Endothelial Dependent Relaxation Demonstrated In Vivo in ...

494 STROKE VOL 17, No 3, MAY-JUNE 1986

effects of vasoactive intestinal polypeptide. Amer J Physiol 238: tion and vascular effects of vasoactive intestinal polypeptide in
H449-H456, 1980 skeletal muscle of the cat. Cell Tiss Res 207: 55-64, 1980
49. Heistad DD, Marcus ML, Said SI, Gross PM: Effect of acetylcho- 55. Williams TJ: Vasoactive intestinal polypeptide is more potent than
line and vasoactive intestinal peptide on cerebral blood flow. Amer prostaglandin E2 as a vasodilator and oedema potentiator in rabbit
J Physiol 239: H73-H80, 1980 skin. Brit J Pharmacol 77: 505-509, 1982

50 Wei EP, Kontos HA, Said SI: Mechanism of action of vasoactive 56. Beed EA, O'Dorisio MS, O'Dorisio TM, Gaginella TS: Demon-
intestinal polypeptide on cerebral arteries. Amer J Physiol 239: stration of a functional receptor for vasoactive intestinal polypep-
H765-768, 1980 tide on Molt 4bt lymphoblasts. Regulatory Peptides 6: 1-12, 1983

51 Wilson DA, O'Neill JT, Said SI, Traystman RJ: Vasoactive intesti- 57. Undem BJ, Dick EC, Buckner CK: Inhibition by vasoactive intesti-
nal polypeptide and the canine cerebral circulation. Circ Res 48: nal polypeptide of antigen-induced histamine release from guinea-
138-148, 1981 pig minced lung. Eur J Pharmacol 88: 247-250, 1983

52. Traystman RJ, Kontos HA, Heistad DD: Vasodilator action of VIP 58. Moskowitz MA, Liu-Chen L-Y, Mayberg MR, Zervas NT, Han
on cerebral vessels. In Vasoactive Intestinal Peptide, ed. Said SI. DH, Brody MR: Cat pia-arachnoid contains immunoreactive sub-
Raven Press, pp. 161-168, 1982 stance P within neurons projecting from the fifth cranial nerve. In
Cerebrovascular Diseases, eds. Reivich M, Hurtig HI. Raven
53. Busija DW, Heistad DD: Effects of cholinergic nerves on cerebral Press, pp. 363-371, 1983
blood flow in cats. Circ Res 48: 62-69, 1981
59. Moskowitz MA: Neurobiology of vascular headaches. Ann Neurol
54. Jarhult J, Hellstrand P, Sundler F: Immunohistochemical localiza- 6: 157-168, 1984

Endothelial Dependent Relaxation Demonstrated
In Vivo in Cerebral Arterioles

WILLIAM I. ROSENBLUM, M.D.

SUMMARY The endothelium of mouse pial arterioles was injured in situ with a light/dye technique. The
response of the arterioles to acetylcholine or to bradykinin was compared before and after the injury. All
vessels failed to dilate after injury. In fact the predominant response now became constriction. The injured
vessels were still capable of dilating to papaverine. Uninjured vessels continued to dilate to acetylcholine or
bradykinin. The data show that relaxation of pial arterioles to acetylcholine or bradykinin is dependent on a
normal endothelium. This is in keeping with demonstrations by others that an endothelial dependent
relaxing factor or factors is(are) the mediator of the dilation to either acetylcholine or bradykinin. The
present demonstration of such endothelial dependence is important because in contrast with the bulk of the
literature it deals with in vivo, rather than in vitro data, and with microcirculation rather than large vessels.
It is also important because it concerns brain circulation. The data suggests that endothelial injury, known
to occur in a wide variety of pathologic states, could enhance vasospastic potential by eliminating dilating
influences and/or converting them to constricting forces.

Stroke Vol 17, No 3, 1986

SEVERAL YEARS AGO, Furchgott and coworkers1-2 mediated by EDRF depends upon the species of ani-
demonstrated that the relaxation of large arteries to mal, and the vascular bed.1"3
acetylcholine (ACh) was really mediated by some sub-
stance released from endothelium. The substance was Endothelial dependence was originally demonstrat-
called endothelial relaxing factor (EDRF). EDRF has ed in vitro by comparing responses of vessels with and
now been demonstrated by many workers.14 Its without endothelium. The endothelium was totally re-
chemical nature has not yet been identified, but it is moved by some mechanical abrasion. Some workers
extremely labile. '•5 Among candidates for EDRF have have questioned the physiologic significance of
been free radicals, carbonyl compounds, products of EDRF, at least insofar as ACh liberated from perivas-
lipoxygenase activity14 and substances produced by cular nerves, is concerned. They express doubt that
the action cytochrome P-450.6 In addition to ACh a ACh could diffuse all the way through adventitia and
number of vasoactive substances have also been found muscularis to initiate endothelial dependent relax-
to trigger EDRF. Use of pharmacologic probes sug- ation,3 or they suggest that the ACh passing through
gests that there may be more than one EDRF, and thus the vessel wall would first trigger an effect mediated
relaxation by two different endothelial dependent dila- by muscarinic receptors on the muscle itself.7 This
tors may be mediated by two different EDRF's.1-9 effect is constriction, not relaxation, and is observed in
However whether the actions of a vasoactive agent are arteries from which endothelium has been removed.'•2
However ACh applied externally to smaller vessels
From the Department of Pathology (Neuropathology), Medical Col- (arterioles) in situ, definitely produces relaxation.8-9
lege of Virginia, Richmond, Virginia 23298. Recently, endothelial injury of such arterioles in vivo
has been shown to result in conversion of the relax-
This work was supported by grant #HL-36091. ation to constriction.10
Address correspondence to: Dr. William I. Rosenblum, Box 17,
MCV Station, Richmond, Virginia 23298. Such in vivo studies of microvessels and endothelial
Received July 30, 1985; accepted September 20, 1985. dependent relaxation are rare. Confirmation of such

Downloaded from http://stroke.ahajournals.org/ by guest on January 29, 2016

ENDOTHELIAL DEPENDENT DILAJlOWRosenblum 495

data is extremely important since virtually all investi- greatly attenuate the onset of damage and aggregation.
gations concerning EDRF to date have occurred in If one interrupts the mercury light immediately after
vitro. Moreover, unlike the in vitro studies with larger aggregates appear in the venules, no aggregates will
vessels, the in vivo demonstration of EDRF in cerebral ever appear in the arterioles. We utilized this maneuver
arterioles10 employed a technique which injured but did to investigate the effects of endothelial injury in the
not remove large amounts of endothelium.12 We wish arterioles, without the complicating factor of platelet
to report a study in which the endothelium of cerebral aggregation in the arteriole. In order to prevent endo-
arterioles (pial arterioles) was also injured without thelial injury from progressing and aggregates from
cuasing denudation. Our data confirms the report of developing in the arteriole we switched, at this point in
Kontos10 showing dependence of ACh induced relax- each experiment, to a 100 Watt halogen lamp, which
ation upon a healthy endothelium in cerebral arte- illuminated the field via a fibreoptic light guide. A blue
rioles. In addition we show a similar dependence of the filter was interposed in the light path, and between
dilating response to bradykinin (BK). These are impor- periods of observation, the halogen lamp itself was
tant observations because acetylcholine and other dila- blacked out.
tors with potential dependence upon an EDRF, are
plentiful in brain and may be released in brain damage One further manipulation of the optical system was
which is frequently accompanied by endothelial in- performed. The initial measurements of diameter and
jury. Vasospasm may follow parenchymal injury and response to vasoactive agents were made with a 6.5 x
has heretofore been considered predominantly in terms Ultropak objective. However the damage was pro-
of the presence or absence of constrictor substances. duced with an 11 x objective and immersion attach-
The present data suggest an additional possibility — ment in place. The higher magnification focuses the
namely that constriction may be accentuated by loss of light on a smaller spot with resultant increase in light
dilating action from agents whose action depends upon intensity. With this system, platelet aggregates began
an intact endothelium. Indeed dilators like ACh may to form after approximately 6 minutes of continuous
not only lose their capacity to relax vessels after endo- exposure to mercury light following injection of sodi-
thelial injury, but may then become constrictors in um fluorescein. After 6 minutes the 11 X objective
their own right. was replaced with the 6.5 x objective and the halogen
lamp replaced the mercury.
Methods
The responses to vasoactive agents were monitored
Male mice, ICR strain (Dominion Labs, Dublin through the Ultropak microscope via its trinocular tube
VA) were used. The preparation has been described in and a TV camera and monitor using the Baez20 image
great detail.13"15 Briefly, a tracheostomy and crani- splitter. Approximately 15 minutes elapsed between
otomy were performed in mice anesthetized with ure- each determination of microvascular response. The 6
thane. The dura was stripped and the pial vessels ex- minute period of exposure to noxious light/dye was
posed and kept moist with a suffusate of artificial inserted into the 15 minute period between applica-
CSF,15-16 flowing at 2 ml per minute at 37°C, and pH of tions of the vasoactive agent. The vasoactive agent was
7.35.l5 The mice were maintained at 37°C. The field administered as a 1 ml bolus over a 30 second period in
was observed with a Leitz Ultropak microscope using artificial CSF at pH 7.35 and 37°C. The changes in
two different light sources at different times during an diameter were continuously recorded with the image
experiment. splitter and the maximal change was used in all calcu-
lations and in the tables.
Prior to the initiation of endothelial damage a 200
Watt mercury lamp with heat and UV filters, and a The agents used were acetylcholine chloride (ACh)
green filter was used. The light from this lamp and and bradykinin triacetate (BK) (80 )ttg/ml of each).
filters is innocuous and is the light used by us for many Papaverine HC1 was also used (320 /xg/ml). The inten-
years when studying responses of pial arterioles tion was to search for a qualitative difference in re-
(e.g.14- " ) . However if 0.2 ml of 2% sodium fluores- sponse to either ACh or BK following endothelial in-
cein is injected via tail vein, the combination of light jury (i.e. to see whether dilation was eliminated or
and dye damages the endothelium. This technique has converted to constriction). Papaverine served as a
been completely described in a number of reports check against "non specific" elimination of dilator re-
(e.g.17"19). The aggregates will fluoresce if a BG-12 sponses. Papaverine was chosen as a control because
(blue) filter is substituted for the green, and if appro- relaxation to papaverine is independent of endothelial
priate barrier filters18 are used. Endothelial damage damage and EDRF.2
occurs first in venules, later in arterioles, in the illumi-
nated field. Ultrastructural studies show that the dam- In each craniotomy site only a single arteriole was
age is initially manifest only by endothelial lucencies chosen for monitoring response to the dilator. This
and vacuoles, and progresses to denudation only if the selection was arbitrary, the only criterion being the
noxious illumination persists.19 No damage is ever size of the arteriole (30-45/A I.D.).
seen in the smooth muscle. Platelet aggregation begins
immediately after the initial appearance of the endo- Results
thelial lucencies and vacuoles.l9 By reducing the inten-
sity of the noxious light at the focal plane one can Table 1 shows that prior to endothelial damage ACh
relaxed the arterioles to 117 ± 4% (M ± SD) of their
resting diameter. Following damage of endothelium
ACh constricted the arterioles to 91 ± 6% of their

Downloaded from http://stroke.ahajournals.org/ by guest on January 29, 2016

496 STROKE VOL. 17, No 3, MAY-JUNE 1986

TABLE 1 Responses of Pial Arterioles to Acetylcholine Are Re- nally illuminated field) showed a normal response to
versed after Endothelial Injury BK (relaxation to 115 ± 6% of resting diameter).

Before After Discussion
Injury Injury
We have shown in this paper that acetylcholine
Baseline diameter (jx) 36±4 35 ± 8 (ACh) and bradykinin (BK) will not relax pial arte-
rioles in vivo, if the endothelium of these arterioles has
Response to ACh* 17±4t -9±6 been injured. Earlier studies19 showed that endotheli-
um injured with our light/dye technique displays only
All values are mean ± SD. N = 12 mice. minimal ultrastructural evidence of damage, prior to
onset of platelet aggregation at the injured site. This is
*80 /ig/ml. Response expressed as a % of baseline diameter. the set of circumstances which existed during the pres-
t( + ) = dilation; ( — ) = constriction. Prior to injury ACh dilat- ent study. It would appear that the endothelial depend-
ed arterioles to 117 ± 4% of baseline. After injury ACh constricted ent relaxing factor(s) required forrelaxationby ACh or
arterioles to 91 ± 6% of baseline; p < 0.001, paired ( test. BK can be eliminated without denuding endothelium.

resting diameter. By "resting diameter" we mean the Heretofore endothelial denudation has been the rule
diameter immediately preceding application of ACh. when EDRF has been eliminated especially in the usu-
This diameter was constant for at least 3 minutes of al in vitro demonstration of EDRF1"* but even in the
baseline measurements preceding each application of rare in vivo demonstration of EDRF." The only prior
ACh. Following endothelial damage ACh failed to exception to this use of denudation has been the work
dilate any of the 12 arterioles. Indeed 10 of the 12 of Kontos et al also using pial arterioles.'0 They used
constricted. cats and showed that acute elevations of blood pressure
eliminated the relaxation of pial arterioles by ACh.
A fifteen minute period was allowed for recovery They had previously shown that the acute pressure
from the ACh. Papaverine (320 /ig/ml) was then ap- elevation produced endothelial lesions which resemble
plied. Each of the 12 arterioles was dilated by the those produced by our light/dye technique, and did not
papaverine. The mean increase in diameter was 20 ± result in denudation.12
7%. Thus capacity to dilate was still present after en-
dothelial injury and had not been eliminated by dete- Kontos et al only investigated the effect of endothe-
rioration of the preparation or by some nonspecific lial injury on the response to ACh.10 We have not only
event occurring during the course of the experiment. confirmed the reversal of that response after injury of
pial arterioles, but have extended thisfindingto BK as
As an additional check against nonspecific alter- well. With respect to our own technique, we are unable
ations in the preparation we took four of the mice and to say whether the endothelium has been irreversibly
followed the measurement of response to papaverine injured in spite of the minimal ultrastructural damage.
with still another measurement of response to ACh. It may be that examination of endothelium at longer
But this time we selected an arteriole that had been intervals after insult would reveal cell death signifying
outside the microscopic field illuminated by the mer- irreversible injury during the original exposure to
cury light. These arterioles should not have suffered light/dye.
endothelial injury and should have displayed normal
responses to ACh. This prediction was confirmed. The Kontos et al have presented pharmacologic data in-
arterioles were relaxed to 113 ± 3% of their resting dicating that, in pial arterioles of the cat, the EDRF for
diameter by the ACh. ACh is different from the EDRF for BK.9 Our study
was not designed to investigate the nature of either
Table 2 displays the response to BK. These mice EDRF.
were different animals from those used to study ACh.
Prior to endothelial damage BK dilated the arterioles to Kontos et al9 were able to eliminate the dilation to
116 ± 3% (M ± SD) of resting diameter. Following BK by applying free radical scavengers to the vessels.
damage each of these five arterioles was constricted by Under such circumstances the vessels were unrespon-
BK. The constriction reduced arteriole diameter to 87 sive to BK, neither dilating nor constricting. In our
± 4% of baseline. study, the response to BK became one of constriction
following endothelial injury. Vasoconstriction by BK
Papaverine was still able to dilate each vessel, the is unusual but not unknown. For example, BK con-
overall dilation reaching 117 ± 6% of resting diame- stricts canine cerebral arterioles in vitro.21
ter. Examination of an undamaged arteriole in each
mouse (arteriole selected from those outside the origi- Both our data and that of Kontos et al10 shows con-
version of the response to ACh dilation to constriction
TABLE 2 Responses of Pial Arterioles to Bradykinin Are Re- following endothelial injury of pial arterioles. The
versed after Endothelial Injury constriction is thought to depend upon muscarinic re-
ceptors on vascular smooth muscle.1-3
Before After
Injury Injury Our data permit an expansion of the concepts con-
cerning cerebral vasospasm and regulation of cerebral
Baseline diameter 36±4 35±9 blood flow in pathologic states. Ischemia, subarach-
Response to BK* 16±3t -13±4 noid hemorrhage, head trauma, and hypertension may
each be associated with endothelial injury. Vasocon-
All values are mean ± SD. N = 5 mice.
*80 jtg/ml. Response expressed as a % of baseline diameter.
t( + ) = dilation; ( - ) = constriction Prior to injury BK dilated
arterioles to 116 ± 3% of baseline. After injury BK constricted
arterioles to 87 ± 4% of baseline; p < 0.001, paired / test.

Downloaded from http://stroke.ahajournals.org/ by guest on January 29, 2016

ENDOTHELIAL DEPENDENT DlLAT\OWRosenblum 497

striction could be enhanced in all these states by en- cular responses in mouse mesentery. Microcirc Endoth Lymph 2:
dogenous agents like ACh whose normal effect is re- 195-205, 1985
laxation but whose action will be a constricting one if 9. Kontos HA, Wei EP, Povlishock JT, Christman CW: Oxygen
endothelium is injured. Moreover, even a failure to radicals mediate the cerebral artenolar dilation from arachidonate
exert a relaxing influence may be sufficient to enhance and bradykinin in cats. Circ Res 55: 295-303, 1984
constriction in the presence of other constricting sub- 10. Kontos HA, Wei EP: Reversal of acetylcholine induced cerebral
stances. Enhanced vasoconstriction, even if it falls vasodilation after acute hypertension. Microvasc Res 29: 231,
short of "spasm" may be an unwanted response in the 1985
injured brain. Heretofore such responses and especial-
ly "spasm" have been viewed as consequences of an 11. Carvalho MH, Furchgott RF: Vasodilation of the rabbit mesenteric
outpouring of vasoconstrictors. Clearly this view must vascular bed by acetylcholine and A23187. The Pharmacologist
be modified to consider the consequences of lost relax- 22: 223, 1981
ing influences in brains with endothelial injury.
12. Kontos HA, Wei EP, Dietrich D, Navan RM, Povlishock JT,
References Ghatak NR, Ellis EF, Patterson JL Jr: Mechanism of cerebral
arteriolar abnormalities after acute hypertension. Am J Physiol
1. Furchgott RF, Cherry PD, Zawadiki JV, Jothianandan D: Endothe- 240: H511-H527, 1981
lial cells as mediators of vasodilation of arteries. J Cardiovasc
Pharmacol 6: 5336-5343, 1984 13. Rosenblum Wl, Zweifach BW: Cerebral microcirculation in the
mouse brain. Arch Neural 9: 414-^23, 1963
2. Furchgott RF: Role of endothelium in responses of vascular smooth
muscle. Circ Res S3: 557-573, 1983 14. Rosenblum Wl: Pial arteriolar responses in the mouse brain revisit-
ed. Stroke 7: 283-287, 1976
3. Vanhoutte PM, Cohen RA: Effects of acetylcholine on the coronary
artery. Fed Proc 43: 2878-2880, 1984 15. Rosenblum WI: Constriction of pial arterioles produced by prosta-
glandin F ^ . Stroke 6: 293-297, 1975
4. Singer HA, Peach MJ: Endothelial dependent relaxation of rabbit
aorta. J Pharmacol Exp Therap 227: 790-795, 1983 16. Elliott KAC, Jasper HH: Physiologic salt solutions for brain sur-
gery. J Neurosurg 6: 140-152, 1949
5. Griffith TM, Edwards DH, Lewis MJ, Newby AC, Henderson AH:
The nature of endothelium derived vascular relaxant factor. Nature 17. Rosenblum WI, El-Sabban F: Platelet aggregation in the cerebral
308: 645-647, 1984 microcirculation: Effects of aspirin and other agents. Circ Res 40:
320-328, 1977
6. Singer HA, Saye JA, Peach MJ: Effects of cytochrome P-450
inhibitors on endothelium dependent relaxation in rabbit aorta. 18. Rosenblum WI: Fluorescence induced in platelet aggregates as a
Blood Vessels 21: 223-230, 1984 guide to luminal contours in the presence of platelet aggregation
Microvasc Res 15: 103-106, 1978
7. Lee JFT: Cholinergic mechanism in the large cat cerebral artery.
Circ Res 50: 870-879, 1982 19. Povlishock JT, Rosenblum WI, Sholley MM, Wei EP: An ultras-
tructural analysis of endothelial change paralleling platelet aggre-
8. Rosenblum WI, Bryan D, Choi SC. Effect of diabetes on microvas- gation in a light/dye model of microvascular insult. Am J Path 110:
148-160, 1983

20. Baez S: Recording of microvascular dimensions with an image
splitter television microscope J Appl Physiol 21: 229-301, 1966

21. Toda N: Actions of bradykinin on isolated cerebral and peripheral
arteries. Am J Physiol 232: H267-H274, 1977

Downloaded from http://stroke.ahajournals.org/ by guest on January 29, 2016

Endothelial dependent relaxation demonstrated in vivo in cerebral arterioles.
W I Rosenblum

Stroke. 1986;17:494-497
doi: 10.1161/01.STR.17.3.494

Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1986 American Heart Association, Inc. All rights reserved.
Print ISSN: 0039-2499. Online ISSN: 1524-4628

The online version of this article, along with updated information and services, is located on the
World Wide Web at:

http://stroke.ahajournals.org/content/17/3/494

Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in
Stroke 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 Stroke is online at:
http://stroke.ahajournals.org//subscriptions/

Downloaded from http://stroke.ahajournals.org/ by guest on January 29, 2016