626
Brief Communication
Venous and Arterial Endothelia: Different Dilator
Abilities in Dog Vessels
Charles L. Seidel and James LaRochelle
Downloaded from http://circres.ahajournals.org/ by guest on June 17, 2017
It has been demonstrated by other investigators that endothelium-dependent vasodilators are more
effective on arterial tissue than on venous tissue. The purpose of this study was to determine if this was
due to a difference in the sensitivity of arterial and venous smooth muscle to the endothelial dilator
(EDRF) or to a difference in the ability of arterial and venous endothelia to release EDRF. To
differentiate between these two possibilities, an in vitro "sandwich" preparation was used in which the
mechanical response to endothelium-dependent dilators of a de-endothelialized vessel was determined
when "sandwiched" with an endothelialized vessel. Using dog femoral artery and saphenous vein, it
was determined that acetylcholine (ACh), the ionophore A23187, and thrombin were endotheliumdependent dilators of the femoral artery, but their dilatory ability was significantly less in the
saphenous vein. However, if the de-endothelialized saphenous vein was "sandwiched" with an
endothelialized femoral artery, both ACh and A23187 significantly relaxed the vein. No relaxation of
the de-endothelialized femoral artery occurred when it was "sandwiched" with an intact saphenous
vein. Sodium nitroprusside, thought to act by a mechanism similar to EDRF, relaxed equally the
saphenous vein and femoral artery. These observations suggest that the difference in responsiveness
between femoral arteries and saphenous veins to endothelium-dependent dilators is due more to
differences in the ability of their endothelia to release EDRF than to an inability of their smooth muscle
to respond to EDRF. (Circulation Research 1987;60:626-630)
I
n 1980, Furchgott and Zawadzki1 demonstrated
that the vasodilation produced by acetylcholine
(ACh) required the presence of the endothelium.
Since then, many reports have appeared describing 1)
the endothelial dependence of a large number of vasodilators2"4; 2) the antagonistic effect of the endothelium
on the action of vasoconstrictors5"7; and 3) its role in
the dilator response to increased flow.8 The nature of
the substance released by the endothelium that mediates this vasodilation is unknown 239 but has been
termed endothelial derived relaxing factor (EDRF).
DeMey and Vanhoutte10 and Vanhoutte and Miller"
surveyed the endothelial dependence of several vasodilators in a variety of arteries and veins from a dog.
They reported that ACh, ATP, thrombin, and arachidonic acid induced a sustained relaxation of arteries
that was dependent on the presence of the endothelium;
however, the veins exhibited only transient relaxations
that were endothelium dependent. This difference in
the magnitude of endothelium-dependent response to
these dilators could be due to a difference in the sensitivity of arterial and venous smooth muscle to the
released EDRF or to a difference in the ability of arteriFrom the Section of Cardiovascular Sciences, Departments of
Medicine and Physiology, Baylor College of Medicine, Houston,
Tex.
Research supported by National Institutes of Health grant
HL23815. Computational assistance was provided by the CLINFO
project, supported by the Division of Research Resources of the
National Institutes of Health under grant RR00350.
Address for reprints: Dr. C.L. Seidel, Section of Cardiovascular
Science, Department of Medicine, Baylor College of Medicine,
One Baylor Plaza, Houston, TX 77030.
Received May 2, 1986; accepted December 30, 1986.
al and venous endothelia to release EDRF. It was the
purpose of this work to differentiate between these two
possibilities. The results suggest that the difference
lies in the relative ability of arterial and venous endothelia to release EDRF in response to certain
vasodilators.
Materials and Methods
Adult mongrel dogs of either sex were killed with
sodium pentobarbital and the femoral arteries and saphenous veins removed. The vessels were placed in a
4° C salt solution (PSS, pH 8) of the following composition (mM): NaCl 132, KC1 4.7, MgSO 4 -7H 2 O 1.2,
NaHCO3 18, CaCl2 2, and glucose 5. While in this
solution, adhering fat and connective tissue were removed by careful dissection. The vessels were cut into
rings 5-8 mm long and then cut into open sheets.
Vessel sheets were suspended between 2 stainless
steel clips similar to those described by Herlihy and
Murphy12 and oriented in an isothermic (37° C) tissue
bath so that the circumference of the original ring was
the vertical dimension of the vessel sheet. The lower
clip was attached to an immovable support, while the
upper clip was attached to a Grass model FT03 isometric force transducer (Grass Instrument Co., Quincy,
Mass.). The transducer in turn was attached to a calibrated, movable stage by which vessel sheet length
could be adjusted. Each vessel sheet was stretched
to twice its unstretched length, which in preliminary experiments had been determined to be the optimal length for force development. The output of the
force transducer was displayed on a Grass polygraph
(model 7).
Seidel and LaRochelle
Venous and Arterial Endothelia
Downloaded from http://circres.ahajournals.org/ by guest on June 17, 2017
To determine the endothelial dependence of the response of the femoral artery and saphenous vein to
vasodilators, the endothelium was removed from some
preparations by rubbing. Rings of vessels were rolled
by one arm of a jeweler's forceps along the surface of a
paraffin block submerged in PSS. These rings were
then cut open and placed in the clips as described
above. Tissues were silver stained13 at the end of the
experiment to determine if the endothelium remained
intact in preparations that had not been intentionally
de-endothelialized, as well as to verify that the endothelium had been removed in those sheets that had
been rubbed.
Since the rubbing procedure had the potential of
damaging the underlying layers of smooth muscle
cells, we reasoned that if such damage occurred, the
maximum contractile response of de-endothelialized
preparations to norepinephrine (NE) would be less
than that of intact preparations. Comparison of the
maximum contractile response of these 2 types of preparations from the same vessel segment indicated that
no attenuation had occurred (Table 1). We concluded
that the rubbing procedure did not damage the smooth
muscle.
To discriminate between differences in arterial and
venous smooth muscle cells or in arterial and venous
endothelial cells, "sandwiches" of intact and de-endothelialized vessel sheets were prepared in a manner
similar to that described originally by Furchgott and
Zawadski.' A sheet of de-endothelialized vessel was
placed in the stainless steel clips as described above,
and a sheet of intact (endothelialized) vessel was
placed against it so that the two lumenal surfaces were
in contact. The intact vessel sheet was oriented so that
its circumferential length was perpendicular to that of
the de-endothelialized vessel sheet and attached by
four 3-mm-long hemoclips placed at each of the corners. Because the intact vessel was oriented perpendicular to the direction of force measurement and held
in place only at discrete points by the hemoclips, its
smooth muscle did not contribute to the force generated by the de-endothelialized preparation; however, its
Table 1. Norepinephrine (NE) ED50 and Maximum Response
of Vessel Preparations
Maximum NE
response
NE ED50
7
/i
(mN/mm2)
(10 M)
Preparation
S
28
6.8±0.9
s*
20
3.7±O.5t
10.5±1.2
11.8±1.6
S*/F
F
F*
F*/F
F*/S
17
29
12
18
3
5.2±0.7t
8.3±1.3
83 + 7
70 + 9
91 + 12
82 + 7
73+10
89 + 9
120 + 4
7.9+1.7
All values means + SEM. S = intact saphenous; F = intact
femoral; S* = de-endothelialized saphenous; F* = de-endothelialized femoral; S*/F, F*/F, and F*/S = various "sandwich" preparations in which force was measured on the de-endothelialized vessel.
tp<0.05 relative to endothelialized preparation.
627
endothelium could influence the contractile response
of the smooth muscle cells in the de-endothelialized
preparation. The ability of the intact vessel sheet to
contribute to the contractile response of the de-endothelialized vessel sheet was tested by comparing the
maximum contractile response of intact and "sandwich" preparations. No significant difference was detected (Table 1). In addition, the maximum response of
the de-endothelialized sheet before and after the intact
preparation had been cut away was also determined in
some preparations, and no change was observed (data
not included).
Vessel preparations were continuously exposed to
PSS bubbled with either 95% O 2 -5% CO 2 or 20%
oxygen-5% CO 2 (pH 7.4) and allowed to equilibrate
for at least 1 hour before the beginning of the experiment to allow the reestablishment of normal cellular
ion gradients. After this period, the preparations were
contracted at least 3 times by adding 30 mM KCl to the
bath solution to ensure that all preparations were responding reproducibly before the start of the experiment. The cumulative concentration-response relation
for NE was determined, the ED50 concentration calculated by probit analysis, and this concentration added
to the bath to produce a level of contractile activity.
The cumulative concentration-response relation for
ACh (Sigma Chemical Co., St. Louis, Mo.), A23187
(Calbiochem), thrombin (bovine thrombin, Sigma), or
sodium nitroprusside (Roche Laboratories, Nutley,
N.J.) was then determined. The response to different
concentrations of these agents was expressed as percent of the response to the ED50 concentration of NE
before the addition of the agent.
The statistical significance of the mean responses of
intact and de-endothelialized vessel preparations to a
given concentration of agent was tested by the Student's t test. For multiple comparisons, analysis of
variance was performed and the Duncan's Multiple
Range test used to compare individual means. In all
cases, a probability value less than or equal to 0.05 was
considered significant.
Results
Table 1 compares the ED50 values for and the maximum responses to NE in the various vessel preparations. No difference in maximum response was observed between the various preparations of saphenous
vein (S) or femoral artery (F). The ED50 values for deendothelialized preparations (S* and F*) were significantly reduced compared to intact preparations (S vs.
S* and F vs. F*). However, when "sandwich" preparations were examined, the ED50 values for the de-endothelialized vessel in the "sandwich" increased and
was significantly greater than the de-endothelialized
preparations (S* vs. S*/S; F* vs. F*/F or F*/S) and not
different from the intact preparation (S vs. S*/S; F vs.
F*/F or F*/S).
ACh and A23187 relaxed the femoral artery in a
concentration-dependent manner that required the
presence of the endothelium (Figure 1A). However,
ACh and A23187 produced minimal relaxation of the
Circulation Research
628
o
Vol 60, No 4, April 1987
-w •—
100
-A
10
Q
s*
50
UJ
O
m
0
Q
UJ
0.2
0.4
0.6
0.8
1
Thrombin Concentration (u/ml)
FIGURE 2. Effect ofthrombin on contractile response to EDso
concentration ofNE in endothelialized (F, S) and de-endothelialized (F*, S*) femoral arteries (F) and saphenous veins (S).
Values with * indicate significant difference from the de-endothelialized preparation. Error bars not included for clarity.
UJ
s
a
A23187 S
D——
UJ
Downloaded from http://circres.ahajournals.org/ by guest on June 17, 2017
A23187S-
-9
-8
-7
-6
ACh. or A23187 Cone. (logM)
FIGURE 1. Effect of ACh and A23187 on contractile response
to ED50 concentration ofNE in endothelialized (F, S) and deendothelialized (F*, S*) femoral arteries (F) and saphenous
veins (S). Values with * are significantly different from deendothelialized preparation; + indicates a significant difference from response to ACh in intact preparations. Error bars
not included for clarity.
saphenous vein, which was unaltered by removal of
the endothelium (Figure IB). At high concentrations,
ACh caused a further increase in tension development.
Thrombin, on the other hand, produced an endothelium-dependent dilation in both femoral artery and saphenous vein, but the magnitude of dilation was significantly less in the vein (Figure 2).
To determine if the observed difference in response
of the femoral artery and saphenous vein to ACh and
A23187 was due to a difference in their smooth muscle
or endothelial cells, the response of "sandwich" preparation was examined. Two kinds of "sandwiches"
were studied: 1) those in which an intact and de-endothelialized vein and artery were combined (heterologous), and 2) those in which intact and de-endothelialized tissues came from the same vessel type
(homologous). In either "sandwich", the contractile
response of the de-endothelialized preparation was recorded.
Figure 3 qualitatively illustrates typical mechanical
responses to ACh and A23187 of "sandwich" preparations and intact and de-endothelialized vessels. ACh
and A23187 were able to elicite relaxation in the saphenous vein only when it was "sandwiched" with an
intact femoral artery. The summary of experiments
using ACh are illustrated in Figure 4. Similar results
were obtained with A23187 (data not shown).
The homologous femoral "sandwich" (F*/F) relaxed
significantly when compared to the de-endothelialized
femoral (F*) but the heterologous "sandwich" (F*/S)
did not (Figure 4A). Unlike the intact saphenous vein
(Figure IB), the heterologous saphenous "sandwich"
(S*/F) relaxed to low concentrations of ACh while at
high concentrations the contractile effect was attenuated (Figure 4B).
EDRF is thought to act via an increase in tissue
cyclicguanosine 3',5' monophosphate (GMP) concentration.14 To determine if another cGMP dependent
vasodilator also had a differential effect on saphenous
veins and femoral arteries, the concentration-response
relation for sodium nitroprusside was determined on
endothelialized vessels. As illustrated in Figure 5A,
there was no significant difference between the re-
-j
F*/S
i
i
i
i
i
•
mm* ^ /"
I
J
^E
•
»
II
-8 -s -7
-e
N
\
V
-»
S*/F
S*/F
•t '
-8 -7
-6
-s| W
Removed
F
FIGURE 3. Typical polygraph traces of mechanical response
to ACh (top 6 traces) and A23187 (bottom trace) of various
vessel preparations contracted with EDso concentration ofNE.
Horizontal time scale equals 5 minutes for all traces; vasodilator concentration is given as — log M below each trace. Note
relaxation induced by ACh and A23187 of de-endothelialized
saphenous vein "sandwiched" with intactfemoral artery (S*IF,
last two traces) and absence of relaxation to ACh when opposite
"sandwich" is used (F*/S, 5th trace). In last trace, when F is
removed from S*, dilation effect of A23187 is removed.
Seidel and LaRochelle
150
125
LLJ
100
z
75
o
50
*
25
IO
o
LU
0
150
iev"—
ACh F«
ACh F«/F
ACh F
-Jh
B
125
LU 100
S
75
Q
UJ
ACh S«/F
50
Cr-
Downloaded from http://circres.ahajournals.org/ by guest on June 17, 2017
25
0
ACh F«/F
D-
-9
629
Venous and Arterial Endothelia
-8
-7
-6
ACh. Cone. (logM)
FIGURE 4. Comparison of the effect of ACh on contractile
response to ED50 concentration of NE of de-endothelialized
vessels (F*, S*) in various "sandwich" preparations (F*/S;
F*IF; S*/F). The * indicates values that are significantly different from the de-endothelialized preparation. Error bars not
included for clarity.
sponse of these 2 vessels to sodium nitroprusside.
Forstermann et al' 5 observed that the half-life of
EDRF was prolonged if the medium was equilibrated
with 20% rather than 95% O2. To determine if the
difference in dilator response of the saphenous vein
and femoral artery to ACh was due to the partial pressure of O 2 , the response to ACh was determined on
vessels equilibrated with 20% oxygen and 5% CO2
(Po2 = 140-145 mm Hg). Lowering the Po2 enhanced
the relaxing effect of ACh in both the femoral artery
and saphenous vein (Figure 5B), however, the differential effectiveness of ACh on the vein and artery was
not eliminated.
Discussion
The femoral artery relaxes when exposed to ACh
and A23187 and this relaxation requires the presence
of an intact endothelium (Figure 1A); however, the
saphenous vein exhibits minimal relaxation to ACh or
A23187 (Figure IB). These observations are similar to
those of DeMey and Vanhoutte.10 This difference in
responses between the femoral artery and saphenous
vein could be due to the inability of the saphenous vein
smooth muscle cells to respond to the EDRF released
by these agents from the saphenous endothelium and/
or an inability of the saphenous endothelium to release
EDRF. We reasoned that if the saphenous vein smooth
muscle cells were unresponsive to EDRF, they would
not relax when exposed to the EDRF released by femoral artery endothelial cells. If the saphenous vein
endothelium did not release an EDRF in response to
ACh or A23187, it would be incapable of relaxing a
de-endothelialized femoral artery. Using various vessel "sandwich" preparations, these possibilities were
tested.
The de-endothelialized saphenous vein, when
"sandwiched" with an intact femoral artery, relaxed at
low concentrations of ACh, and at high concentrations
the contractile response was attenuated (Figure 4B).
De-endothelialized femoral artery, however, did not
relax in response to ACh if "sandwiched" with an
intact saphenous vein (Figure 4A). The ability of the
saphenous vein and femoral artery to relax similarly
with sodium nitroprusside (Figure 5A), thought to act
by increasing cell cGMP content analogous to EDRF, 14
indicates that the attenuated EDRF effect in the vein is
not due to a general inability of its smooth muscle cells
to respond to cGMP mediated dilators. Finally, lowering the oxygen partial pressure of the PSS from 722
mm Hg (95% O2) to 140-145 mm Hg in an attempt to
increase the half-life of EDRF15 did not remove the
differential effect of ACh on saphenous veins and femoral arteries (Figure 5B). These studies, therefore,
suggest that the saphenous vein smooth muscle cells
are capable of relaxing in response to EDRF or other
cGMP-dependent dilators, but that their endothelial
cells are either incapable of releasing EDRF or do not
release sufficient EDRF to cause relaxation.
In contrast to ACh and A23817, thrombin relaxed
the saphenous vein in an endothelium-dependent manner (Figure 2); however, the magnitude of relaxation
was significantly less than in the artery. This observation indicates that the saphenous vein endothelium is
100
r
o
75-
10
O
SO
u
*
25
-9
-8
-7
-6
Na Nitro. Cone. (logM)
-9 -8
-7
-6
ACh Cone. (logM)
FIGURE 5. Panel A: Effect of sodium nitroprusside on contractile response to ED50 concentration ofNE ofendothelialized
femoral arteries (F) and saphenous veins (S). Error bars not
included for clarity (n = 6). Panel B: Effect of ACh on contractile response to EDS0 concentration of NE in endothelialized
femoral arteries and saphenous veins when preparations were
equilibrated with either 95% or 20% oxygen. *indicates a significant effect of lower oxygen. Error bars not included for
clarity (n = 6).
630
Downloaded from http://circres.ahajournals.org/ by guest on June 17, 2017
capable of releasing sufficient EDRF to produce dilation in response to some endothelium-dependent vasodilators (e.g., thrombin) but that the response is still
less than that of the artery.
The observation that thrombin relaxed the saphenous vein is different from that reported by DeMey and
Vanhoutte9 who observed contraction. This difference
in response to thrombin may be due to the source of
thrombin used. Because commercial preparations of
thrombin are heterogenous, it is possible that different
vasoactive components may be present in different
preparations.
Removal of the endothelium from either the femoral
artery or the saphenous vein did not change the maximum response of the vessel (Table 1), which is similar
to the observation of Carrier and White16 for rat aorta.
Other investigators observed that de-endothelialization
increased the NE maximum response and suggested
that NE may stimulate the release of an endotheliumderived vasodilator. 56 DeMey and Vanhoutte, 10 on the
other hand, observed a significant reduction in maximum response when the endothelium was removed
from canine femoral arteries and saphenous veins. We
have no explanation for these divergent observations.
The NE ED50 concentration was reduced by de-endothelialization in both the saphenous vein and femoral artery (Table 1); however, "sandwiches" were not
different from intact preparations. Lues and Schumann6 and Carrier and White16 observed similar increases in sensitivity to a variety of a-adrenergic agonists following removal of the endothelium from rat
aorta. The explanation for this increased sensitivity is
unknown but could be due to 1) the rubbing procedure
damaging the endogenous nerve endings and affecting
their ability to inactivate NE in the de-endothelialized
vessel; 2) the endothelium serving as an uptake site for
NE16; 3) NE causing the release of an endotheliumdependent dialtor5-6; or 4) the spontaneous release of
EDRF from the endothelium. 917
In summary, contracted canine saphenous veins dilated less in response to ACh, A23187, and thrombin
than did femoral arteries. The data presented is consistent with the hypothesis that this difference is due to an
inability of saphenous vein endothelia to release sufficient EDRF to produce a similar magnitude of relaxation rather than that the saphenous smooth muscle
cells are insensitive to EDRF.
Circulation Research
Vol 60, No 4, April 1987
References
1. Furchgott RF, Zawadzki JV: The obligatory role of endothelial
cells in the relaxation of arterial smooth muscle by acetylcholine. Nature 1980;288:373-376
2. Furchgott RF: Role of endothelium in responses of vascular
smooth muscle. Circ Res 1983;53:557-573
3. Peach MJ, Singer HA, Loeb AL: Mechanisms of endotheliumdependent vascular smooth muscle relaxation. Biochem Pharmacol 1985;34:1867-1874
4. Furchgott RF: The role of endothelium in the responses of
vascular smooth muscle to drugs. Ann Rev Pharmacol Toxicol
1984;24:175-197
5. Cocks TM, Angus JA: Endothelium-dependent relaxation of
coronary arteries by noradrenaline and serotonin. Nature
1983;305:627-630
6. Lues I, Schumann H-J: Effect of removing the endothelial cells
on the reactivity of rat aortic segments to different a-adrenoceptor agonists. Naunyn Schmiedebergs Arch Pharmacol
1984;328:160-163
7. Egleme C, Godfraind T, Miller RC: Enhances responsiveness
of rat isolated aorta to clonidine after removal of the endothelial
cells. BrJ Pharmacol 1984;81:16-18
8. Pohl U, Holtz J, Busse R, Bassenge E: Crucial role of endothelium in the vasodilator response to increased flow in vivo.
Hypertension 1986;8:37-44
9. Griffith TM, Edwards DH, Lewis MJ, Newby AC, Henderson
AH: The nature of endothelium-derived vascular relaxant factor. Nature 1984;308:645-647
10. DeMey JC, Vanhoutte PM: Heterogeneous behavior of the
canine arterial and venous wall. Circ Res 1982;51:439-447
11. Vanhoutte PM, Miller VM: Heterogeneity of endotheliumdependent responses in mammalian blood vessels. J Cardiovasc Pharmacol 1985;7(suppl 3):S12-S223
12. Herlihy JT, Murphy RA: Length-tension relationship of
smooth muscle of the hog carotid artery. Circ Res
1973;33:275-283
13. Poole JCF, Sanders AG, Florey HW: The regeneration of the
aortic endothelium. J Pathol Bacteriol 1985;75:133—141
14. Rapoport RM, Draznin MD, Murad F: Endothelium-dependent vasodilator and nitrovasodilator induced relaxation may be
mediated through cyclic GMP formation and cyclic GMP dependent protein phosphorylation. Trans Assoc Am Physicians
1983;96:19-30
15. Forstermann U, Trogisch G, Busse R: Species-dependent differences in the nature of endothelium-derived vascular relaxing
factor. Eur J Pharmacol 1984;106:639-643
16. Carrier GO, White RE: Enhancement of alpha-1 and alpha-2
adrenergic agonist-induced vasoconstriction by removal of endothelium in rat aorta. J Pharmacol Exper Ther 1985;232:
682-687
17. Martin W, Villani GM, Jothianandan D, Furchgott RF: Selective blockade of endothelium-dependent and glyceryl trinitrate-induced relaxation by hemoglobin and by methylene blue
in the rabbit aorta. J Pharmacol Exp Ther 1985;232:708-716
KEY WORDS • endothelial derived relaxing factor • saphenous vein • femoral artery • relaxation
Venous and arterial endothelia: different dilator abilities in dog vessels.
C L Seidel and J LaRochelle
Downloaded from http://circres.ahajournals.org/ by guest on June 17, 2017
Circ Res. 1987;60:626-630
doi: 10.1161/01.RES.60.4.626
Circulation Research is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1987 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/60/4/626
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/