Tumor Necrosis Factor-  Mediates Hemolysis-Induced Vasoconstriction and the Cerebral Vasospasm Evoked by Subarachnoid Hemorrhage

Tumor Necrosis Factor-� Mediates Hemolysis-InducedVasoconstriction and the Cerebral Vasospasm Evoked by

Subarachnoid HemorrhageCarmine Vecchione, Alessandro Frati, Alba Di Pardo, Giuseppe Cifelli, Daniela Carnevale,

Maria Teresa Gentile, Rosa Carangi, Alessandro Landolfi, Pierluigi Carullo, Umberto Bettarini,Giovanna Antenucci, Giada Mascio, Carla Letizia Busceti, Antonella Notte, Angelo Maffei,

Gian Paolo Cantore, Giuseppe Lembo

Abstract—Hypertension can lead to subarachnoid hemorrhage and eventually to cerebral vasospasm. It has been suggestedthat the latter could be the result of oxidative stress and an inflammatory response evoked by subarachnoid hemorrhage.Because an unavoidable consequence of hemorrhage is lysis of red blood cells, we first tested the hypothesis on carotidarteries that the proinflammatory cytokine tumor necrosis factor-� contributes to vascular oxidative stress evoked byhemolysis. We observed that hemolysis induces a significant increase in tumor necrosis factor-� both in blood and invascular tissues, where it provokes Rac-1/NADPH oxidase–mediated oxidative stress and vasoconstriction. Further-more, we extended our observations to cerebral vessels, demonstrating that tumor necrosis factor-� triggered thismechanism on the basilar artery. Finally, in an in vivo model of subarachnoid hemorrhage obtained by the administrationof hemolyzed blood in the cisterna magna, we demonstrated, by high-resolution ultrasound analysis, that tumor necrosisfactor-� inhibition prevented and resolved acute cerebral vasoconstriction. Moreover, tumor necrosis factor-� inhibitionrescued the hemolysis-induced brain injury, evaluated with the method of 2,3,5-triphenyltetrazolium chloride and by thehistological analysis of pyknotic nuclei. In conclusion, our results demonstrate that tumor necrosis factor-� plays acrucial role in the onset of hemolysis-induced vascular injury and can be used as a novel target of the therapeutic strategyagainst cerebral vasospasm. (Hypertension. 2009;54:150-156.)

Key Words: cytokines � cerebrovascular disease � oxidant stress � inflammation � subarachnoid hemorrhage

Hypertension can lead to the rupture of a cerebral aneu-rysm, resulting in sustained subarachnoid hemorrhage

(SAH) and eventually cerebral vasospasm.1 This clinicalcondition is often lethal and provokes severe disability inmost survivors.2 Pathophysiologically, the presence of bloodin the subarachnoid space induces acute vasoconstriction,causing hypoperfusion of the surrounding tissues with dev-astating consequences for patients.3,4

A substantial amount of studies has been conducted in theeffort to predict and treat cerebral vasospasm.5,6 Severalpharmacological strategies have been proposed to increasecerebral blood flow in arteries susceptible to vasospasm.5–7

However, their therapeutic effects are limited. Their failurecould depend on our lack of knowledge about the pathophys-iological mechanisms involved in cerebral vasospasm. Clin-ical and experimental studies have clarified that brain injuryafter SAH is a biphasic event with an acute ischemic insult atthe time of initial bleed, followed by a generalized cerebral

vasospasm. This secondary event is related to the severity ofacute vasoconstriction.8,9 Therefore, targeting early patho-physiological mechanisms could interfere with the molecularcascade, leading to the perpetuation of cerebral vasospasmafter SAH.

When a cerebral aneurysm undergoes a rupture, bleedingand clot formation occur on the brain surface, where severalmajor blood vessels lay. The contact between blood and theextraluminal wall of arteries provokes vascular injury.10 Inparticular, blood extravasation induces erythrocyte degrada-tion, leading to hemoglobin release, which evokes vascularoxidative damage by activating NADPH oxidase.11 The latteris constituted by several subunits, which need to be assem-bled to exert full enzymatic activity. Among these subunits,there is a small G protein, Rac-1, that collects the transduc-tion of several intracellular signalings, converging them onNADPH oxidase activation.12,13 So far, there are no data onthe possible involvement of Rac-1 signaling in the vascularinjury evoked by hemolysis.

Received December 16, 2008; first decision January 8, 2009; revision accepted April 25, 2009.From the Departments of Angiocardioneurology (C.V., A.D.P., G.C., M.T.G., R.C., A.L., P.C., U.B., G.A., G.M., C.L.B., A.N., A.M., G.L.) and

Neurosurgery (A.F., G.P.C.), IRCCS Neuromed, Pozzilli (IS), Italy; Department of Cell Biology and Neurosciences (D.C.), Istituto Superiore di Sanita,Rome, Italy; and Department of Experimental Medicine (G.L.), Sapienza University, Rome, Italy.

Correspondence to Giuseppe Lembo, Sapienza University of Rome and Department of Angiocardioneurology, Neuromed Institute IRCCS, LocCamerelle, 86077 Pozzilli (IS), Italy. E-mail [email protected]

© 2009 American Heart Association, Inc.

Hypertension is available at http://hyper.ahajournals.org DOI: 10.1161/HYPERTENSIONAHA.108.128124

150 by guest on February 18, 2016http://hyper.ahajournals.org/Downloaded from by guest on February 18, 2016http://hyper.ahajournals.org/Downloaded from by guest on February 18, 2016http://hyper.ahajournals.org/Downloaded from by guest on February 18, 2016http://hyper.ahajournals.org/Downloaded from by guest on February 18, 2016http://hyper.ahajournals.org/Downloaded from by guest on February 18, 2016http://hyper.ahajournals.org/Downloaded from by guest on February 18, 2016http://hyper.ahajournals.org/Downloaded from by guest on February 18, 2016http://hyper.ahajournals.org/Downloaded from by guest on February 18, 2016http://hyper.ahajournals.org/Downloaded from by guest on February 18, 2016http://hyper.ahajournals.org/Downloaded from

On the other hand, free hemoglobin represents a proinflam-matory stimulus that promotes the accumulation of oxygenradicals and upregulates the expression of endothelial andleukocyte adhesion molecules, thereby recruiting macro-phages and neutrophils to the site of hemorrhage.14,15 Thisinflammatory response has been suggested to play a role inthe onset of cerebral vasospasm after SAH.16 In this study, wetested the hypothesis that overproduction of an inflammatorytrigger, eg, tumor necrosis factor-� (TNF-�), contributes tovascular oxidative stress, which sustains ischemic braininjury after SAH. The first aim of the present study was toinvestigate the mechanisms of hemolysis-induced vasocon-striction on murine carotid vessels, for which the size allowsthe analysis of intracellular signaling and its handling withgenetic probes. Then, we extended our studies to cerebralvessels, eg, the basilar artery, to verify whether the moleculartargets identified on carotid vessels could also play a role inthe hemolysis-induced cerebral vasoconstriction. Finally, wetested the relevance of our findings in an in vivo murinemodel of SAH.17 In this last set of experiments, beyondhistological examination, we developed a novel ultrasoundimaging analysis, which allows a real-time in vivo evaluationof the onset and progression of cerebral vasospasm.

Materials and MethodsFor a detailed description, please see the online data supplementavailable at http://hyper.ahajournals.org.

ResultsHemolyzed Blood Induces VasoconstrictionThrough Oxidative StressTo dissect the molecular mechanisms underlying the vaso-constriction that occurs after contact with blood, we firstrealized experiments on isolated carotid arteries incubatedwith whole or hemolyzed blood. The addition of hemolyzedbut not whole blood induced a significant vasoconstriction(Figure 1A) as compared with whole blood. On the samevessels, hemolyzed blood induced an increased dihydro-ethidium (DHE) staining (Figure 1B), revealing an enhancedsuperoxide production. Furthermore, exposure to Tiron sig-nificantly blunted vasoconstriction and oxidative stress in-duced by hemolyzed blood (Figure 1A and 1B). K�-evoked

vascular contraction was not affected by Tiron exposure(maximum vasoconstriction: 816�16 versus 802�11 mg;n�10; P value not significant). Altogether, these data indi-cate that hemolyzed blood is able to induce vasoconstrictionthrough the activation of oxidative stress mechanisms.

Rac-1/NADPH Oxidase Pathway MediatesHemolysis-Induced Vascular Oxidative StressBecause Rac-1 can be involved in the intracellular signalingconverging on NADPH oxidase activation and, consequently,in vascular oxidative stress, we focused our attention on thisprotein. Interestingly, hemolyzed blood caused an increase inRac-1 activity as compared with whole blood (Figure S1A).Most important, selective inhibition of Rac-1 by a dominant-negative mutant (AdN17) significantly blunted the action ofhemolyzed blood on both oxidative stress and vasoconstrictionas compared with the vessels treated with an empty adenovirus(Ad0; Figure S1B and S1C). In contrast, K�-evoked vasocon-striction was unaffected by AdN17 (maximum vasoconstriction:770�17 versus 781�11 mg; n�5; P value not significant).

Moreover, mice with a genetic deletion of p47phox, acytoplasmic subunit of the NADPH oxidase, were resistant tohemolysis-induced oxidative stress and vasoconstriction (Fig-ure S1D and S1E). K�-evoked vasoconstriction was compa-rable between wild-type and knockout mice (maximum va-soconstriction: 800�14 versus 793�13 mg; n�5; P value notsignificant). These results clearly demonstrate that the Rac-1/NADPH oxidase pathway plays a crucial role in thehemolysis-induced vascular oxidative stress.

TNF-� Is a Vasoconstrictive Cytokine and IsCrucial for Hemolysis-Induced Vascular InjuryIt is well known that oxidative stress is generated duringinflammatory processes. In this study, we demonstratedboth by mRNA transcription and protein expression thathemolysis evoked a marked increase in TNF-� both inblood (Figure 2A and 2B) and in vessels (Figure 2C),suggesting that this inflammatory cytokine could play arole in the genesis of oxidative stress and increased vasculartone in vessels exposed to hemolyzed blood. To verify thishypothesis, we analyzed TNF-� effects on isolated vessels.

0

10

20

30

40

50

60

DH

E Fl

uore

scen

ce (F

AU

)

B

0

10

20

30

40

50

60

A

Forc

e (%

of K

)+

WB HB HB+Tiron

*

#

WB HB HB+Tiron

*

#

Figure 1. Hemolyzed blood (HB) induces vaso-contriction and oxidative stress. A, Vascularresponse to whole blood (WB) or to hemolyzedblood alone and in the presence of an antioxi-dant agent (Tiron). B, Oxidative stress in carotidarteries, evaluated as DHE fluorescence. Repre-sentative images and quantification (n�5).*P�0.01 vs WB; #P�0.01 vs HB alone.

Vecchione et al TNF-� and Cerebral Vasospasm 151

by guest on February 18, 2016http://hyper.ahajournals.org/Downloaded from

TNF-� evoked a dose-dependent vasoconstriction (Figure2D), lower than that evoked by K� (� ID: 39�2 versus65�3; n�5; P�0.01).

Interestingly, TNF-� was also able to activate vascularRac-1, as shown by the increase in the Rac-1/p21-activatedkinase (PAK) complex (Figure 2E). To demonstrate therelevance of Rac-1 in the vasoconstriction induced by TNF-�,we selectively inhibited Rac-1, infecting the vessels withAdN17, either before the administration of TNF-� or after theonset of TNF-�–evoked vasoconstriction. AdN17 inhibitedTNF-�–induced vasoconstriction both before (Figure 2F) andafter (� ID: from 38�7 to 2�1 �m; P�0.01) TNF-�administration, whereas the use of an empty virus had noeffect (� ID: from 40�6 to 42�7 �m; P value not signifi-cant). These results clearly demonstrate that TNF-� is avasoconstrictor cytokine that realizes its effect through anintracellular signaling pathway involving Rac-1. Most impor-tant, blockade of TNF-� by a specific antibody blunted theeffects of hemolyzed blood on Rac-1 activation, oxidativestress, and vasoconstriction (Figure 3A through 3C). Incontrast, the vasoconstriction evoked by K� was unaffectedby the TNF-� antibody (maximum vasoconstriction: 784�18versus 773�16 mg; n�5; P value not significant). These datareveal the crucial role of the TNF-�/Rac-1 pathway in themolecular cascade converging on hemolysis-induced vascularcontraction.

TNF-� Mediates the Vasoconstrictive Effect ofHemolyzed Blood on Basilar ArteryTo verify whether the molecular target identified on carotidvessels, namely, TNF-�, also plays a role in the hemolysis-induced cerebral vasoconstriction, we performed experimentson basilar arteries. Also, in this experimental setting, wefound that hemolyzed blood induced a significant vasocon-striction as compared with whole blood (Figure S2A). More-over, hemolyzed blood induced an increased DHE fluores-cence (Figure S2B), demonstrating enhanced oxidative stress.The administration of Tiron significantly blunted vasoconstric-tion and oxidative stress induced by hemolyzed blood. Also incerebral vessels the administration of exogenous TNF-� evokeda dose-dependent vasoconstriction (Figure S2C).

To evaluate the role of NO in the vascular effects evokedby TNF-� and hemolyzed blood, we performed some exper-iments during the inhibition of NO obtained with NG-nitro-L-arginine methyl ester. Our results demonstrate that NG-nitro-L-arginine methyl ester exposure increased basalvascular tone (43�2 mg) but did not significantly affectTNF-� (maximum vasoconstriction: 80�3 versus 88�4 mg;n�5; P value not significant) and hemolyzed blood-inducedvasoconstriction (maximum vasoconstriction: 95�6 versus104�5 mg; n�5; P value not significant). Importantly,TNF-� inhibition blunted the effects of hemolyzed blood onvasoconstriction (Figure S2A). K�-evoked vascular contrac-

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

WB HB0

2

4

6

8

10

A B

0

2

4

6

8

10

LPS Basal HB

C

D

0

10

20

30

40

50

2.5 5.0 10.0 ng/ml Control TNFα

E

0

10

20

30

40

50

WB HB

F

Ad0 AdN17

TNFα

/β-a

ctin

TNFα

/β-a

ctin

∆ ID

(µM

)

Fold

Indu

ctio

n m

RN

AR

ac-1

Act

ivity

(fold

indu

ctio

n)

∆ ID

(µM

)

*

*

*

*

*

* *

*

TNFαβ-actin

TNFαβ-actin

Rac-1/PAKRac-1

Ctrl TNFα

TNFα

0

0.5

1.0

1.5

2.0

Figure 2. Hemolyzed blood (HB) increases the expression of TNF-�, which induces vasoconstriction via Rac-1 activation. A, Represen-tative blot and quantification of TNF-� protein expression in plasma from whole blood (WB) and HB (n�5). B, mRNA expression ofTNF-� in lymphocytes from WB and HB (*P�0.01 vs WB; n�9). C, TNF-� protein expression in carotid arteries in the basal conditionand in the presence of lipopolysaccharide or HB. Representative Western blot and quantification are shown (n�4; *P�0.01 vs basal).D, Dose-dependent vascular response to TNF-� (n�5; *P�0.01 vs basal). E, Rac-1 activity in carotid arteries treated (TNF-�) or not(control [CTRL]) with exogenous TNF-� protein (10 ng/mL). Representative Western blotting and quantification, corrected for total Rac-1protein (n�5). F, Vascular response to TNF-� (10 ng/mL) after infection with Ad0 or AdN17 (*P�0.02 vs Ad0; n�4).

152 Hypertension July 2009

by guest on February 18, 2016http://hyper.ahajournals.org/Downloaded from

tion was not affected by Tiron (maximum vasoconstriction:261�11 versus 256�10 mg; n�10; P value not significant)or TNF-� inhibition (maximum vasoconstriction: 269�8versus 275�8 mg; n�5; P value not significant). These dataindicate that TNF-� is involved in the action of hemolyzedblood on vascular tone also in cerebral arteries.

Inhibition of TNF-� Rescues Cerebral VasospasmWe evaluated the effects of TNF-� inhibition in a murinemodel of SAH obtained by injecting hemolyzed blood intothe cisterna magna. In this model, we observed a significantlumen narrowing of the basilar artery associated with thick-ening of the vascular wall and corrugation of the internalelastic lamina, as well as with increased oxidative stress(Figure 4A and 4B). In particular, mice exposed to SAHshowed a 50% reduction in cross-sectional area of the basilarartery as compared with control mice. Importantly, inhibitionof TNF-�, realized by infusion of a highly selective anti–TNF-� monoclonal antibody, prevented the phenotypicchanges observed in the basilar artery after injection ofhemolyzed blood (Figure 4A and 4B).

To monitor the onset and development of cerebral vaso-constriction in real time, changes in vascular diameter of theanterior cerebral artery (ACA) were also examined by high-resolution ultrasound analysis. In the murine model of SAH,an initial decrease in the ID of ACA was observed 30 minutesafter exposure to hemolyzed blood (data not shown). Suchphenomenon reached its maximal extension (46%) after �60minutes (Figure 5A) and remained evident for the entireobservation period (120 minutes). In contrast, in control mice,saline injection into the cisterna magna did not modify theACA diameter (data not shown). Interestingly, administrationof the anti–TNF-� antibody before the infusion of hemolyzedblood impaired the vasoconstriction of ACA (Figure 5B),whereas it did not exert any effect in control mice.Strikingly, the anti–TNF-� antibody was also able toresolve the already established ACA vasoconstriction in-duced by hemolysis (Figure 5C).

Finally, we evaluated ACA response and cerebral tissueviability after 2 and 5 days from the infusion of hemolyzedblood into cisterna magna in control conditions and duringTNF-� inhibition by either intra-arterial or intraperitonealadministration of an anti–TNF-� antibody. Importantly, thislate analysis demonstrated that early TNF-� inhibition is alsoable to rescue the ACA vasoconstriction observed at 2 (data

0

3

6

9

WB HB HB+Ab-TNFα

A

0

10

20

30

40

50

DH

E F

luor

esce

nce

(FA

U)

B

0

10

20

30

40

50

60

CR

ac-1

Act

ivity

(fold

indu

ctio

n)

WB HB HB+Ab-TNFα HB HB+Ab-TNFα

Forc

e (%

of K

)+

* *

# #

#

Rac-1/PAK

Rac-1

Figure 3. TNF-� mediates Rac-1 activation, oxidative stress, and vasoconstriction induced by hemolyzed blood (HB). A, Rac-1 activityin carotid arteries after the addition of whole blood (WB) or HB alone and in presence of anti–TNF-� antibody (Ab-TNF�). Representa-tive Western blotting and quantification, corrected for total Rac-1 protein (n�6). B, Representative high-power micrographs and quanti-fication of DHE dyeing in carotid arteries treated with WB or HB alone and in the presence of anti–TNF-� antibody (n�6). C, Vascularresponse in carotid arteries to HB alone and in the presence of anti–TNF-� antibody (n�6). *P�0.01 vs WB, #P�0.01 vs HB alone.

0

20

40

60

80

DH

E Fl

uore

scen

ce (F

AU

)

B

*

#

A

Control SAH SAH+infliximab

Control SAH SAH+infliximab

Figure 4. TNF-� inhibition blocks vasoconstriction and oxidativestress induced by hemolyzed blood (HB) in vivo. A, Immunohis-tochemical analysis and (B) representative high-power micro-graphs and quantification of DHE fluorescence of basilar arterysections from mice injected in the cisterna magna with saline(control; n�5) or HB pretreated with a nonimmune IgG (SAH;n�5) or with Infliximab (SAH�infliximab; n�6; �20 magnifica-tion; top right squares are at �100 magnification). *P�0.01 vswhole blood; #P�0.01 vs HB alone).

Vecchione et al TNF-� and Cerebral Vasospasm 153

by guest on February 18, 2016http://hyper.ahajournals.org/Downloaded from

not shown) and 5 days after the infusion of hemolyzed bloodinto the brain (Figure S3A). Moreover, brain 2,3,5-triphenyl-tetrazolium chloride staining, carried out 5 days after, showeda lighter positivity for this in brain sections from micesubjected to SAH in contrast with control mice. This markeddifference in 2,3,5-triphenyltetrazolium chloride stainingdemonstrated that mice subjected to SAH have a reducedbrain viability, as a consequence of generalized cerebralischemia. Strikingly, treatment with the anti–TNF-� antibodyrestored 2,3,5-triphenyltetrazolium chloride staining in brainsections from mice subjected to SAH (Figure S3B), thusindicating that early TNF-� inhibition is able to dramaticallyimprove the generalized cerebral ischemia and the impairedbrain viability induced by infusion of hemolyzed blood intothe cisterna magna. Furthermore, brains from SAH miceshowed pyknotic nuclei mainly localized in the posteriorcerebral cortex, indicating neuronal damage. Interestingly,infliximab pretreatment protected from neurodegeneration(Figure 6).

DiscussionIn this study, we demonstrated that TNF-� release was crucialfor hemolysis-induced cerebral vasospasm in a murine modelof SAH. More important, TNF-� inhibition not only pre-vented cerebral vasospasm, but was also able to resolvevasospasm when it was already established.

After the rupture of a cerebral aneurysm, blood does notremain fixed in the subarachnoid space but squeezes aroundmaking contact with the extraluminal wall of the arteries. Atthe same time, erythrocyte hemolysis and consequent releaseof oxyhemoglobin occur.10 In our hands, injection of hemo-lyzed blood in the cisterna magna proved that hemolysisrepresents a decisive step in inducing cerebral vasospasm.This effect can be reproduced easily in isolated cerebralvessels, where exposure to hemolyzed blood evokes a markedvasoconstriction. This methodological approach has allowedus to characterize intermediate vascular phenotypic changes

leading to vasoconstriction. In particular, our data demon-strate that the exposure of isolated cerebral vessels to hemo-lyzed blood induces a strong oxidative stress. This event iscrucial for hemolysis-induced vasoconstriction, because theuse of antioxidant agents rescues the hemolysis effect onvascular tone. Interestingly, both hypersensitivity to hydroxylradicals in the basilar artery and a decreased availability of NOin a subarachnoid space have been reported after SAH.18–20

However, our data show that the inhibition of NO synthesisdid not significantly modify the vasoconstrictor effect ofhemolyzed blood in isolated vessels, indicating that othermechanisms are involved in reactive oxygen species–medi-ated vasoconstriction. In agreement with our data, it has been

0.0

0.1

0.2

0.3

0.4

A

0.0

0.1

0.2

0.3

0.4

B

0.0

0.1

0.2

0.3

0.4

0 60 120 min

CID

(mm

)

ID (m

m)

ID (m

m)

0 60 120 min0 60 120 min

**

*

#

Figure 5. TNF-� inhibition can both prevent and resolve the vasospasm induced by hemolyzed blood (HB). A, Echographic analysis ofACA at 0, 60, and 120 minutes after injection of HB in cisterna magna (n�6) in mice pretreated with a nonimmune IgG; arrow indicatesACA. B, Echographic analysis of ACA at 0, 60, and 120 minutes after injection of HB in cisterna magna of mice pretreated with inflix-imab (n�5). C, Echographic analysis of ACA at 0, 60, and 120 minutes after injection of HB in cisterna magna of mice treated withinfliximab after the occurrence of vasospasm (60 minutes after injection of HB; n�5). Representative images and quantification of IDare shown (*P�0.01 vs 0 minutes; #P�0.01 vs 60 minutes).

0

10

20

30

40

50

60

70

SAH SAH+infliximab

B

n p

icno

tic n

ucle

i/mm

2

°#

A SAH SAH+infliximab

Figure 6. TNF-� inhibition prevents neurodegeneration inducedby SAH. A, Representative photomicrographs of hematoxylin-eosin staining of brain sections of nonimmune IgG (SAH; n�3)and infliximab-treated SAH mice (SAH�infliximab; n�4); �100magnification. Arrows indicate pycnotic nuclei. B, Quantificationof neuronal damage as number of pycnotic nuclei per millimetersquared (#P�0.05 vs SAH).

154 Hypertension July 2009

by guest on February 18, 2016http://hyper.ahajournals.org/Downloaded from

reported that reactive oxygen species can exert also a directvasoconstriction through modulation of calcium levels and/orarachidonic acid metabolism.18,21,22 Therefore, because oxi-dative stress appears to be a good intermediate phenotypetoward cerebral vasospasm, we have investigated the molec-ular mechanisms involved in the generation of vascularoxidative stress stimulated by hemolysis. We have focusedour attention on NADPH oxidase, largely expressed incerebral arteries.23,24 On this issue, a biphasic effect ofNADPH oxidase–induced reactive oxygen species generationon vascular tone has been observed. In particular, a lowamount of reactive oxygen species induces vasorelaxation,whereas high levels have been reported to induce vasocon-striction in cerebral arteries.23,24 NADPH oxidase is activatedduring SAH, thus causing an increase of superoxide forma-tion and impairing self-regulating vasodilation.24,25 Theseprevious data fully support our results showing that vesselswith genetic ablation of p47phox are protected fromhemolysis-evoked oxidative stress, clearly demonstrating aninvolvement of NADPH oxidase. Interestingly, hemolyzedblood is able to activate Rac-1. This activation is crucial foroxidative stress and vasoconstriction. Thus, Rac-1 intracellu-lar signaling is an important requisite for hemolysis-inducedvascular injury. These results are in agreement with previousstudies reporting that oxidative stress plays a significant rolein the development of acute brain injury and cerebral vaso-spasm after SAH.26,27 Oxidative stress is also a main compo-nent of inflammatory processes and is generated as a responseto several inflammatory cytokines. Among the latter, TNF-�can be considered a possible candidate for hemolysis-inducedvasoconstriction, because it has been reported that thiscytokine is augmented in the subarachnoid space after SAHand correlates with brain damage.28 However, no mechanisticrelationships have been reported to date.

Our data demonstrate increased TNF-� levels in hemo-lyzed blood. This result is supported by previous evidenceshowing an increased TNF-� release from circulating mac-rophages exposed to hemoglobin.15 More important, in thisstudy we demonstrated for the first time that hemolysis is alsoa stimulus for TNF-� release in cerebral vascular tissue,likely activating Toll-like receptor 4, which has been de-scribed to be overexpressed in the basilar artery afterSAH.29,30 These results strengthen the hypothesis that TNF-�could play a role in the onset of cerebral vasospasm. Thistheory finds a strong support in our evidence that TNF-� isable to evoke a direct vasoconstrictor effect on isolatedcerebral vessels through Rac-1 activation. Such evidenceextends previous observations showing that TNF-� impairsendothelium-dependent vasorelaxation.31 More important, theinhibition of TNF-� is able to counteract the effects ofhemolysis on Rac-1 activation, oxidative stress, and vasocon-striction, thus demonstrating that TNF-� is crucial for theabnormal cerebral vascular tone induced by hemolysis.

For this reason, we targeted TNF-� to verify its relevancein an in vivo model of SAH. Strikingly, our data demonstratethat the inhibition of TNF-� rescues the development of earlycerebral vasospasm evoked by the injection of hemolyzedblood into the cisterna magna. Our analysis was accom-plished not only by histological evaluation of structural

changes in the basilar artery, but also by a novel ultrasoundimaging technique that provides continuous real-time moni-toring of vascular tone of ACA by evaluating its changes inID. This analysis allows both temporal and spatial character-izations of cerebral vasospasm and can also be proposed toevaluate the effectiveness of novel therapeutic interventionson cerebral vascular tone. With this approach, we showed theonset and the development of the ACA vasospasm induced byhemolyzed blood for the first time. More importantly, wewere able to detect the beneficial effect of TNF-� inhibitionin the prevention of vasoconstriction. Excitingly, the block-ade of TNF-� activity is also able to resolve the earlyestablished cerebral vasospasm, thus revealing that the inhi-bition of this cytokine is important not only for the onset butalso for the perpetuation of hemolysis-induced cerebral va-soconstriction. This conclusion is strongly supported by ourlate analysis, focused on the brain damage accomplished 2and 5 days after SAH. In fact, the early inhibition of TNF-�is also able to counteract the chronic vasospasm, thus exertinga real protection for brain against ischemia. Therefore, ourstrategy focusing on the early phase of cerebral vasospasmhas allowed us to reveal a novel molecular mechanism thatcan be useful to more efficiently fight the long-term injuryoccurring after SAH. Therefore, the investigation of the earlyphase of cerebral vasoconstriction is not trivial, as depictedby the results of this study and other previous reports showingthat late brain injury is strictly related to early molecularevents.8,9 On the other hand, the fact that even in our study theinhibition of TNF-� leaves a slight residual cerebral hypo-perfusion at late phase reveals that other mechanisms canparticipate in the action of hemolyzed blood on brain injury,as depicted by previous reports.32–34

However, so far, the proposed therapies for cerebralvasospasm have mainly targeted the delayed phase of cere-bral vasospasm,35,36 and this approach could explain thefailure in the treatment of patients with cerebral vasospasm.In fact, oxyhemoglobin appears early after SAH,9 thus trig-gering the whole course of events flowing into cerebralvasoconstriction and hypoperfusion. On the other hand, theuse of a symptomatic treatment of cerebral vasospasm withvasodilators has several limitations, because the hypotensiveeffect favors cerebral ischemia,37 which is further aggravatedby impaired autoregulation of cerebral blood flow after therupture of the intracranial aneurysm.

In conclusion, our results propose the use of TNF-�inhibitors as a novel therapeutic strategy against cerebralvasospasm in humans. This translation is strongly facilitatedby the fact that these drugs are already used in clinical practicein the treatment of several inflammatory diseases.38,39

PerspectivesIn this study, we identified a novel therapeutic target againstcerebral vasoconstriction after SAH, a condition associatedwith elevated blood pressure levels. In particular, we showedthat an inflammatory cytokine, TNF-�, mediated the delete-rious effects of hemolyzed blood on vessels, both in vitro andin vivo. Neutralization of TNF-� by administration of inflix-imab, a TNF-� antibody used in clinical practice, was able toprevent and resolve cerebral vasospasm in a murine model.

Vecchione et al TNF-� and Cerebral Vasospasm 155

by guest on February 18, 2016http://hyper.ahajournals.org/Downloaded from

Thus, future research will be aimed at evaluating the efficacyof this treatment in patients with SAH, which might limit thislethal consequence of hypertension.

Source of FundingThis work was partly supported by the Italian Ministry of Health.

DisclosuresNone.

References1. Schievink WI. Intracranial aneurisms. New Eng J Med. 1997;336:28–40.2. Longstreth WT Jr, Nelson LM, Loepsell TD, van Belle G. Clinical course

of spontaneous subarachnoid hemorrhage; a population-based study inKing Country, Washington. Neurology. 1993;43:712–718.

3. Hop JW, Rinkel GJ, Algra A, van Gijn J. Quality of life in patients andpartners after aneurysmal subarachnoid hemorrhage. Stroke. 1998;29:798–804.

4. Charpentier C, Charpentier C, Audibert G, Guillemin F, Civit T, DucrocqX, Bracard S, Hepner H, Picard L, Laxenaire MC. Multivariate analysisof predictors of cerebral vasospasm occurrence after aneurysmal sub-arachnoid hemorrhage. Stroke. 1999;30:1402–1408.

5. Vajkoczy P, Horn P, Bauhuf C, Munch E, Hubner U, Ing D, Thome C,Poeckler-Schoeninger C, Roth H, Schmiedek P. Effect of intra-arterialpapaverine on regional cerebral blood flow in hemodynamically relevantcerebral vasospasm. Stroke. 2001;32:498–505.

6. Sen J, Belli A, Albon H, Morgan L, Petzold A, Kitchen N. Triple-Htherapy in the management of aneurysmal subarachnoid haemorrhage.Lancet Neurol. 2003;2:614–621.

7. Pluta RM, Dejam A, Grimes G, Gladwin MT, Oldfield EH. Nitriteinfusions to prevent delayed cerebral vasospasm in a primate model ofsubarachnoid hemorrhage. JAMA. 2005;293:1477–1484.

8. Sehba FA, Bederson JB. Mechanisms of acute brain injury after sub-arachnoid hemorrhage. Neurol Res. 2006;28:381–398.

9. Cahill J, Calvert JW, Zhang JH. Mechanisms of early brain injury aftersubarachnoid hemorrhage. J Cereb Blood Flow Metab. 2006;26:1341–1353.

10. Macdonald RL, Weir BK. A review of hemoglobin and the pathogenesisof cerebral vasospasm. Stroke. 1991;22:971–982.

11. Hasan RN, Schafer AI. Hemin upregulates Egr-1 expression in vascularsmooth muscle cells via reactive oxygen species ERK-1/2-Elk-1 andNF-kappaB. Circ Res. 2008;102:42–50.

12. Vecchione C, Aretini A, Marino G, Bettarini U, Poulet R, Maffei A,Sbroggio M, Pastore L, Gentile MT, Notte A, Iorio L, Hirsch E, TaroneG, Lembo G. Selective Rac-1 inhibition protects from diabetes-inducedvascular injury. Circ Res. 2006;98:218–225.

13. Vecchione C, Patrucco E, Marino G, Barberis L, Poulet R, Aretini A,Maffei A, Gentile MT, Storto M, Azzolino O, Brancaccio M, Colussi GL,Bettarini U, Altruda F, Silengo L, Tarone G, Wymann MP, Hirsch E,Lembo G. Protection from angiotensin II-mediated vasculotoxic andhypertensive response in mice lacking PI3K�. J Exp Med. 2005;201:1217–1228.

14. Sadrzadeh SM, Eaton JW. Hemoglobin-mediated oxidant damage to thecentral nervous system requires endogenous ascorbate. J Clin Invest.1988;82:1510–1515.

15. McFaul SJ, Bowman PD, Villa VM. Hemoglobin stimulates the release ofproinflammatory cytokines from leukocytes in whole blood. J Lab ClinMed. 2000;135:263–269.

16. Dumont AS, Dumont RJ, Chow MM, Lin CL, Calisaneller T, Ley KF.Cerebral vasospasm after subarachnoid hemorrhage: putative role ofinflammation. Neurosurgery. 2003;53:123–135.

17. Matz PG, Sundaresan S, Sharp FR, Weinstein PR. Induction of HSP70 inrat brain following subarachnoid hemorrhage produced by endovascularperforation. J Neurosurg. 1996;85:138–145.

18. Li J, Li W, Liu W, Altura BT, Altura BM. Mechanisms of hydroxylradical-induced contraction of rat aorta. Eur J Pharmacol. 2004;499:171–178.

19. Sehba FA, Schwartz AY, Chereshnev I, Bederson JB. Acute decrease incerebral nitric oxide levels after subarachnoid hemorrhage. J Cereb BloodFlow Metab. 2000;20:604–611.

20. Nishihashi T, Trandafir CC, Wang A, Ji X, Shimizu Y, Kurahashi K.Hypersensitivity to hydroxyl radicals in rat basilar artery after subarach-noid hemorrhage. J Pharmacol Sci. 2006;100:234–236.

21. Cosentino F, Sill JC, Katusic ZS. Role of superoxide anions in themediation of endothelium dependent contractions. Hypertension. 1994;23:229–235.

22. Jernigan NL, Walker BR, Resta TC. Reactive oxygen species mediateRhoA/Rho kinase-induced Ca2� sensitization in pulmonary vascularsmooth muscle following chronic hypoxia. Am J Physiol Lung Cell MolPhysiol. 2008;295:L515–L529.

23. Miller AA, Drummond GR, Schmidt HH, Sobey CG. NADPH oxidaseactivity and function are profoundly greater in cerebral versus systemicarteries. Circ Res. 2005;97:1055–1062.

24. Didion SP, Faraci FM. Effects of NADH and NADPH on superoxidelevels and cerebral vascular tone. Am J Physiol Heart Circ Physiol.2002;282:H688–H695.

25. Shin HK, Lee JH, Kim KY, Kim CD, Lee WS, Rhim BY, Hong KW.Impairment of autoregolatory vasodilation by NAD(P)H oxidase-dependent superoxide generation during acute stage of subarachnoidhemorrhage in rat pial artery. J Cereb Blood Flow Met. 2002;22:869–877.

26. McGirt MJ, Parra A, Sheng H, Higuchi Y, Oury TD, Laskowitz DT,Pearistein RD, Warner DS. Attenuation of cerebral vasospasm aftersubarachnoid hemorrhage in mice overexpressing extracellularsuperoxide dismutase. Stroke. 2002;33:2317–2323.

27. Zheng JS, Zhan RY, Zheng SS, Zhou YQ, Tong Y, Wan S. Inhibition ofNADPH oxidase attenuates vasospasm after experimental subarachnoidhemorrhage in rats. Stroke. 2005;36:1059–1064.

28. Mathiesen T, Edner G, Ulfarsson E, Andersson B. Cerebrospinal fluidinterleukin-1 receptor antagonist and tumor necrosis factor-alpha fol-lowing subarachnoid hemorrhage. J Neurosurg. 1997;87:215–220.

29. Figueiredo RT, Fernandez PL, Mourao-Sa DS, Porto BN, Dutra FF, AlvesLS, Oliveira MF, Oliveira PL, Graca-Souza AV, Bozza MT. Character-ization of heme as activator of Toll-like receptor 4. J Biol Chem. 2007;282:20221–20229.

30. Zhou ML, Wu W, Ding YS, Zhang FF, Hang CH, Wang HD, Cheng HL,Yin HX, Shi JX. Expression of Toll-like receptor 4 in the basilar arteryafter experimental subarachnoid hemorrhage in rabbits: a preliminarystudy. Brain Res. 2007;1173:110–116.

31. Vila E, Salaices M. Cytokines and vascular reactivity in resistancearteries. Am J Physiol Heart Circ Physiol. 2005;288:H1016–H1021.

32. Kai Y, Hirano K, Maeda Y, Nishimura J, Sasaki T, Kanaide H. Pre-vention of the hypercontractile response to thrombin by proteinase-acti-vated receptor-1 antagonist in subarachnoid hemorrhage. Stroke. 2007;38:3256–3265.

33. Xie A, Aihara Y, Bouryi VA, Nikitina E, Jahromi BS, Zhang ZD,Takahashi M, Macdonald RL. Novel mechanism of endothelin-1 inducedvasospasm after subarachnoid hemorrhage. J Cereb Blood Flow Metab.2007;27:1692–1701.

34. Hansen-Scwartz J, Ansar S, Edvinsson L. Cerebral vasoconstriction aftersubarachnoid hemorrhage-role of changes in vascular receptor phenotype.Front Biosci. 2008;13:2160–2164.

35. Harada T, Mayberg MR. Inhibiton of delayed arterial narrowing by theiron-chelating agent deferoxamine. J Neurosurg. 1992;77:763–767.

36. Haley EC Jr, Kassell NF, Alves WM, Weir BK, Hansen CA. Phase II trialof tirilazad in aneurysmal subarachnoid hemorrhage: a report of theCooperative Aneurysm Study. J Neurosurg. 1995;82:786–790.

37. Pluta RM, Oldfield EH, Boock RJ. Reversal and prevention of cerebralvasospasm by intracarotid infusions of nitric oxide donors in a primatemodel of subarachnoid hemorrhage. J Neurosurg. 1997;87:746–751.

38. Scott DL, Kingsley GH. Tumor necrosis factor inhibitors for rheumatoidarthritis. New Eng J Med. 2006;355:704–712.

39. Sands BE, Anderson FH, Bernstein CN, Chey WY, Feagan BG, FedorakRN, Kamm MA, Korzenik JR, Lashner BA, Onken JE, Rachmilewitz D,Rutgeerts P, Wild G, Wolf DC, Marsters PA, Travers SB, Blank MA, vanDeventer SJ. Infliximab maintenance therapy for fistulizing Crohn’sdisease. New Eng J Med. 2004;350:876–885.

156 Hypertension July 2009

by guest on February 18, 2016http://hyper.ahajournals.org/Downloaded from

Gian Paolo Cantore and Giuseppe LemboGiovanna Antenucci, Giada Mascio, Carla Letizia Busceti, Antonella Notte, Angelo Maffei,

Maria Teresa Gentile, Rosa Carangi, Alessandro Landolfi, Pierluigi Carullo, Umberto Bettarini, Carmine Vecchione, Alessandro Frati, Alba Di Pardo, Giuseppe Cifelli, Daniela Carnevale,

Vasospasm Evoked by Subarachnoid Hemorrhage Mediates Hemolysis-Induced Vasoconstriction and the CerebralαTumor Necrosis Factor-

Print ISSN: 0194-911X. Online ISSN: 1524-4563 Copyright © 2009 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Hypertension doi: 10.1161/HYPERTENSIONAHA.108.128124

2009;54:150-156; originally published online May 26, 2009;Hypertension. 

http://hyper.ahajournals.org/content/54/1/150World Wide Web at:

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

http://hyper.ahajournals.org/content/suppl/2009/05/26/HYPERTENSIONAHA.108.128124.DC1.htmlData Supplement (unedited) at:

  http://hyper.ahajournals.org//subscriptions/

is online at: Hypertension 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 this process is available in the

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

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

by guest on February 18, 2016http://hyper.ahajournals.org/Downloaded from

ONLINE SUPPLEMENT

TNFalpha mediates haemolysis-induced vasoconstriction and the cerebral

vasospasm evoked by subarachnoid hemorrhage

Carmine Vecchione, Alessandro Frati, Alba Di Pardo, , Giuseppe Cifelli, Daniela Carnevale, Maria Teresa Gentile, Rosa Carangi, Alessandro Landolfi, Pierluigi Carullo, Umberto Bettarini, Giovanna Antenucci, Giada Mascio, Carla Letizia Busceti, Antonella Notte, Angelo Maffei, Gian Paolo Cantore, Giuseppe Lembo.

Animals 95 male C57BL6 mice were used (10-12 weeks old; Charles River). Additional experiments were performed using age and weight-matched p47phox–/– (KO, n=5) mice on a C57Bl/6 background (WT, n=5). Mice were anesthetized by i.p. injection of xylazine/ketamine (20mg/kg and 110mg/kg ; Phoenix Pharmaceuticals). All experimental procedures were approved by our institutional review committee and were in accordance with the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals.

Blood and cerebral-spinal fluid Cerebral-spinal fluid (CSF) samples and blood were taken from hydrocephalus normotensive patients for therapeutical reasons and analysis. They were analyzed (cytochemical and cultural analysis) to ensure absence of infections. Blood was haemolyzed by shaking for several minutes and centrifugation for 10 min at 3000 rpm, and left at room temperature. The study was approved by our institutional review committee and the subjects gave informed consent.

Evaluation of vascular reactivity After anesthesia, mice were decapitated and carotid and basilar arteries were dissected out, cleaned of adhering perivascular tissue, and placed in cold Krebs-Henseleit buffer (mmol/L: NaCl 118.3, KCl 4.7, CaCl2 2.5, MgSO4·7H2O 1.2, KH2PO4 1.2, NaHCO3 25, glucose 5.6). Carotid arteries were mounted on either pressure or wire myograph. Basilar arteries were mounted on wire myograph. Vascular reactivity was tested as previously described (1-2) and vasoconstriction has been expressed as percentage of maximum K+ (80mmol/L)-evoked force.

Evaluation of Rac-1 activity Given the great amount of tissue needed to characterize molecular events in our study we used carotid arteries. Rac-1 activity has been evaluated as previously described (1-2).

Evaluation of oxidative stress in vessels Analysis of superoxide production was assessed by dihydroethidium (DHE) assays as previously described (1-2).

Adenoviral infection of carotid artery Pressure myograph was used to allow adenoviral infection of carotid arteries, as previously described (1-2). Carotid arteries were infected with 109 pfu/mL AdN17, an adenoviral vector containing a dominant negative mutant of Rac-1 and Green Fluorescent Protein (GFP) as a reporter gene, or with a vector containing only GFP (Ad0) as control, as previously described (1-2). This experimental procedure was impossible to perform, in our experimental system, in basilar artery for the small size of the vessel.

Evaluation of TNFα expression in blood and vessels TNFα expression was evaluated by both real-time PCR and Western blotting. Lymphocytes were isolated from whole and haemolyzed blood and total RNA was extracted using TRIzol reagent (Invitrogen). Total RNA (100 ng) from each sample was transcribed into cDNA using the RT-PCR Superscript III kit (Invitrogen). 2 µl (10% of reverse transcription reaction) of each cDNA preparation were subsequently used as template for PCR containing 1 µmol/L of TNFα primers and 7.875 µl of SYBR green PCR master mix (Applied Biosystems). Real-time PCR was performed using an ABI Prism 7500 Sequence Detection System (Applied Biosystems) under the following conditions: 50°C for 2 minutes; 95°C for 5 minutes; 40 cycles 95°C for 45 seconds; and 58°C for 1 minute. TNFα gene expression levels were determined using the Relative Quantification (∆∆Ct) Study of 7500 System SDS Software (Applied Biosystems). Efficiencies of the real time primers were previously tested by PCR. Western blot analysis was performed on proteins obtained before and after blood haemolysis. Protein content was measured using the Bradford method. Aliquots were subjected to SDS-PAGE, transferred onto nitrocellulose membrane and probed with mouse anti-TNFα (Chemicom, International, CA) overnight at 4°C and with anti-mouse secondary antibody (Amersham). To normalize for protein quantity, membranes were incubated with anti β-actin antibody. TNFα expression was evaluated also in vascular tissue. Briefly, carotid arteries were treated with whole blood, haemolyzed blood and LPS as positive control. The presence of the cytokine was evaluated 40 minutes after treatment by western blot as above described. Murine model of SAH After anesthesia, mice were placed in a prone position with the head flexed by approximately 30°, and the atlanto-occipital membrane was exposed. A 1.5 cm linear incision was performed on the posterior scalp of the mouse, on the midline about 5mm above and 10mm below a horizontal plane joining the external auditory meatus. The right femoral artery was exposed and cannulated with a polyethylene catheter (PE-10) and 0.04 ml of autologous blood was withdrawn. An equivalent volume of normal saline solution was replaced i.p. after blood removal. Subsequently exposed atlanto-occipital membrane was punctured with a 30-gauge steel needle, and 0.04 ml of cerebrospinal fluid was aspirated percutaneously from the cisterna magna, and then 0.04 ml of blood was slowly injected at rate of 1.2 ml/h for 2min by infusion pumps (Harvard apparatus) in cisterna magna through a PE-10 catheter, placed in a horizontal position and blocked in situ by fibrin glue tissucol (Baxter). Saline solution was injected in control mice. The animals were then tilted with tail up for 10 min in order to diffuse the blood into the subarachnoid space. Then, the catheter was cut just out of the entry point end, which was previously closed with a microclip to avoid the possible escape of fluid during removal of the tube. Finally, the occipital muscles were sutured and the skin was closed. 2,3,5-Triphenyltetrazolium chloride and hematoxylin-eosin staining for the evaluation of cerebral tissue damage 2,3,5-triphenyltetrazolium chloride (TTC) staining was used to differentiate viable tissue from that with impaired viability (3). After 5 days from SAH, brains were quickly isolated, placed in cold PBS and sectioned in to serial 1 mm-thick coronal slices. Isolation and slicing of the brain were completed within 10 min after the mouse had been decapitated. The brain slices were transferred in

TTC incubation medium (0.05% w/v in PBS) for 30 min at 37°C. After staining, slices were washed in PBS and photographed; red colour intensity was assessed by a computer assisted image analysis system (Spot, Universal Imaging). Brain cross-sections (10 µm in thickness) were assessed as described above. Classical hematoxylin-eosin staining was performed to evaluate neuronal damage. Cell death quantification was obtained by counting pycnotic cells showing an intense nuclear condensation, and a rounded shape, in 600µm distant rostro-caudal brain serial sections of the two treated groups (Bregma levels from –2.06 to –3.16). Data are expressed as the number of pycnotic nuclei per mm2. Evaluation of vascular diameter To evaluate the in vivo onset and development of vasospasm after the injection of haemolyzed blood, we monitored the anterior cerebral artery (ACA) by ultrasound analysis using a high-resolution imaging system (Vevo 770; VisualSonics) equipped with a 55 MHz transducer (4). 15 minutes after injection, the mice were anesthetized and laid prone on a platform with all legs taped. The head was placed on the left side. To reveal the anterior cerebral artery the probe was oriented to the anterior side of mouse skull, along the axis connecting the eye and the ear. Vessel diameter was recorded using an internal computer assisted measurement analysis system 30, 60, 120 minutes, and 2 and 5 days after injection of haemolyzed blood. Basilar artery diameter was evaluated ex vivo. Two hours after the injection of haemolyzed blood, mice were subjected to intracardiac perfusion-fixation. In particular, they were perfused with 0,1 mol/L PBS, followed by 25 ml of 4% paraformaldehyde at a flow rate of 2,5 ml/min for 10min. The brain was removed, immersed in the same fixative overnight at 4 °C, rinsed, dehydrated and embedded in paraffin. Cross-sections (5 µm in thickness) of the brain were cut (Leica RM2245 microtome), and placed on microscope slides for hematoxylin-eosin staining to evaluate structural changes in basilar artery. Measurement of vessel diameter was performed using a computer assisted image analysis system (Spot, Universal imaging). Infusion of anti-TNFα antibody (Infliximab) in mice Administration of Infliximab (Schering Plough), or a non-immune IgG as control, was performed in the model of SAH by either intra-arterial approach (3mg/kg) through the cannulated femoral artery before or 60 min after the injection of haemolyzed blood, or by i.p. injection (4.5mg/kg) during surgical procedures. Statistical analysis Results are shown as mean ± SEM. Data were analyzed by Student’s t-test or 2-way ANOVA followed by Bonferroni post-hoc analysis, as appropriate, using SPSS 14.0 software.

References

1. Vecchione C, Aretini A, Marino G, Bettarini U, Poulet R, Maffei A, Sbroggiò M, Pastore L, Gentile MT, Notte A, Iorio L, Hirsch E, Tarone G, Lembo G. Selective Rac-1 inhibition protects from diabetes-induced vascular injury. Circ Res. 2006;98:218-225.

2. Vecchione C, Patrucco E, Marino G, Barberis L, Poulet R, Aretini A, Maffei A, Gentile MT, Storto M, Azzolino O, Brancaccio M, Colussi GL, Bettarini U, Altruda F, Silengo L, Tarone G, Wymann MP, Hirsch E, Lembo G. Protection from angiotensin II-mediated vasculotoxic and hypertensive response in mice lacking PI3K . J Exp Med. 2005;201:1217–1228.

3. Chiamulera C, Terron A, Reggiani A, Cristofori P. Qualitative and quantitative analysis of the progressive cerebral damage after middle cerebral artery occlusion in mice. Brain Res. 1993;606:251-258.

4. Zacchigna L, Vecchione C, Notte A, Cordenonsi M, Dupont S, Maretto S, Cifelli G, Ferrari A, Maffei A, Fabbro C, Braghetta P, Marino G, Selvetella G, Aretini A, Colonnese C, Bettarini U, Russo G, Soligo S, Adorno M, Bonaldo P, Volpin D, Piccolo S, Lembo G, Bressan GM. Emilin1 links TGF-beta maturation to blood pressure homeostasis. Cell. 2006;124:929-942.