Subarachnoid Hemorrhage and Endothelial L-Arginine Pathway in

392

Subarachnoid Hemorrhage and EndothelialL-Arginine Pathway in Small Brain Stem

Arteries in Dogs

Zvonimir S. Katusic, MD, PhD; James H. Milde;Francesco Cosentino, MD; and Bora S. Mitrovic, MD

Background and Purpose: Experiments were designed to determine the effect of subarachnoid hemor-rhage on endothelium-dependent relaxations in small arteries of the brain stem. A "double-hemorrhage"canine model of the disease was used, and the presence of vasospasm in the basilar artery was confirmedby angiography.Methods: Secondary branches of both untreated basilar arteries (inner diameter, 324±11 ,um; n=12)

and arteries exposed to subarachnoid hemorrhage for 7 days (inner diameter, 328±12 jum; n=12) weredissected and mounted on glass microcannulas in organ chambers. Changes in the intraluminal diameterof pressurized arteries were measured using a video dimension analyzer.

Results: In untreated arteries, 10-11 to 10-7 M vasopressin, 1010 to 10-6 M bradykinin, and 10` to 10-6M calcium ionophore A23187 caused endothelium-dependent relaxations. At 10-6 and 3 x 10-4M the nitricoxide synthase inhibitor NG-nitro-L-arginine methyl ester (L-NAME) abolished relaxations to vasopressinand produced small but significant rightward shifts of the concentration-response curves to bradykininand A23187. At 10-3 M L-arginine prevented the inhibitory effect of L-NAME. Subarachnoid hemorrhageabolished relaxations to vasopressin but did not affect relaxations to bradykinin or A23187.

Conclusions: These studies suggest that in small arteries of the brain stem vasopressin causesrelaxations by activation of the endothelial L-arginine pathway. This mechanism of relaxation is selectivelyinhibited by subarachnoid hemorrhage. Preservation of endothelium-dependent relaxations to bradykininand A23187 is consistent with the concept that small arteries are resistant to vasospasm aftersubarachnoid hemorrhage. (Stroke 1993;24:392-399)KEY WoRDs * brain stem * nitric oxide * vasospasm * dogs

Ni itric oxide production in endothelial cells via theL-arginine pathway plays a key role in theregulation of cerebral arterial tone.' Activa-

tion of this pathway mediates basal and stimulatedproduction of nitric oxide and endothelium-dependentrelaxations in small as well as large cerebral arteries.2-6

Subarachnoid hemorrhage impairs endothelium-de-pendent relaxations in cerebral arteries.7-12 This maycontribute to the development of cerebral vasospasm inlarge arteries. By contrast, it has been suggested thatvasospasm does not occur in small arteries.13 The reason

for this differential effect of autologous blood on largeversus small cerebral arteries is unknown but maydepend on differences in the endothelial regulation ofsmooth muscle tone. The effect of subarachnoid hem-orrhage on relaxations mediated by the activation of

From the Departments of Anesthesiology (Z.S.K., J.H.M., F.C.,B.S.M.) and Pharmacology (Z.S.K.), Mayo Clinic, Rochester,Minn.Supported in part by US Public Health Service grants HL 44116

and GM 44486; by the American Heart Association, MinnesotaAffiliate; and by the Mayo Foundation.Address for reprints: Zvonimir S. Katusic, MD, PhD, Depart-

ment of Anesthesiology, Mayo Clinic, 200 First Street SW, Roch-ester, MN 55905.

Received January 14, 1992; final revision received October 12,1992; accepted November 9, 1992.

See Editorial Comment, page 399endothelial cells in small arteries of the brain stem hasnot been studied. Thus, the present study was designedto determine the role of the endothelial L-argininepathway in the reactivity of small brain stem arteries tovasopressin, bradykinin, and A23187 and to examinethe effect of subarachnoid hemorrhage on endothelium-dependent relaxations to these agonists.

Materials and MethodsAnimal Model of Subarachnoid Hemorrhage

Mongrel dogs of either sex weighing 14-18 kg wereused. During general anesthesia with 15 mg/kg thiopen-tal and 15-25 mg/kg i.v. pentobarbital and controlledventilation, a transfemoral angiogram of the basilarartery was obtained. Subsequently, the cisterna magnawas aseptically punctured with a No. 22 spinal needle,and 5 ml cerebrospinal fluid was removed. With theanimal in a 300 head-down position, 5 ml autologousvenous blood was injected through the spinal needleover 2 minutes. After 15 minutes in the head-downposition, the animal was returned to its cage. Two dayslater (on day 2), the injection of venous blood into thecisterna magna was repeated under general anesthesia.On day 7, angiography was performed, with the animals

by guest on March 28, 2018

http://stroke.ahajournals.org/D

ownloaded from

http://stroke.ahajournals.org/

Katugic et al Cerebral Vasospasm and Endothelium 393

FIGURE 1. Angiograms of ca-nine basilar artery before (leftpanel) and 1 week after (rightpanel) injections of autologousblood into cisterna magna. Notevasospasm after intracisternalinjections.

under anesthesia and controlled ventilation, to confirmthe presence of vasospasm'74 (Figure 1). During theangiography procedures, arterial blood gases were mon-

itored. The diameters of the basilar arteries were mea-

sured on the angiograms under optical magnification.The diameters before and after the intracisternal injec-tions of blood were compared, and the data are ex-

pressed as a percentage of the value before the injec-tions. The procedures and handling of the animals were

reviewed and approved by the Institutional AnimalCare and Use Committee of the Mayo Foundation.

2nd branch

Middlecerebral a.

2nd

FIGURE 2. Schematic representation of anatomy of caninebasilar artery and its secondary branches used in presentstudy. a., artery.

In Vitro StudiesThe brain was removed from dogs (untreated animals

or those exposed to subarachnoid hemorrhage) anesthe-tized with 30 mg/kg i.v. sodium pentobarbital andplaced into cold modified Krebs-Ringer bicarbonatesolution of the following millimolar composition (con-trol solution): 118.3 NaCI, 4.7 KCl, 2.5 CaC12, 1.2MgSO4, 25.0 NaHCO3, 0.026 calcium-ethylenediamine-tetraacetic acid, and 11.1 glucose. Segments 2-3 mmlong of a secondary branch of the basilar artery (Figure2) were carefully dissected using a dissection micro-scope. The arteries were transferred to an arteriographfilled with oxygenated (94%O 2 and 6% CO) controlsolution and then mounted onto microcannulas15 (Liv-ing System Instrumentation, Burlington, Vt.). Controlsolution circulated from a 250 ml oxygenated reservoirthrough the arteriograph chamber at a flow rate of 12ml/min. Temperature was continuously monitored(model 7000 H, Jenco Electronics) to maintain thevessel environment at 37±0.5°C. All experiments were

performed in the presence of 10` M indomethacin toprevent the activity of cyclooxygenase.The arteriograph was placed on the stage of an

inverted microscope (Diaphot-TMD, Nikon) that had avideo camera attached to the viewing tube. The signalderived from the video image of the vessel was pro-cessed by an electronic system (Living System Instru-mentation) for the continuous measurement and re-

cording of both inner diameter and wall thickness.

TABLE 1. Inner Diameters of Canine Brain Stem Arteries Withand Without Endothelium at 50 mm Hg

SubarachnoidUntreated hemorrhage

Endothelium With Without With

Control 324+11 268+10* 328+12Indomethacin 316+11 232+7* 308+12U46619 223-+- 13t 162 -+ 2*t 194-+- 1t

Values are mean+SEM gm; n-5-12 dogs.*p<0.05 different from untreated with endothelium.tp<0.05 different from control or indomethacin.

Iq

k,.1



ownloaded from


394 Stroke Vol 24, No 3 March 1993

TABLE 2. Inner Diameters of Canine Brain Stem Arteries inAbsence and Presence of L-NAME and L-Arginine Plus L-NAMEat 50 mm Hg

Drug(s) Absence PresenceL-NAME

10-4 M 395+17 363±173x10-4 M 346±11 324+14*

L-Arginine (10`3 M)+L-NAME (10-4 M) 378±28 377±32

L-NAME, NG-nitro-L-arginine methyl ester. Values are mean±SEM ,um; n=4 or 5 dogs.

*p<0.05 different from absence.

c0

aX

ax)Xmi

0-* With EndothellumA-A Without Endotheilum

In some experiments the endothelium was removed byintraluminal perfusion with 0.5% 3-[(3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate (CHAPS) for 30seconds.616'17 Removal of the endothelium was confirmedby the absence of relaxations to 10`6 M bradykinin.Responses of the pressurized arteries (in the absence

of intraluminal flow) were measured at a transmuralpressure of 50 mm Hg. This pressure was found to beoptimal for contractions of the small brain stem arteriesas assessed by repeated exposures to 3 x 10`8 M U46619at various transmural pressures.

After equilibration, the vessels were contracted with3 x 10` to 10-7M U46619 (Table 1). When a steady tonewas established vasopressin, bradykinin, or A23187 wasadded extraluminally in a cumulative fashion. Only oneconcentration-response curve was made per prepara-tion with endothelium or without endothelium. Therelaxations were expressed as a percentage of themaximal increase in intraluminal diameter obtained inresponse to 3 x 10-4 M papaverine.

DrugsThe following drugs were used: [Arg8]-vasopressin

(Sigma Chemical Co., St. Louis, Mo.), calcium iono-phore A23187 (Sigma), L-arginine hydrochloride (Sig-ma), bradykinin (Sigma), CHAPS (Sigma), 9,11-dide-oxy-9a,11a-methanoepoxyprostaglandin F2a (U46619;Cayman Chemical Co., Ann Arbor, Mich.), indometh-

10 9 8 7 6

Bradykinin (-log M)

FIGURE 4. Concentration-response curves to bradykinin incanine brain stem arteries with endothelium and withoutendothelium. Relaxations were obtained during contractionsto 3x10-8 to 10-7M U46619. Data are mean ±SEMpercent-age of maximal relaxation induced by papaverine. *p<0.05difference between preparations.

acin (Sigma), molsidomine (SIN-1; Cassela AG, Frank-furt, FRG), NG-nitro-L-arginine methyl ester (L-NAME; Sigma), and papaverine hydrochloride (Sigma).

Stock solutions of the drugs were prepared freshevery day. The stock solution of indomethacin wasprepared in an equal-molar concentration of Na2CO3.The stock solution of 10`6 M A23187 was prepared in1.5 x 10- M dimethyl sulfoxide. The drugs were dis-solved in distilled water. Concentrations of the drugsare expressed as the final molar concentration in thebath solution. The incubation periods were 30 minutesfor indomethacin and 15 minutes for L-NAME orL-arginine plus L-NAME. Indomethacin did not signif-icantly affect resting diameter (Table 1); 3 x 10`4 ML-NAME caused a significant reduction of resting di-ameter (Table 2).

Calculations and StatisticsData are given as mean±SEM. In each set of exper-

iments n equals the number of animals studied. Statis-

+20

-50

*-* With EndotheliumA-A Wlthout Endothellum

-100

11 10 9 8 7

Vasopressin (-log M)

FIGURE 3. Concentration-response curves to vasopressin in

canine brain stem arteries with endothelium and withoutendothelium. Relaxations were obtained during contractionsto 3 x10-8 to 10-7M U46619. Data are mean +SEMpercent-age of maximal relaxation induced by papaverine. *p<0.05difference between preparations.

La)a)Ea

a)

0)

C)-c

+50 0-0 With EndotheltumA-A Without Endothelium

9 8 7 6

A23187 (-log M)

FIGURE 5. Concentration-response curves to A23187 incanine brain stem arteries with endothelium and withoutendothelium. Relaxations were obtained during contractionsto 3 x 10-8 to 10-7M U46619. Data are mean+SEMpercent-

age of maximal relaxation induced by papaverine. *p<0.05difference between preparations.

c0

a)XlY



ownloaded from


Katusit et al Cerebral Vasospasm and Endothelium 395

TABLE 3. Effect of SIN-1 in Canine Brain Stem Arteries With and Without Endothelium Contracted With U46619

SIN-1 (-log M)Endothelium 9 8.5 8 7.5 7 6.5 6 5.5 5 4.5 4With 2+10 2±1 0+5 0+6 2±11 14±11 26±13 49+13 64+11 79±8 89±5Without 3±2 5±3 8±4 12±4 27±9 46±12 61±10 76+7 85±5 91±3 95±1

Values are mean±SEM % increase in diameter; n=6 dogs. U46619 concentration, 10-8 to 10-7 M.

tical evaluation was done by paired and unpaired Stu-dent's t tests. Means were considered significantlydifferent when the probability value was less than 0.05.

ResultsEffect of Endothelial Removal on Relaxations toVasopressin, Bradykinin, A23187, and SIN-iDuring contractions induced with 3 x 108 to 10` M

U46619, 10-11 to 10-7 M vasopressin, 10- 10 to 10-6 Mbradykinin, and 10-9 to 10`6 M A23187 caused concen-tration-dependent relaxations. Chemical removal of theendothelium abolished these relaxations (Figures 3-5).In arteries without endothelium, A23187 caused con-centration-dependent contractions (Figure 5). Removalof endothelial cells did not affect relaxations to SIN-1(Table 3).

Effect ofL-NAME on Relaxations to Vasopressin,Bradykinin, and A23187

Endothelium-dependent relaxations to vasopressinwere abolished in the presence of 10`4 and 3 x 10` ML-NAME (Figure 6). In contrast, L-NAME reducedrelaxations to bradykinin and A23187 but did not affectthe maximal relaxation to these agonists (Figures 7 and8). In the presence of 10-3 M L-arginine the inhibitoryeffects of L-NAME on relaxations to vasopressin andbradykinin were reversed (Figures 6 and 7). In ringswith endothelium, L-NAME (3x 10-4 M) did not affectrelaxations to papaverine (3 x 10'4M; 99+1%, n =5, and98+1%, n=5, in the absence and in the presence ofL-NAME, respectively).

AngiographyDogs with subarachnoid hemorrhage developed cere-

bral vasospasm (diameter of basilar artery 57+7% ofdiameter before intracisternal injections of blood, n =6;Figure 1).

Effect of Subarachnoid Hemorrhage onEndothelium-Dependent Relaxations to Vasopressin,Bradykinin, and A23187

In small arteries of the brain stem with endotheliumobtained from animals with developed vasospasm of thebasilar artery, relaxations to vasopressin were abolished(Figure 9). The relaxations to bradykinin and A23187were not affected except for those at 10-9 M bradykinin(Figures 10 and 11).

Effect of Subarachnoid Hemorrhage on Relaxationsto SIN-i

During contractions induced with 3 x 108 to 10` MU46619, 10-9 to 10-4 M SIN-1 caused concentration-dependent relaxations in arteries with endothelium.Subarachnoid hemorrhage did not affect the relaxationsto SIN-1 (Figure 12).

Effect of Subarachnoid Hemorrhage onArterial Wall Thickness

At an intra-arterial pressure of 50 mm Hg, wallthickness of the small brain stem arteries was notaffected by subarachnoid hemorrhage (28+ 2 gim, n= 12and 33+3 ,m, n=12 for untreated dogs and animalsreceiving intracisternal blood, respectively).

DiscussionThis is the first study to examine the effect of sub-

arachnoid hemorrhage on endothelium-dependent re-laxations in small brain stem arteries. The major newfinding is that subarachnoid hemorrhage selectivelyinhibits relaxations to vasopressin mediated by activa-tion of the endothelial L-arginine pathway.The relaxations to vasopressin, bradykinin, and

A23187 were abolished by chemical removal of endo-thelial cells, demonstrating that the relaxations to theseagonists are mediated by the production and release of

*-* ControlA-A L-NAME (10o4 M)*-* L-NAME (3x 10)4)

c0

aa3W._

*-* ControlA-A L-ARG (10 M)

+ 4L-NAME (10 M)


FIGURE 6. Concentration-re-sponse curves to vasopressin incanine brain stem arteries withendothelium in the absence (con-trol) and presence of N'~nitro-L-arginine methyl ester (L-NAMJE)(left panel) and in the absence(control) and presence of L-argi-nine (L-ARG) plus L-NAME(right panel). Relaxations wereobtained during contractions to3x10-8 to 10-7M U46619. Dataare mean±SEM percentage ofmaximal relaxation induced bypapaverine. *p<0.05 differencebetween control and L-NAME-treated arteries.

c0aa'F




ownloaded from


396 Stroke Vol 24, No 3 March 1993

*-* ControlA-, L-NAME (1064 Mi)*-* L-NAME (3x10 4 M) +20

axC) -500ct

-1 -100

Bradykinin (-log M)

*-* ControlA-A L-ARG (1°63 Mi)

+(

.4L-NAME (10 Mi)

9 8 7 E

Bradykinin (-log M)

FIGURE 7. Concentration-re-sponse curves to bradykinin incanine brain stem arteries withendothelium in the absence(control) and presence of NG-nitro-L-arginine methyl ester (L-NAME) (left panel) and in theabsence (control) and presenceof L-arginine (L-ARG) plusL-NAME (right panel). Relax-ations were obtained during con-tractions to 3x10-8 to 10-!MU46619. Data are mean±SEMpercentage ofmaximal relaxationinduced by papaverine. *p<Q0.05difference in EC50 for prepara-tions (-logEC50=72±0.2, n=5and 6.8+0.2 n=5 for controland L -NAME, respectively).

endothelium-derived relaxing factor(s). Our results alsodemonstrate that in arteries without endothelium theresting diameter was significantly reduced, suggestingthat continuous production of endothelium-derived re-laxing factor(s) exerts an inhibitory effect on smoothmuscle cells under basal conditions. This explanation isconsistent with a significant reduction of the innerdiameter induced by the nitric oxide synthase inhibitorL-NAME in arteries with endothelium.The reactivity of smooth muscle cells to a prostaglan-

din H2/thromboxane A2 receptor agonist (U46619), anitric oxide donor (SIN-1), or a phosphodiesteraseinhibitor (papaverine) appears to be preserved afterremoval of endothelial cells. Furthermore, in prepara-tions without endothelium, A23187 caused a concentra-tion-dependent decrease in diameter, demonstratingthat nonreceptor-mediated mechanisms of smooth mus-cle contraction are intact.The results of the present study confirm our previous

reports suggesting that the relaxations to vasopressin in

+20

0o

c

0

a

a._

x-50

canine basilar artery as well as in small arteries of thebrain stem are mediated by increased activity of nitricoxide synthase and production of nitric oxide fromL-arginine.36,18 In contrast, L-NAME produced onlysmall rightward shifts of the concentration-responsecurves to bradykinin and A23187 and did not affectmaximal relaxations to these agonists. Thus, it appearsthat the inhibitory effects of bradykinin and A23187 areprimarily due to increased production of relaxing fac-tor(s) other than nitric oxide. This conclusion is consis-tent with results reported from several different labora-tories. In isolated porcine coronary artery, endothelium-dependent relaxations to bradykinin are not inhibited bythe nitric oxide synthase inhibitor NG-monomethyl-L-arginine (L-NMMA).1920 Similarly, L-NMMA does notprevent endothelium-dependent relaxations to acetyl-choline or substance P in rabbit hind limb resistancearteries,21 suggesting that the L-arginine/nitric oxidepathway is not involved. Our experiments were per-formed in the presence of the cyclooxygenase inhibitor

+20

0

c0

C)

0

Q)

x

tY

-100 .. _ 1 ]

9 8 7 6

A23187 (-log M)

FIGURE 8. Concentration-response curves to A23187 incanine brain stem arteries with endothelium in the absence(control) and presence of NG-nitro-L-arginine methyl ester(L-NAME). Relaxations were obtained during contractions to3X10-8 to 10-7 MU46619. Data are mean+SEMpercentageof maximal relaxation induced by papaverine. *p<0.05 dif-ference between control and L-NAME-treated arteries.

-50

-10oo

10 9 8 7


FIGURE 9. Concentration-response curves to vasopressin inbrain stem arteries from untreated dogs (control) and ani-mals exposed to subarachnoid hemorrhage (SAH). Relax-ations were obtained during contractions to 3 x 10-8 to 10-7MU46619. Data are mean SEM percentage of maximal relax-ation induced by papaverine. *p<0.05 difference betweenpreparations.

+20

ca-i

D



ownloaded from


Katu&it et al Cerebral Vasospasm and Endothelium 397

+20

0

c0

CXam(DXiXi

-50

0-0 ControlA-A SAH

c0

0XGQ)

Q-

-100 [ _

10 9 8 7 6

Bradykinin (-log M)

FIGURE 10. Concentration-response curves to bradykinin inbrain stem arteries from untreated dogs (control) and animalsexposed to subarachnoid hemorrhage (SAH). Relaxations wereobtained during contractions to 3x10-8 to 10-7 M U46619.Data are mean±SEM percentage of maximal relaxation in-duced bypapaverine. *p<0.0odifference between preparations.

indomethacin, indicating that prostacyclin or oxygen-derived free radicals as possible mediators can be ruledout.2223 The precise mechanism of the relaxations tobradykinin and A23187 in small canine brain stem arter-ies is unknown, but it is likely that the componentresistant to inhibition by L-arginine analogues is medi-ated by a hyperpolarizing factor or some other unknownsubstance released from the endothelium.24

Angiographically we detected a significant decrease inthe diameter of the basilar artery after injections ofautologous blood into the cisterna magna. However, westudied the effect of subarachnoid hemorrhage on endo-thelium-dependent relaxations in secondary branches ofthe basilar artery that are below the limits of resolutionof angiography. Vasospasm of the basilar artery mayreduce dye entry into the more distal vessels, makingquestionable the value of angiography in quantifyingnarrowing in smaller arteries.25 Therefore, we quantifiedonly changes in diameter of the basilar artery.

+20

0

ci0

-50

-100

9 8 7

A23187 (-log M)

FIGURE 11. Concentration-response curves to A23187 inbrain stem arteries from untreated dogs (control) and ani-mals exposed to subarachnoid hemorrhage (SAH). Relax-ations were obtained during contractions to 3 X 10-8 to 10-7MU46619. Data are mean ±SEM percentage of maximal relax-ation induced by papaverine.

+20

0

-50

-100

9 5

SIN-1 (-log M)FIGURE 12. Concentration-response curves to molsidomine(SIN-i) in brain stem arteries from untreated dogs (control)and animals exposed to subarachnoid hemorrhage (SAH).Relaxations were obtained during contractions to 3x 108 to10'7M U46619. Data are mean +SEMpercentage ofmaximalrelaxation induced by papaverine.

The role of small arteries in the pathophysiology ofcerebral vasospasm during subarachnoid hemorrhagehas been studied in monkeys.26 The authors were able todetect spontaneous irregular increase in tone. However,since diameters of small arteries could not be estimatedfrom angiography, it remains unknown whether thisincreased contractility detected in vitro was translatedinto narrowing of these vessels in vivo.26 We have notdetected spontaneous increase in tone in isolated ca-nine small arteries exposed to SAH. This may be due tospecies differences, regional heterogeneity (cerebralversus brain stem arteries), or methodologies used toinduce SAH. Nevertheless, the results of the presentstudy are consistent with the concept that in smallarteries subarachnoid hemorrhage may lead to alter-ations in vascular reactivity without affecting the me-chanical properties of the tissue.25 This is supported bythe fact that optimal pressures for the reactivity ofuntreated and vasospastic arteries were identical. Inaddition, we did not detect any differences in wallthickness between the two preparations.

Subarachnoid hemorrhage abolished the endotheli-um-dependent relaxations to vasopressin, whereas therelaxations to bradykinin and A23187 were not affected(except for those at 10` M bradykinin). In contrast tosmall arteries, in the basilar artery the endothelium-dependent relaxations to bradykinin are abolished bysubarachnoid hemorrhage.27 Furthermore, the relax-ations to bradykinin in the basilar artery are mediatedby activation of endothelial nitric oxide synthase and areabolished by L-arginine analogues5,28 (F. Cosentino andZ.S. Katusi6, unpublished observation). The inhibitoryeffects of nitric oxide synthase inhibitors and subarach-noid hemorrhage on endothelium-dependent relax-ations strongly suggests that the development of cere-bral vasospasm is associated with inactivation of therelaxations mediated by the endothelial L-arginine path-way. However, the exact cellular mechanism of thesubarachnoid hemorrhage-induced inhibition of therelaxations to vasopressin remains unknown. Relax-ations to the nitric oxide donor SIN-1 were not inhibitedin arteries obtained from dogs with developed vaso-



ownloaded from


398 Stroke Vol 24 No 3 March 1993

spasm, suggesting that, unlike in basilar artery,29 vascu-lar smooth muscle cells have preserved reactivity tonitric oxide. This finding also suggests that the produc-tion, release, or transfer of nitric oxide rather than thereactivity of smooth muscle cells to nitric oxide may beimpaired by subarachnoid hemorrhage. We do not havean explanation for the differential effect of subarach-noid hemorrhage on reactivity of smooth muscle cells tonitric oxide in large versus small cerebral arteries.

It has been reported that 25% of a series of 42patients with subarachnoid hemorrhage had increasedconcentrations of vasopressin in either the plasma orcerebrospinal fluid.30 We did not measure circulatinglevels of vasopressin, and it is impossible to rule out thatprolonged exposure to high concentrations of vasopres-sin may contribute to alterations of V1-vasopressinergicreceptors present on endothelial cells.6

All experiments on arteries exposed to subarachnoidhemorrhage were performed in the presence of thecyclooxygenase inhibitor indomethacin. In our previousstudies we demonstrated that 10-5 M indomethacininhibits the production of prostanoids and prevents theformation of endothelium-derived contracting factor(s)generated by the cyclooxygenase pathway in the metab-olism of arachidonic acid.31 Thus, the inhibition ofendothelium-dependent relaxations to vasopressin bysubarachnoid hemorrhage cannot be explained by in-creased metabolism of arachidonic acid and release ofcontracting factor(s) from endothelial cells. On theother hand, the presence of indomethacin in our exper-iments imposes a limitation concerning the role ofcyclooxygenase in the mediation of endothelium-depen-dent responses. Further studies are needed to charac-terize the importance of endothelial arachidonic acidmetabolism in the regulation of smooth muscle tone insmall brain stem arteries.The results of the present study demonstrate that in

small canine brain stem arteries endothelium-dependentrelaxations to vasopressin are mediated by activation ofthe L-arginine pathway. This mechanism of relaxationappears to be selectively inhibited by subarachnoid hem-orrhage. Unlike in large cerebral arteries, the endotheli-um-dependent relaxations to bradykinin and A23187 arenot abolished by L-NAME or subarachnoid hemorrhage.The selective preservation of endothelial function mayhelp to explain the observation that cerebral vasospasmdoes not occur in small arteries.13

AcknowledgmentsThe authors would like to thank Leslie Phelps for technical

assistance, Rebecca Wilson and Robert Lorenz for preparingthe figures, and Janet Beckman for typing the manuscript. Wewould also like to thank Dr. Daniel Rufenacht, RichardWiener, William Gallagher, and Richard Koenig for help withangiography, intracisternal injections of blood, and arterialblood gas measurements, as well as Dr. Jurgen Reden ofCassella AG, FRG, for a generous supply of SIN-1.

References1. Moncada S, Palmer RMJ, Higgs EA: Nitric oxide: Physiology,

pathophysiology, and pharmacology. Pharmacol Rev 1991;43:109-142

2. Faraci FM: Role of nitric oxide in regulation of basilar artery tonein vivo. Am J Physiol 1990;259:H1216-H1221

3. Katusic ZS, Moncada S, Vanhoutte PM: Inhibitory effect ofN0-monomethyl-L-arginine on endothelium-dependent relaxationsto vasopressin, in Moncada S, Higgs EA (eds): Nitric Oxide From

L-Arginine: A Bioregulatory System. New York/Amsterdam, Else-vier Science Publishing Co, Inc, 1990, pp 69-72

4. Rosenblum WI, Nishimura H, Nelson GH: Endothelium-dependentL-Arg- and L-NMMA-sensitive mechanisms regulate tone of brainmicrovessels. Am J Physiol 1990;259:H1396-H1401

5. Mayhan WG: Impairment of endothelium-dependent dilatation ofbasilar artery during chronic hypertension. Am J Physiol 1990;259:H1455-H1462

6. Katusii ZS: Endothelial L-arginine pathway and regional cere-bral arterial reactivity to vasopressin. Am J Physiol 1992;262:H1557-H1562

7. Kim P, Sundt TM Jr, Vanhoutte PM: Alterations in endothelium-dependent responsiveness of the canine basilar artery after sub-arachnoid hemorrhage. J Neurosurg 1988;69:239-246

8. Kim P, Lorenz RR, Sundt TM Jr, Vanhoutte PM: Release ofendothelium-derived relaxing factor after subarachnoid hemor-rhage.JNeurosurg 1989;70:108-114

9. Nakagomi T, Kassell NF, Sasaki T, Fujiwara S, Lehman RM,Johshita H, Nazar GB, Torner JC: Effect of subarachnoid hemor-rhage on endothelium-dependent vasodilation. J Neurosurg 1987;66:915-923

10. Nakagomi T, Kassell NF, Sasaki T, Fujiwara S, Lehman RM,Torner JC: Impairment of endothelium-dependent vasodilationinduced by acetylcholine and adenosine triphosphate followingexperimental subarachnoid hemorrhage. Stroke 1987;18:482-489

11. Kanamuru K, Weir BKA, Findlay JM, Krueger CA, Cook DA:Pharmacological studies on relaxation of spastic primate cerebralarteries in subarachnoid hemorrhage. J Neurosurg 1989;71:909-915

12. Byrne JV, Griffith TM, Edwards DH, Harrison TJ, Johnston KR:Investigation of the vasoconstrictor action of subarachnoid haemo-globin in the pig cerebral circulation in vivo. BrJ Pharmacal 1989;97:669-674

13. Nihei H, Kassell NF, Dougherty DA, Sasaki T: Does vasospasmoccur in small pial arteries and arterioles of rabbits? Stroke 1991;22:1419-1425

14. Varsos VG, Liszczak TM, Han DH: Delayed cerebral vasospasm isnot reversible by aminophylline, nifedipine, or papaverine in a"two-hemorrhage" canine model. J Neurosurg 1983;58:11-17

15. Osol G, Halpern W: Spontaneous vasomotion in pressurized cere-bral arteries from genetically hypertensive rats. Am J Physiol 1988;254:H28-H33

16. Tesfamariam B, Halpern W: Endothelium-dependent and endo-thelium-independent vasodilation in resistance arteries fromhypertensive rats. Hypertension 1988;11:440-444

17. Diederich D, Yang Z, Buhler FR, Luscher TF: Impaired endothe-lium-dependent relaxations in hypertensive resistance arteriesinvolve cyclooxygenase pathway. Am J Phvsiol 1990;258:H445-H451

18. Cosentino F, Katusic ZS: NG-Nitro-L-arginine selectively inhibitsstimulated production of EDRF. (abstract) Circulation 1991;84(suppl II):II-653

19. Richard V, Tanner FC, Tschudi M, Luscher TF: Different activa-tion of L-arginine pathway by bradykinin, serotonin, and clonidinein coronary arteries. Am JPhysiol 1990;259:H1433-H1439

20. Cowan CL, Cohen RA: Two mechanisms mediate relaxation bybradykinin of pig coronary artery: NO-dependent and -indepen-dent responses. Am J Physiol 1991;261:H830-H835

21. Mugge A, Lopez JAG, Piegors DJ, Breese KR, Heistad DD:Acetylcholine-induced vasodilatation in rabbit hindlimb in vivo isnot inhibited by analogues of L-arginine. Am J Physiol 1991;260:H242-H247

22. Marshall JJ, Kontos HA: Endothelium-derived relaxing factors.Hypertension 1990;16:371-386

23. Yang S-T, Mayhan WG, Faraci FM, Heistad DD: Mechanisms ofimpaired endothelium-dependent cerebral vasodilatation in responseto bradykinin in hypertensive rats. Stroke 1991;22:1177-1182

24. Komori K, Vanhoutte PM: Endothelium-derived hyperpolarizingfactor. Blood Vessels 1990;27:238-245

25. Bevan JA, Bevan RD: Arterial wall changes in chronic cerebrova-sospasm: In vitro and in vivo pharmacological evidence. Annu RevPharmacol Toxicol 1988;28:311-329

26. Bevan JA, Bevan RD, Frazee JG: Functional arterial changes inchronic cerebrovasospasm in monkeys: An in vitro assessment ofthe contribution to arterial narrowing. Stroke 1987;18:472-481

27. Kim P, Lorenz RR, Sundt TM Jr, Vanhoutte PM: Release ofendothelium-derived relaxing factor after subarachnoid hemor-rhage. JNeurosurg 1989;70:108-114



ownloaded from


Katusit et al Cerebral Vasospasm and Endothelium 399

28. Katusi6 ZS, Marshall JJ, Kontos HA, Vanhoutte PM: Similarresponsiveness of smooth muscle of the canine basilar artery toEDRF and nitric oxide. Am J Physiol 1989;257:H1235-H1239

29. Kim P, Schini VB, Sundt TM Jr, Vanhoutte PM: Reduced pro-duction of cGMP underlies the loss of endothelium-dependentrelaxations in canine basilar artery after subarachnoid hemor-rhage. Circ Res 1992;70:248-256.

30. Mather HM, Ang V, Jenkins JS: Vasopressin in plasma and CSF ofpatients with subarachnoid haemorrhage. J Neurol Neurosurg Psy-chiatry 1981;44:216-219

31. Katusi6 ZS, Vanhoutte PM: Superoxide anion is an endotheli-um-derived contracting factor. Am J Physiol 1989;257:H33-H37

Editorial Comment

Endothelium plays an important role in regulation ofcerebral vascular tone through the production andrelease of endothelium-derived relaxing factor (EDRF),which has been identified as nitric oxide (NO) or anNO-containing compound.1-4 Impaired endothelium-dependent relaxation of large cerebral arteries aftersubarachnoid hemorrhage has been described in hu-mans and experimental animal models and may contrib-ute to vasospasm.5-10The present study by Katusic et al examined endo-

thelium-dependent relaxation of small arteries of thebrain stem in a canine model of subarachnoid hemor-rhage. Subarachnoid hemorrhage selectively impairedendothelium-dependent relaxation in response to vaso-pressin but not in response to bradykinin or the calciumionophore A23187. Because an inhibitor of NO-syn-thase and subarachnoid hemorrhage produced similarreductions in responses to vasopressin, these findingssuggest that subarachnoid hemorrhage selectively im-pairs the L-argininelNO (EDRF) pathway in endothe-lium. In contrast to large cerebral arteries and smallarteries taken from the surface of the brain, a recentstudy suggests that endothelium-dependent responsesare not impaired in penetrating arterioles after sub-arachnoid hemorrhage.11The mechanism that accounts for impairment of

endothelium-dependent relaxation in response to spe-cific agonists after subarachnoid hemorrhage is notclear. Impaired endothelium-dependent relaxation maybe due to abnormal receptor function, impaired produc-tion of EDRF, or simultaneous production of an endo-thelium-derived contracting factor. Impaired endothe-lial function was observed in the present experiments inthe presence of indomethacin, suggesting that forma-tion of a cyclooxygenase-derived contracting factor didnot contribute to abnormal responses. The finding thatrelaxation of arteries in response to an NO donor wasnormal suggests that impaired endothelium-dependent

responses were not due to reduced activity of solubleguanylate cyclase and formation of cyclic GMP inresponse to EDRF.10

Frank M. Faraci, PhD, Guest EditorDepartment of Internal Medicine

The University of Iowa College of MedicineIowa City, Iowa

References1. Faraci FM: Role of nitric oxide in regulation of basilar artery tone

in vivo. Am J Physiol 1990;259:H1216-H12212. Mayhan WG: Impairment of endothelium-dependent dilatation of

basilar artery during chronic hypertension. Am J Physiol 1990;259:H1455-H1462

3. Rosenblum WI, Nishimura H, Nelson GH: Endothelium-dependentL-arg- and L-NMMA-sensitive mechanisms regulate tone of brainmicrovessels. Am J Physiol 1990;259:H1396-H1401

4. Wei EP, Kukreja R, Kontos HA: Effects of inhibition of nitricoxide synthesis on cerebral vasodilation and endothelium-derivedrelaxing factor from acetylcholine. Stroke 1992;23:1623-1629

5. Edwards DH, Byrne JV, Griffith TM: The effect of chronic sub-arachnoid hemorrhage on basal endothelium-derived relaxing fac-tor activity in intrathecal cerebral arteries. J Neurosurg 1992;76:830-837

6. Hatake K, Wakabayashi I, Kakishita E, Hishida S: Impairment ofendothelium-dependent relaxation in human basilar artery aftersubarachnoid hemorrhage. Stroke 1992;23:1111-1117

7. Hongo K, Kassell NF, Nakagomi T, Sasaki T: Subarachnoid hem-orrhage inhibition of endothelium-derived relaxing factor in rabbitbasilar artery. J Neurosurg 1988;69:247-253

8. Kanamaru K, Weir BKA, Findlay JM, Krueger CA, Cook DA:Pharmacological studies on relaxation of spastic primate cerebralarteries in subarachnoid hemorrhage. JNeurosurg 1989;71:909-915

9. Kim P, Sundt TM, Vanhoutte PM: Alterations in endothelium-dependent responsiveness of the canine basilar artery after sub-arachnoid hemorrhage. J Neurosurg 1988;69:239-246

10. Kim P, Schini VB, Sundt TM, Vanhoutte PM: Reduced productionof cGMP underlies the loss of endothelium-dependent relaxationsin the canine basilar artery after subarachnoid hemorrhage. CircRes 1992;70:248-256

11. Vollmer DG, Takayasu M, Dacey RG: An in vitro comparativestudy of conducting vessels and penetrating arterioles after exper-imental subarachnoid hemorrhage in the rabbit. J Neurosurg 1992;77:113-119



ownloaded from


Z S Katusic, J H Milde, F Cosentino and B S Mitrovicarteries in dogs.

Subarachnoid hemorrhage and endothelial L-arginine pathway in small brain stem

Print ISSN: 0039-2499. Online ISSN: 1524-4628 Copyright © 1993 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Stroke doi: 10.1161/01.STR.24.3.392

1993;24:392-399Stroke.

http://stroke.ahajournals.org/content/24/3/392World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://stroke.ahajournals.org//subscriptions/

is online at: Stroke Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. Permissions and Rights Question and Answer process is available in the

Request Permissions in the middle column of the Web page under Services. Further information about thisOnce the online version of the published article for which permission is being requested is located, click

can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.Strokein Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:



ownloaded from

http://stroke.ahajournals.org/content/24/3/392

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://stroke.ahajournals.org//subscriptions/


Documents

Subarachnoid Hemorrhage and Endothelial L-Arginine Pathway in