figure 3. Effects of modulating flow at 10 (A) and 30 (B) dyne/cm2
shear by 50% to 200% on 7 ENDO and 6 EPI vessels. Modulation
of flow did not affect the steady flow effect, even for amplitudes
where flow reversal was present in the vessel.
did not significantly change the vessel tone. (P⫽NS, two-way
ANOVA). In 2 tested vessels, oscillating flow profiles at
lower (down to 0.1 Hz) and higher (2 Hz) frequencies and
similar shear values also did not induce flow-dependent
dilation (data not shown).
The effects of flow modulation on 7 ENDO and 6 EPI
arterioles are depicted in Figure 3. Modulating shears of 10
and 30 dyne/cm2 at 1.5 Hz did not significantly change the
level of flow-induced dilation (P⫽NS, two-way ANOVA).
Even at amplitudes of 150% and 200%, when retrograde flow
occurred, and the root-mean-square level of shear stress was
increased by respectively 46% and 73% as compared with
steady flow, the diameter remained constant during the 3
minutes of flow modulation. This was the case for both
ENDO and EPI arterioles. Modulation of shear values around
averages of 60 and 100 dyne/cm2 up to 200% also did not
cause additional dilation (data not shown).
The order of the steady, oscillating, and modulated flow
protocols was randomized. In those cases where steady flow
was applied after the other protocols, normal dilation occurred. Sensitivity to bradykinin was tested in all vessels after
finishing all protocols and found to be normal, indicating
undamaged endothelial cells (data not shown).
In 10 vessels (5 ENDO, 5 EPI), the effect of L-NNA was
tested. In this group, shear-induced dilation before L-NNA
was significant above 5 dyne/cm2. In ENDO and EPI vessels,
incubation with 10 ␮mol/L L-NNA in the absence of flow for
30 minutes induced a 3% to 4% constriction. L-NNA completely abolished dilation over the whole range of shear
Oscillating Flow Effects in Coronary Arterioles
1347
Figure 4. Incubation of 5 ENDO (A) and 5 EPI (B) arterioles with
10 ␮mol/L L-NNA completely inhibited steady flow-induced dilation in both types of vessels. P⬍0.01, ANOVA, with vs without
L-NNA.
stresses in ENDO and EPI vessels (Figure 4; P⬍0.05,
two-way ANOVA).
In 7 ENDO (127⫾10-␮m inner diameter) and 6 EPI
(132⫾12-␮m inner diameter), steady and oscillating flow
protocols with shears up to 30 dyne/cm2 were applied before
and after incubation for 30 minutes with 100 U/mL SOD.
SOD had no effect on the basal diameter of the vessels.
Figures 5A and 5B show that the magnitude of the response
to steady flow remained unchanged (P⫽NS, two-way
ANOVA). Whereas oscillating flow in the absence of SOD
had no effect on vessel diameter, Figures 5C and 5D show
that after treatment with SOD such oscillating flow caused
significant dilation in all vessels (P⬍0.05, 2-way ANOVA;
eg, at 30 dyne/cm2: 0.0⫾1.2% versus 19.5⫾2.3% in ENDO
and 0.0⫾1.1% versus 11.5⫾4.3% in EPI vessels, before
versus after incubation). In ENDO vessels in the presence of
SOD, the response to oscillating flow at 30 dyne/cm2 peakpeak even exceeded the dilation to steady flow (P⬍0.05,
paired t test, n⫽7).
Although SOD did not affect the magnitude of dilation to
steady shear, it did affect the latency of the dilatory response
(Figure 6). We found a 21.8⫾2.7-second delay to 25% of the
dilatory response after onset of steady flow, which was
independent of the level of shear (Figure 6A; P⫽NS; n⫽7
ENDO⫹6 EPI). After incubation with SOD, the delay time
was reduced from 23.4⫾1.5 to 14.3⫾2.1 seconds at 10
dyne/cm2 (Figure 6B, n⫽5; P⬍0.05, one-way ANOVA). In
the presence of SOD, the delay time for start of dilation after
onset of oscillating flow was slightly larger than for steady
flow (16.3⫾1.8 second; P⫽0.044).
We tested in 4 ENDO vessels (128.5⫾11.2 ␮m) whether
NO mediates oscillating flow-induced dilation in presence of
1348
Circulation Research
June 27, 2003
Figure 5. In 7 ENDO and 6 EPI arterioles, steady and oscillating flow with
shears up to 30 dyne/cm2 were tested
before and after incubation with 100
U/mL SOD. Treatment with SOD did not
affect the magnitude of steady flowinduced dilation (left panels; P⫽NS,
ANOVA). SOD completely uncovered an
oscillating flow-induced dilation in EPI
vessels. ENDO vessels in the presence
of SOD dilated even more during oscillating flow than during steady flow (right
panels).
Downloaded from http://circres.ahajournals.org/ by guest on June 15, 2017
SOD. Oscillating flow with peak-peak shears up to 30
dyne/cm2 was applied. L-NNA (10 ␮mol/L) reduced the
dilation in the presence of SOD by approximately 50%
(Figure 7A; P⬍0.05, two-way ANOVA). The remaining
dilation could not be attributed to an H2O2 mediated mechanism, because in 5 ENDO vessels (134.3⫾12.1-␮m inner
diameter), 80 U/mL catalase did not significantly diminish
the SOD effect during oscillating flow (Figure 7B; P⫽NS,
SOD versus SOD⫹catalase). In two vessels tested, this
concentration of catalase reduced the dilation to 10⫺5 mol/L
extrinsic H2O2 from 50.7⫾10.0% to 8.8⫾1.0%. In other
experiments, raising the catalase concentration from 80 to
1000 U/mL only modestly impaired oscillating flow-induced
dilation: from 9.1⫾1.0% (SOD) to 7.3⫾1.0% (SOD⫹1000
U/mL catalase) at 10 dyne/cm2 (n⫽2), and from 18.4% to
16.5% at 30 dyne/cm2 (n⫽1).
Five ENDO arterioles (129.7⫾6.8 ␮m) were incubated
with 30 ␮mol/L of the NAD(P)H oxidase inhibitor DPI. The
dilation to oscillating flow was recorded for peak-peak shears
of 5 to 30 dyne/cm2. Oscillating flow-induced dilation was
absent before incubation with DPI, but became significant at
10 dyne/cm2 in the presence of DPI (Figure 7C). The dilation
was approximately 20% to 30% less in the presence of DPI as
compared with SOD, as seen in Figures 5C, 7A, and 7B.
Indeed, in 3 of these 5 vessels, we also tested the effect of
SOD and found it to be significantly larger. DPI did not affect
steady flow-induced dilation.
The above effects on coronary vessels were obtained after
preconstriction by U46619. We tested in 9 first order rat
cremaster arterioles (159⫾19 ␮m) whether vessels with basal
tone also lack responsiveness to oscillating flow. Basal tone
developed to approximately 65% of the maximal diameter.
Figure 8 indicates that steady flow caused dilation above 25
dyne/cm2. Furthermore, sinusoidal modulation of flow at
100% did not affect the dilation in also these vessels.
However, also in these vessels, dilation to oscillating flow at
1.5 Hz was absent up to peak-peak amplitudes of 90 dyne/cm2
(P⫽NS, steady versus modulated flow; P⬍0.01 steady versus
oscillating; P⫽NS, oscillating flow versus no flow, ANOVAs).
The effect of SOD was not tested in these vessels with basal
tone. We were able to cannulate 3 ENDO vessels smaller than
100 ␮m, which developed basal tone that averaged 56.6⫾0.9%
from the maximal dilated diameter. In these vessels, SOD
incubation also uncovered oscillating flow-induced dilation of
18.8⫾2.6% of maximal diameter at a peak-peak shear of 30
dyne/cm2.
Discussion
This study presents several novel findings. First, dilation in
response to steady flow occurred not only in epicardial
arterioles but also in endocardial vessels, which responded to
a greater degree than did the epicardial vessels. The dilation
was NO dependent, being completely blocked by L-NNA.
Second, oscillating flow without a net forward component did
not cause flow-dependent dilation. However, in both EPI and
ENDO vessels, dilation to oscillating flow did occur after
incubation with superoxide dismutase. In fact, in the presence
of SOD, ENDO vessels dilated to a greater degree in response
to oscillating flow than to steady flow. This dilation was
partially mediated by NO; the remaining dilation occurred via
a mechanism not related to H2O2, because catalase was
without effect. Third, DPI, an NAD(P)H oxidase inhibitor,
also uncovered a dilation to oscillating flow in ENDO
vessels. Fourth, the delay time for steady flow-induced
dilation decreased in the presence of SOD in both ENDO and
EPI vessels.
Sorop et al
Oscillating Flow Effects in Coronary Arterioles
1349
Downloaded from http://circres.ahajournals.org/ by guest on June 15, 2017
Figure 6. A, Delay time between onset of steady flow and 25%,
50%, and 75% of the total dilation development was independent of shear. B, Five individual experiments showing the delay
time to 25% dilation. Delay time in response to steady flow
onset decreased significantly in the presence of SOD.
Methodology
The possibility to independently manipulate left and right
cannula pressure by computers allowed us to apply flow
wave-forms as would occur in the beating myocardium at
constant intraluminal pressure. We are confident that luminal
pressure was constant, capacitive flow was absent, and flow
was accurately predicted from cannula pressures and cannula
resistances in the current study for the following reasons:
first, a sudden or oscillating change in luminal pressure would
have resulted in simultaneous distension of the vessel. Previous studies3 demonstrated that pressure variations of less
than 5 mm Hg can be detected in this way. No such changes
in diameter were observed even though gradients in cannula
pressures in the order of 100 mm Hg were required to
generate the highest shear stresses. Second, in previous
experiments at zero flow, intraluminal pressure as measured
by servo-null pipettes penetrated through the vessel wall
matched the cannula pressures, also for fast oscillations at
high amplitudes.3 Third, the low estimated Womersley number (see Materials and Methods) indicates a fully developed
parabolic velocity profile that instantaneously followed the
applied pressure gradient. Fourth, we found the predicted
centerline velocity to match the velocity of red blood cells in
separate experiments.
Basal tone in porcine coronary vessels of the size we
studied is variable and generally quite shallow. Therefore, we
preconstricted vessels by U46619. We do not believe that this
preconstrictor is involved in the diverging effects of steady
Figure 7. A, In 4 ENDO vessels, 10 ␮mol/L LNNA inhibited by
approximately 50% the oscillating flow-induced dilation in the
presence of SOD (P⬍0.01, ANOVA, L-NNA vs SOD). B, In 5
ENDO vessels, catalase did not affect this dilation (P⫽NS,
ANOVA, catalase versus SOD). C, In 5 ENDO, DPI also uncovered the dilation to oscillating flow (P⬍0.01, ANOVA, DPI vs
control).
and oscillating flow, because the cremaster arterioles, as well
as the few coronary arterioles below 100 ␮m that we were
able to cannulate, which did have basal tone, responded in the
same way to the various flow patterns. The applied shears
included clearly superphysiological levels, in order to dem-
Figure 8. Effects of steady, oscillating, and modulated flow with
100% amplitude on 9 rat cremaster arterioles with basal tone.
Steady flow induced dilation, whereas oscillating flow did not.
100% modulation of flow also did not affect the dilation initially
induced by steady flow.
1350
Circulation Research
June 27, 2003
onstrate absence of oscillating flow-induced dilation over a
wide range. We do not believe that the lower shear responsiveness of EPI vessels relates to damage during dissection
because similar levels of preconstriction were obtained by 1
␮mol/L U46619 in ENDO and EPI vessels and our previous
studies3 showed similar concentration-response curves to
adenosine and bradykinin in both vessel types.
In the first experimental group of vessels studied, having
diameter around 150 to 180 ␮m, we found a half-maximal
effect of shear at approximately 20 dyne/cm2 (Figure 2).
Smaller arterioles (110 to 150 ␮m) were used in most of the
other experiments. In these vessels the dilation reached a
plateau at shears of 10 dyne/cm2. The observation that the
smaller vessels are more sensitive to shear is in accordance
with data of Kuo et al9 on epicardial arterioles, showing that
shear sensitivity is the largest in 89-␮m vessels, whereas
larger vessels were less sensitive to flow.
Downloaded from http://circres.ahajournals.org/ by guest on June 15, 2017
NO and O2ⴚ Production
It is now accepted that NO bioavailability within the vascular
wall is negatively influenced by O2⫺ production.24 Both
radicals react extremely rapidly to form the ONOO⫺ (peroxynitrite) anion. Their balance therefore is of major importance in vascular function. Our experiments indicate that
steady and oscillating flow differentially affect this balance.
We suggest that steady flow induced production of NO that
overwhelmed any possible concomitant O2⫺ production, resulting in NO-dependent dilation. In contrast, during oscillating flow, O2⫺ production fully inhibited bioavailability of the
simultaneously produced NO to the smooth muscle cells. In
the presence of extrinsic SOD or DPI, this inhibition was
reversed, resulting in oscillating flow-induced dilation. It is
not clear whether the negative balance during oscillating flow
is due to a higher O2⫺ production or a lower NO production.
Modulation of steady flow up to 200% (Figure 3) neither
reversed nor augmented the dilation. There are alternative
explanations for this finding. Thus, the effects of any putative
extra NO and O2⫺ production during such modulation could
have just canceled. Alternatively, the effect of dynamic flow
components may be different, depending on the presence or
absence of a net forward flow. Thus, superoxide production
by rhythmic flow may have been suppressed by the presence
of a net forward flow. However, it is not clear then why even
a 200% modulation of low shear (eg, changing from a steady
10 dyne/cm2 to shear oscillations between ⫺10 and 30
dyne/cm2; Figure 3A) did not augment the initial dilation
(which was not yet saturated; compare Figure 3B), whereas
oscillations with such amplitudes had clear effects after
removal of the superoxide (Figure 5, right).
In ENDO vessels at 30 dyne/cm2, the response to oscillating flow in the presence of SOD was even larger than that to
steady flow. In addition, part of the former response could not
be inhibited by L-NNA or catalase. This suggests the existence of still another factor produced by oscillating but not
steady flow. Whatever the nature of this factor, it is sensitive
to O2⫺ because no oscillating flow-induced dilation at all was
observed in the absence of SOD. Recently, it was shown that
H2O2 is a possible EDHF in human coronary microvessels.25
One could thus argue that the uncovered dilation to oscillat-
ing flow in the presence of SOD relates to H2O2. However, 80
or 1000 U/mL catalase was without effect, whereas 80 U/mL
almost fully suppressed the dilation to extrinsic H2O2.
We do not believe that steady flow, once developed,
caused large amounts of O2⫺ production, because SOD was
without any effect on the magnitude of steady flow-induced
dilation. However, the onset of steady flow represents a
transient by itself, and this might induce O2⫺ production.
There is a very consistent lag time that precedes the initiation
of vasodilation on onset of flow. The duration of the delay
ranges between 5 seconds in arterioles smaller than 100 ␮m8
to 40 seconds in large conduit arteries.26,27 In the present
study, we found a 20-second delay. On application of SOD, a
40% reduction in delay time was observed, supporting the
view that the onset caused O2⫺ production. The remaining
delay was similar for onset of steady and oscillating flow,
suggesting that similar mechanisms caused the dilation.
Indeed, L-NNA also substantially inhibited the oscillating
flow-induced dilation in the presence of SOD, indicating the
involvement of NO also here.
It was beyond the scope of this article to unambiguously
identify the source of O2⫺ production during oscillating flow
or quantify its production. Experiments on cultured endothelial cells that demonstrate superoxide production in response
to oscillating shear21 support our SOD and DPI data, which
together implicate NAD(P)H oxidase as a shear-induced O2⫺
source. Alternative sources may include xanthine oxidase,
cytochrome P450, or eNOS itself. In addition to its effect on
NAD(P)H oxidase, DPI may inhibit other flavin-containing
oxidases, including eNOS. Therefore, we cannot fully rule
out that eNOS could have contributed as an additional source
of oxygen-free radicals. However, this does not affect our
main thesis that superoxide production during oscillating
flow inactivates nitric oxide.
Using dihydroethidine (DHE), we attempted to perform
dynamic measurements of O2⫺ production in our cannulated
vessels during acute episodes of shear. We now believe that
this approach was unsuccessful because of a number of
complicating factors, including the effect of flow on the
transport rate of the dye to the cells of interest. Tyrosine
nitrosylation neither seemed a way to directly demonstrate
O2⫺ and subsequent ONOO⫺ because these nitrosylation
reactions require chronic models or at least substantial stimulus periods and are therefore not applicable to acute interventions such as studied here. Cytochrome c reduction has
been applied for the measurement of O2⫺. However, this
absorbance technique seems not sensitive enough to use on
perfused arterioles. Although our data strongly suggest a role
for O2⫺ in the effects we observed, future studies must be
directed at online, localized quantification of O2⫺ production
during episodes of acute shear.
Implications for Coronary Physiology
This study demonstrates that not only EPI8 but also ENDO
arterioles show steady flow-induced dilation. Moreover, we
found such dilation to fully depend on NO-dependent pathways. Whether these responses are relevant for coronary flow
regulation depends on in vivo shear stresses in the coronary
circulation. These are poorly defined especially in the suben-
Sorop et al
Downloaded from http://circres.ahajournals.org/ by guest on June 15, 2017
docardium. Stepp et al5 measured microvascular diameters
and velocities in canine epicardial vessels. Basal shear stress
averaged 10 dyne/cm2 in small arteries and ranged between
15 and 23 dyne/cm2 in vessels of 100 to 150 ␮m. This
coincides with the range of values causing submaximal
dilation in EPI vessels of this size. A concern however is that
smaller vessels respond to lower shears, as demonstrated by
Kuo et al9 and confirmed in the present study, whereas based
on coronary branching patterns,28 the in vivo shear stress on
average actually increases in smaller vessels.
The dynamic profiles of flow and shear along the coronary
microcirculation are not extensively quantitated. However,
based on the retrograde coronary flow in systole29 and
subendocardial velocity30 and diameter patterns,31 the highly
pulsatile nature of subendocardial flow is beyond doubt. Our
modulated flow data (Figure 3) make clear that in vitro such
pulsatility does not add an extra component to flowdependent dilation. In that sense, the pulsatility seems irrelevant for subendocardial flow regulation. However, the activity in vivo of eNOS, O2⫺-producing enzymes, and also
intrinsic SODs (notably extracellular SOD324) may differ
from that observed in vitro. Moreover, their activity is
influenced by coronary artery disease. The diverging actions
of purely oscillating and steady flow on the NO-O2⫺ balance
therefore leave room for quite substantial effects of flow
modulation under such conditions.
In conclusion, our data show that steady and oscillating
flow differentially modulate coronary vascular tone via
mechanisms that determine the balance between NO and O2⫺
production in the coronary circulation. Considering the pulsatile nature of subendocardial flow and the vulnerability of
this layer, interfering with this balance may provide future
means for improving subendocardial perfusion.
Acknowledgments
Oana Sorop was supported by grant 98.180 of the Netherlands Heart
Foundation. Terry Sweeney was supported by the University of
Scranton, the Netherlands Organization for Scientific Research grant
B94-188, and NIH grant 1F33 HL072548-01.
References
1. Spaan JA. Mechanical determinants of myocardial perfusion. Basic Res
Cardiol. 1995;90:89 –102.
2. Krams R, Sipkema P, Westerhof N. Varying elastance concept may
explain coronary systolic flow impediment. Am J Physiol. 1989;257:
H1471–H1479.
3. Sorop O, Spaan JA, VanBavel E. Pulsation-induced dilation of subendocardial and subepicardial arterioles: effect on vasodilator sensitivity. Am J
Physiol. 2002;282:H311–H319.
4. Sun D, Huang A, Koller A, Kaley G. Endothelial Kca channels mediate
flow-dependent dilation of arterioles of skeletal muscle and mesentery.
Microvasc Res. 2001;61:179 –186.
5. Stepp DW, Nishikawa Y, Chilian WM. Regulation of shear stress in the
canine coronary microcirculation. Circulation. 1999;100:1555–1561.
6. Shimoda LA, Norins NA, Madden JA. Responses to pulsatile flow in
piglet isolated cerebral arteries. Pediatr Res. 1998;43:514 –520.
7. Miura H, Wachtel RE, Liu Y, Loberiza FRJ, Saito T, Miura M, Gutterman DD. Flow-induced dilation of human coronary arterioles:
important role of Ca2⫹-activated K⫹ channels. Circulation. 2001;103:
1992–1998.
8. Kuo L, Davis MJ, Chilian WM. Endothelium-dependent, flow-induced
dilation of isolated coronary arterioles. Am J Physiol. 1990;259:
H1063–H1070.
Oscillating Flow Effects in Coronary Arterioles
1351
9. Kuo L, Davis MJ, Chilian WM. Longitudinal gradients for endotheliumdependent and -independent vascular responses in the coronary microcirculation. Circulation. 1995;92:518 –525.
10. Kuo L, Chilian WM, Davis MJ. Interaction of pressure- and flow-induced
responses in porcine coronary resistance vessels. Am J Physiol. 1991;261:
H1706 –H1715.
11. Koller A, Seyedi N, Gerritsen ME, Kaley G. EDRF released from microvascular endothelial cells dilates arterioles in vivo. Am J Physiol. 1991;
261:H128 –H133.
12. Koller A, Sun D, Huang A, Kaley G. Corelease of nitric oxide and
prostaglandins mediates flow-dependent dilation of rat gracilis muscle
arterioles. Am J Physiol. 1994;267:H326 –H332.
13. Kajimura M, Michel CC. Inhibition of effects of flow on potassium
permeability in single perfused frog mesenteric capillaries. J Physiol.
199;516:201–207.
14. Huang A, Sun D, Carroll MA, Jiang H, Smith CJ, Connetta JA, Falck JR,
Shesely EG, Koller A, Kaley G. EDHF mediates flow-induced dilation in
skeletal muscle arterioles of female eNOS-KO mice. Am J Physiol.
2001;280:H2462–H2469.
15. Henrion D, Dechaux E, Dowell FJ, Maclour J, Samuel JL, Levy BI,
Michel JB. Alteration of flow-induced dilatation in mesenteric resistance
arteries of L-NAME treated rats and its partial association with induction
of cyclo-oxygenase-2. Br J Pharmacol. 1997;121:83–90.
16. Fisher AB, Chien S, Barakat AI, Nerem RM. Endothelial cellular
response to altered shear stress. Am J Physiol. 2001;281:L529 –L533.
17. Zhao S, Suciu A, Ziegler T, Moore JEJ, Burki E, Meister JJ, Brunner HR.
Synergistic effects of fluid shear stress and cyclic circumferential stretch
on vascular endothelial cell morphology and cytoskeleton. Arterioscler
Thromb Vasc Biol. 1995;15:1781–1786.
18. Chappell DC, Varner SE, Nerem RM, Medford RM, Alexander RW.
Oscillatory shear stress stimulates adhesion molecule expression in
cultured human endothelium. Circ Res. 1998;82:532–539.
19. Silacci P, Formentin K, Bouzourene K, Daniel F, Brunner HR, Hayoz D.
Unidirectional and oscillatory shear stress differentially modulate NOS
III gene expression. Nitric Oxide. 2000;4:47–56.
20. Ziegler T, Silacci P, Harrison VJ, Hayoz D. Nitric oxide synthase
expression in endothelial cells exposed to mechanical forces. Hypertension. 1998;32:351–355.
21. De Keulenaer G, Chappell DC, Ishizaka N, Nerem RM, Alexander RW,
Griendling KK. Oscillatory and steady laminar shear stress differentially
affect human endothelial redox state: role of a superoxide-producing
NADH oxidase. Circ Res. 1998;82:1094 –1101.
22. Helmlinger G, Berk BC, Nerem RM. Calcium responses of endothelial
cell monolayers subjected to pulsatile and steady laminar flow differ.
Am J Physiol. 1995;269:C367–C375.
23. Loudon C, Tordesillas A. The use of the dimensionless Womersley
number to characterize the unsteady nature of internal flow. J Theor Biol.
1998;191:63–78.
24. Fukai T, Folz RJ, Landmesser U, Harrison DG. Extracellular superoxide
dismutase and cardiovascular disease. Cardiovasc Res. 2002;55:
239 –249.
25. Miura H, Bosnjak JJ, Ning G, Saito T, Miura M, Gutterman DD. Role for
hydrogen peroxide in flow-induced dilation of human coronary arterioles.
Circ Res. 2003;92:e31– e40.
26. Edwards JC, Ignarro LJ, Hyman AL, Kadowitz PJ. Relaxation of intrapulmonary artery and vein by nitrogen oxide– containing vasodilators and
cyclic GMP. J Pharmacol Exp Ther. 1984;228:33– 42.
27. Gyllenhammar H. Lucigenin chemiluminescence in the assessment of
neutrophil superoxide production. J Immunol Methods. 1987;97:
209 –213.
28. VanBavel E, Spaan JA. Branching patterns in the porcine coronary
arterial tree: estimation of flow heterogeneity. Circ Res. 1992;71:
1200 –1212.
29. Hoffman JI, Spaan JA. Pressure-flow relations in coronary circulation.
Physiol Rev. 1990;70:331–390.
30. Merkus D. Determinants of Coronary Blood Flow at Low Coronary
Arterial Pressure [dissertation]. Amsterdam, The Netherlands: University
of Amsterdam; 1998.
31. Yada T, Hiramatsu O, Kimura A, Tachibana H, Chiba Y, Lu S, Goto M,
Ogasawara Y, Tsujioka K, Kajiya F. Direct in vivo observation of subendocardial arteriolar response during reactive hyperemia. Circ Res.
1995;77:622– 631.
Effect of Steady Versus Oscillating Flow on Porcine Coronary Arterioles: Involvement of
NO and Superoxide Anion
Oana Sorop, Jos A.E. Spaan, Terrence E. Sweeney and Ed VanBavel
Downloaded from http://circres.ahajournals.org/ by guest on June 15, 2017
Circ Res. 2003;92:1344-1351; originally published online May 22, 2003;
doi: 10.1161/01.RES.0000078604.47063.2B
Circulation Research is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 2003 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7330. Online ISSN: 1524-4571
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circres.ahajournals.org/content/92/12/1344
Data Supplement (unedited) at:
http://circres.ahajournals.org/content/suppl/2003/06/24/92.12.1344.DC1
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Circulation Research can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further information
about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation Research is online at:
http://circres.ahajournals.org//subscriptions/
FIGURE LEGENDS FOR ON-LINE SUPPLEMENTS:
Supplement
Online
Figure 1:1: experiments validating the estimation of shear stress. Top: validation for
two vessels at steady shear. The x-axis shows shear stress as estimated from the applied
pressure gradients, cannula resistances and vascular diameter; the y-axes indicate the
shear stress based on the measured centerline velocity of a few red blood cells in the
perfusate, on the basis of a parabolic velocity profile. The identity line (x=y) is shown.
Middle: Estimated and measured peak-peak shear stress during sinusoidal (1.5 Hz)
oscillating flow without a net forward component. This plot indicates that flow
oscillations linearly follow pressure oscillations, without any damping at this frequency.
Bottom: Effect of modulating flow up to 200% on the average shear stress as determined
from centerline velocity. Pressure gradients were applied that should induce variations
around an average level of 30 dynes/cm2. The time-averaged shear stress based on red
blood cell velocity was close to this value. This figure indicates that large variations in
driving pressure do not cause unintended shifts in the average shear stress.
Supplement
Online
Figure 2:2: response to very low shear stress in 6 sub-endocardial arterioles, 123 ± 12
(SEM) µm in diameter, preconstricted by 1 µM U46619. In order to be able to apply this
low shear stress to the small vessels with sufficient accuracy, these vessels were
cannulated with pipettes having higher resistances than used in the main study.
Online Figure 1
Steady flow
Exp 1
Exp 2
60
2
(dynes/cm )
Measured shear
70
50
40
30
20
10
0
0
10
20
30
40
50
60
70
Estimated shear ( dynes/cm 2)
Oscillating flow
(dynes/cm )
Exp 1
Exp2
40
2
Measured shear
50
30
20
10
0
0
10
20
30
40
50
Estimated shear ( dynes/cm 2)
Modulated flow of 30 dynes/cm 2
(dynes/cm )
2
Measured shear
40
Exp 1
Exp 2
30
20
10
0
0
50
100
150
Pulsatility (%)
200
250
Online Figure 2
Normalized diameter
0.25
Normalized dilation
0.20
0.15
0.10
0.05
0.00
0
1
2
3
4
Initial shear ( dynes/cm2)
5
6